NLP-covid-19 -Text Classification.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"NLP-covid-19 -Text Classification.ipynb","provenance":[],"collapsed_sections":[],"authorship_tag":"ABX9TyNXGsLxu2WZF/BIoMXUKNGD"},"kernelspec":{"name":"python3","display_name":"Python 3"}},"cells":[{"cell_type":"code","metadata":{"id":"hKYbCXDyooKe","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1615007418929,"user_tz":-300,"elapsed":899,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"be743abf-c5a7-4b11-cc34-efd4319db853"},"source":["from google.colab import drive\n","drive.mount(\"/content/gdrive\")"],"execution_count":null,"outputs":[{"output_type":"stream","text":["Drive already mounted at /content/gdrive; to attempt to forcibly remount, call drive.mount(\"/content/gdrive\", force_remount=True).\n"],"name":"stdout"}]},{"cell_type":"markdown","metadata":{"id":"_eG0iXWfhe-H"},"source":["## Download data from kagle"]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"qwmfrNfChQ9Y","executionInfo":{"status":"ok","timestamp":1613492356490,"user_tz":-300,"elapsed":8547,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"233d83e0-529d-4712-d561-16cff686bd72"},"source":["!pip install kaggle\n","!pip install pyunpack\n","!pip install patool"],"execution_count":null,"outputs":[{"output_type":"stream","text":["Requirement already satisfied: kaggle in /usr/local/lib/python3.6/dist-packages (1.5.10)\n","Requirement already satisfied: six>=1.10 in /usr/local/lib/python3.6/dist-packages (from kaggle) (1.15.0)\n","Requirement already satisfied: python-slugify in /usr/local/lib/python3.6/dist-packages (from kaggle) (4.0.1)\n","Requirement already satisfied: certifi in /usr/local/lib/python3.6/dist-packages (from kaggle) (2020.12.5)\n","Requirement already satisfied: python-dateutil in /usr/local/lib/python3.6/dist-packages (from kaggle) (2.8.1)\n","Requirement already satisfied: requests in /usr/local/lib/python3.6/dist-packages (from kaggle) (2.23.0)\n","Requirement already satisfied: urllib3 in /usr/local/lib/python3.6/dist-packages (from kaggle) (1.24.3)\n","Requirement already satisfied: tqdm in /usr/local/lib/python3.6/dist-packages (from kaggle) (4.41.1)\n","Requirement already satisfied: text-unidecode>=1.3 in /usr/local/lib/python3.6/dist-packages (from python-slugify->kaggle) (1.3)\n","Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests->kaggle) (2.10)\n","Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests->kaggle) (3.0.4)\n","Requirement already satisfied: pyunpack in /usr/local/lib/python3.6/dist-packages (0.2.2)\n","Requirement already satisfied: easyprocess in /usr/local/lib/python3.6/dist-packages (from pyunpack) (0.3)\n","Requirement already satisfied: entrypoint2 in /usr/local/lib/python3.6/dist-packages (from pyunpack) (0.2.3)\n","Requirement already satisfied: argparse in /usr/local/lib/python3.6/dist-packages (from entrypoint2->pyunpack) (1.4.0)\n","Requirement already satisfied: patool in /usr/local/lib/python3.6/dist-packages (1.12)\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"bWXYuCUHh2sa"},"source":["#upload the credentials\n","from google.colab import files\n","files.upload()"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"nH51Nq6WiI2l","executionInfo":{"status":"ok","timestamp":1613492358316,"user_tz":-300,"elapsed":3668,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"71597555-92a1-4a40-cca4-865574f0616f"},"source":["\n","#The Kaggle API client expects the json file to be in ~/.kaggle folder so let’s create a new folder and move it inside.\n","!mkdir -p ~/.kaggle\n","!cp kaggle.json ~/.kaggle/\n","!chmod 600 ~/.kaggle/kaggle.json\n","!pip install --upgrade --force-reinstall --no-deps kaggle"],"execution_count":null,"outputs":[{"output_type":"stream","text":["Processing /root/.cache/pip/wheels/3a/d1/7e/6ce09b72b770149802c653a02783821629146983ee5a360f10/kaggle-1.5.10-cp36-none-any.whl\n","Installing collected packages: kaggle\n","  Found existing installation: kaggle 1.5.10\n","    Uninstalling kaggle-1.5.10:\n","      Successfully uninstalled kaggle-1.5.10\n","Successfully installed kaggle-1.5.10\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"eyDkiN6ViiAe","executionInfo":{"status":"ok","timestamp":1613492364736,"user_tz":-300,"elapsed":1519,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"521ccfd1-0824-4e4b-9401-1eb978b30daa"},"source":["!kaggle datasets download -d datatattle/covid-19-nlp-text-classification\n","#Let’s see the imported files:\n","!ls"],"execution_count":null,"outputs":[{"output_type":"stream","text":["covid-19-nlp-text-classification.zip: Skipping, found more recently modified local copy (use --force to force download)\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"y-jU0ubDoLGY"},"source":["#put file name and the dic to store file \n","from pyunpack import Archive\n","Archive('covid-19-nlp-text-classification.zip').extractall('/content/gdrive/MyDrive/Entity_extarction_aftab/experi-different-model-aftab')"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":669},"id":"5YVz0BreomBK","executionInfo":{"status":"ok","timestamp":1615027737333,"user_tz":-300,"elapsed":956,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"f0b036c9-7831-469d-c010-4e655a99ef31"},"source":["import pandas as pd\n","import pandas as pd\n","df_movies_train = pd.read_csv(\"/content/gdrive/MyDrive/Entity_extarction_aftab/experi-different-model-aftab/Corona_NLP_train.csv\",encoding = \"ISO-8859-1\",usecols=['OriginalTweet','Sentiment'])\n","df_movies_train.head(20)\n"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>OriginalTweet</th>\n","      <th>Sentiment</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>@MeNyrbie @Phil_Gahan @Chrisitv https://t.co/i...</td>\n","      <td>Neutral</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>advice Talk to your neighbours family to excha...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Coronavirus Australia: Woolworths to give elde...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>My food stock is not the only one which is emp...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Me, ready to go at supermarket during the #COV...</td>\n","      <td>Extremely Negative</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>As news of the regionÂs first confirmed COVID...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>6</th>\n","      <td>Cashier at grocery store was sharing his insig...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>7</th>\n","      <td>Was at the supermarket today. Didn't buy toile...</td>\n","      <td>Neutral</td>\n","    </tr>\n","    <tr>\n","      <th>8</th>\n","      <td>Due to COVID-19 our retail store and classroom...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>9</th>\n","      <td>For corona prevention,we should stop to buy th...</td>\n","      <td>Negative</td>\n","    </tr>\n","    <tr>\n","      <th>10</th>\n","      <td>All month there hasn't been crowding in the su...</td>\n","      <td>Neutral</td>\n","    </tr>\n","    <tr>\n","      <th>11</th>\n","      <td>Due to the Covid-19 situation, we have increas...</td>\n","      <td>Extremely Positive</td>\n","    </tr>\n","    <tr>\n","      <th>12</th>\n","      <td>#horningsea is a caring community. LetÂs ALL ...</td>\n","      <td>Extremely Positive</td>\n","    </tr>\n","    <tr>\n","      <th>13</th>\n","      <td>Me: I don't need to stock up on food, I'll jus...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>14</th>\n","      <td>ADARA Releases COVID-19 Resource Center for Tr...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>15</th>\n","      <td>Lines at the grocery store have been unpredict...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>16</th>\n","      <td>????? ????? ????? ????? ??\\r\\r\\n?????? ????? ?...</td>\n","      <td>Neutral</td>\n","    </tr>\n","    <tr>\n","      <th>17</th>\n","      <td>@eyeonthearctic 16MAR20 Russia consumer survei...</td>\n","      <td>Neutral</td>\n","    </tr>\n","    <tr>\n","      <th>18</th>\n","      <td>Amazon Glitch Stymies Whole Foods, Fresh Groce...</td>\n","      <td>Extremely Positive</td>\n","    </tr>\n","    <tr>\n","      <th>19</th>\n","      <td>For those who aren't struggling, please consid...</td>\n","      <td>Positive</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["                                        OriginalTweet           Sentiment\n","0   @MeNyrbie @Phil_Gahan @Chrisitv https://t.co/i...             Neutral\n","1   advice Talk to your neighbours family to excha...            Positive\n","2   Coronavirus Australia: Woolworths to give elde...            Positive\n","3   My food stock is not the only one which is emp...            Positive\n","4   Me, ready to go at supermarket during the #COV...  Extremely Negative\n","5   As news of the regionÂs first confirmed COVID...            Positive\n","6   Cashier at grocery store was sharing his insig...            Positive\n","7   Was at the supermarket today. Didn't buy toile...             Neutral\n","8   Due to COVID-19 our retail store and classroom...            Positive\n","9   For corona prevention,we should stop to buy th...            Negative\n","10  All month there hasn't been crowding in the su...             Neutral\n","11  Due to the Covid-19 situation, we have increas...  Extremely Positive\n","12  #horningsea is a caring community. LetÂs ALL ...  Extremely Positive\n","13  Me: I don't need to stock up on food, I'll jus...            Positive\n","14  ADARA Releases COVID-19 Resource Center for Tr...            Positive\n","15  Lines at the grocery store have been unpredict...            Positive\n","16  ????? ????? ????? ????? ??\\r\\r\\n?????? ????? ?...             Neutral\n","17  @eyeonthearctic 16MAR20 Russia consumer survei...             Neutral\n","18  Amazon Glitch Stymies Whole Foods, Fresh Groce...  Extremely Positive\n","19  For those who aren't struggling, please consid...            Positive"]},"metadata":{"tags":[]},"execution_count":108}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"cCygr2-H7V3_","executionInfo":{"status":"ok","timestamp":1615026372668,"user_tz":-300,"elapsed":3927,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"954ecae9-e2f7-4efe-c472-e23b3a4349d2"},"source":["!pip install contractions"],"execution_count":null,"outputs":[{"output_type":"stream","text":["Requirement already satisfied: contractions in /usr/local/lib/python3.7/dist-packages (0.0.48)\n","Requirement already satisfied: textsearch>=0.0.21 in /usr/local/lib/python3.7/dist-packages (from contractions) (0.0.21)\n","Requirement already satisfied: pyahocorasick in /usr/local/lib/python3.7/dist-packages (from textsearch>=0.0.21->contractions) (1.4.1)\n","Requirement already satisfied: anyascii in /usr/local/lib/python3.7/dist-packages (from textsearch>=0.0.21->contractions) (0.1.7)\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"4eBa3QLgsjbY","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1615026372680,"user_tz":-300,"elapsed":2095,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"31eff334-66b6-4e01-a584-73e72135ee92"},"source":["import numpy as np\n","from matplotlib import pyplot as plt\n","import numpy as np\n","import pandas as pd\n","import os\n","import glob\n","import re\n","import nltk\n","import contractions\n","nltk.download('punkt')\n","nltk.download('stopwords')\n","nltk.download('wordnet')\n","from nltk.corpus import stopwords\n","from nltk.corpus import wordnet\n","from nltk.stem import WordNetLemmatizer\n","lemmatizer = WordNetLemmatizer() \n","from gensim.models import Word2Vec"],"execution_count":null,"outputs":[{"output_type":"stream","text":["[nltk_data] Downloading package punkt to /root/nltk_data...\n","[nltk_data]   Package punkt is already up-to-date!\n","[nltk_data] Downloading package stopwords to /root/nltk_data...\n","[nltk_data]   Package stopwords is already up-to-date!\n","[nltk_data] Downloading package wordnet to /root/nltk_data...\n","[nltk_data]   Package wordnet is already up-to-date!\n"],"name":"stdout"}]},{"cell_type":"markdown","metadata":{"id":"HRfXV0Rtqag_"},"source":["#Data analysis"]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":433},"id":"V7HKVF3HqTiK","executionInfo":{"status":"ok","timestamp":1615026422399,"user_tz":-300,"elapsed":1393,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"1cfc1766-5fd0-4349-81cc-9e36d20c83e7"},"source":["values=df_movies_train['Sentiment'].value_counts()\n","class_size=values.tolist()\n","\n","plt.style.use(\"fivethirtyeight\")\n","\n","# Language Popularity\n","slices = class_size\n","labels = ['Positive','Negative','Neutral','Extremely Positive','Extremely Negative']\n","\n","fig = plt.figure(1, figsize=(7,6))\n","ax = fig.add_subplot(111)\n","ax.axis('equal')\n","expode=[0,0,0,0.1,0]\n","\n","ax.pie(slices,labels=labels,wedgeprops={'edgecolor': 'black'},\n","        autopct='%1.1f%%',explode=expode,shadow=True, startangle=90,)\n","\n","plt.title(label=\"samples distribution\", \n","          loc=\"left\", \n","          fontstyle='italic',\n","          fontsize=20\n","          ) \n","plt.tight_layout()\n","\n","plt.show()\n"],"execution_count":null,"outputs":[{"output_type":"display_data","data":{"image/png":"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\n","text/plain":["<Figure size 504x432 with 1 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"nYzPaJNUhgAM","executionInfo":{"status":"ok","timestamp":1615026507618,"user_tz":-300,"elapsed":873,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"0714951e-5546-4bd2-ce44-0ca399f713f5"},"source":["#check null value\n","df_movies_train.isna().sum()\n"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["OriginalTweet    0\n","Sentiment        0\n","dtype: int64"]},"metadata":{"tags":[]},"execution_count":78}]},{"cell_type":"code","metadata":{"id":"rkXxz12VqT2y"},"source":["#check duplicate values\n","def remove_duplicate(dataframe):\n","  duplicate = dataframe[dataframe.duplicated()]\n","  if len(duplicate) !=0:\n","    dataframe.drop_duplicates(keep='first',inplace=True)\n","  else:\n","    pass\n","  return dataframe\n","df_movies_train=remove_duplicate(df_movies_train)"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"ClBtVD-MzVvQ","executionInfo":{"status":"ok","timestamp":1615026515405,"user_tz":-300,"elapsed":813,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"237ab7c4-72d4-4bdc-f3e7-a43808283d4f"},"source":["label_datf=df_movies_train['Sentiment'].copy()\n","label_datf.head(10)"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0               Neutral\n","1              Positive\n","2              Positive\n","3              Positive\n","4    Extremely Negative\n","5              Positive\n","6              Positive\n","7               Neutral\n","8              Positive\n","9              Negative\n","Name: Sentiment, dtype: object"]},"metadata":{"tags":[]},"execution_count":80}]},{"cell_type":"code","metadata":{"id":"uyDKO9LvzWAU","colab":{"base_uri":"https://localhost:8080/","height":669},"executionInfo":{"status":"ok","timestamp":1615031625354,"user_tz":-300,"elapsed":3002,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"e49dedd9-4f0d-45f5-cef7-7d1406f439be"},"source":["# merg postive and Extremely Positive & Negative and Extremely Negative\n","dictionary={'Neutral':'Neutral','Positive':'Positive','Extremely Positive':'Positive','Negative':'Negative','Extremely Negative':'Negative'     \n","}\n","df_movies_train['Sentiment']=df_movies_train['Sentiment'].map(dictionary)\n","\n","df_movies_train.head(20)"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>OriginalTweet</th>\n","      <th>Sentiment</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>@MeNyrbie @Phil_Gahan @Chrisitv https://t.co/i...</td>\n","      <td>Neutral</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>advice Talk to your neighbours family to excha...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Coronavirus Australia: Woolworths to give elde...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>My food stock is not the only one which is emp...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Me, ready to go at supermarket during the #COV...</td>\n","      <td>Negative</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>As news of the regionÂs first confirmed COVID...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>6</th>\n","      <td>Cashier at grocery store was sharing his insig...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>7</th>\n","      <td>Was at the supermarket today. Didn't buy toile...</td>\n","      <td>Neutral</td>\n","    </tr>\n","    <tr>\n","      <th>8</th>\n","      <td>Due to COVID-19 our retail store and classroom...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>9</th>\n","      <td>For corona prevention,we should stop to buy th...</td>\n","      <td>Negative</td>\n","    </tr>\n","    <tr>\n","      <th>10</th>\n","      <td>All month there hasn't been crowding in the su...</td>\n","      <td>Neutral</td>\n","    </tr>\n","    <tr>\n","      <th>11</th>\n","      <td>Due to the Covid-19 situation, we have increas...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>12</th>\n","      <td>#horningsea is a caring community. LetÂs ALL ...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>13</th>\n","      <td>Me: I don't need to stock up on food, I'll jus...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>14</th>\n","      <td>ADARA Releases COVID-19 Resource Center for Tr...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>15</th>\n","      <td>Lines at the grocery store have been unpredict...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>16</th>\n","      <td>????? ????? ????? ????? ??\\r\\r\\n?????? ????? ?...</td>\n","      <td>Neutral</td>\n","    </tr>\n","    <tr>\n","      <th>17</th>\n","      <td>@eyeonthearctic 16MAR20 Russia consumer survei...</td>\n","      <td>Neutral</td>\n","    </tr>\n","    <tr>\n","      <th>18</th>\n","      <td>Amazon Glitch Stymies Whole Foods, Fresh Groce...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>19</th>\n","      <td>For those who aren't struggling, please consid...</td>\n","      <td>Positive</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["                                        OriginalTweet Sentiment\n","0   @MeNyrbie @Phil_Gahan @Chrisitv https://t.co/i...   Neutral\n","1   advice Talk to your neighbours family to excha...  Positive\n","2   Coronavirus Australia: Woolworths to give elde...  Positive\n","3   My food stock is not the only one which is emp...  Positive\n","4   Me, ready to go at supermarket during the #COV...  Negative\n","5   As news of the regionÂs first confirmed COVID...  Positive\n","6   Cashier at grocery store was sharing his insig...  Positive\n","7   Was at the supermarket today. Didn't buy toile...   Neutral\n","8   Due to COVID-19 our retail store and classroom...  Positive\n","9   For corona prevention,we should stop to buy th...  Negative\n","10  All month there hasn't been crowding in the su...   Neutral\n","11  Due to the Covid-19 situation, we have increas...  Positive\n","12  #horningsea is a caring community. LetÂs ALL ...  Positive\n","13  Me: I don't need to stock up on food, I'll jus...  Positive\n","14  ADARA Releases COVID-19 Resource Center for Tr...  Positive\n","15  Lines at the grocery store have been unpredict...  Positive\n","16  ????? ????? ????? ????? ??\\r\\r\\n?????? ????? ?...   Neutral\n","17  @eyeonthearctic 16MAR20 Russia consumer survei...   Neutral\n","18  Amazon Glitch Stymies Whole Foods, Fresh Groce...  Positive\n","19  For those who aren't struggling, please consid...  Positive"]},"metadata":{"tags":[]},"execution_count":160}]},{"cell_type":"code","metadata":{"id":"x58Ole9VzWQK","colab":{"base_uri":"https://localhost:8080/","height":433},"executionInfo":{"status":"ok","timestamp":1615026607702,"user_tz":-300,"elapsed":1206,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"3303464e-dd4a-4591-87af-a0fa58eecc15"},"source":["# samples distribution among differnt classes \n","values=df_movies_train['Sentiment'].value_counts()\n","class_size=values.tolist()\n","\n","plt.style.use(\"fivethirtyeight\")\n","\n","# Language Popularity\n","slices = class_size\n","labels = ['Positive','Negative','Neutral']\n","\n","fig = plt.figure(1, figsize=(7,6))\n","ax = fig.add_subplot(111)\n","ax.axis('equal')\n","expode=[0,0,0.1]\n","\n","ax.pie(slices,labels=labels,wedgeprops={'edgecolor': 'black'},\n","        autopct='%1.1f%%',explode=expode,shadow=True, startangle=90,)\n","\n","plt.title(label=\"samples distribution\", \n","          loc=\"left\", \n","          fontstyle='italic',\n","          fontsize=20\n","          ) \n","plt.tight_layout()\n","plt.show()"],"execution_count":null,"outputs":[{"output_type":"display_data","data":{"image/png":"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\n","text/plain":["<Figure size 504x432 with 1 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"markdown","metadata":{"id":"xksY5wxe5QBd"},"source":["## Data preprocessing"]},{"cell_type":"code","metadata":{"id":"YLRt583CqUJU"},"source":["import string\n","def data_pre_processing(data):\n","  corpus = []\n","  fillerWord = (\"so\",\"yeah\",\"okay\",\"um\",\"uh\",\"mmm\",\"ahan\",\"uh\",\"huh\",\"ahm\",\"oh\",\"sooo\",\"uh\",\"huh\",\"yeh\",\"yah\",\"hmm\",\"bye\")\n","  fillerword_reg= \"bye[.,]|so[.,]|yeah[.,]|okay[.,]|um[.,]|uh[.,]|mmm[.,]|ahan[.,]|uh[.,]|huh[.,]|ahm[.,]|oh[.,]|sooo[.,]|uh[.,]|huh[.,]|yeh[.,]|yah[.,]|hmm[.,]\"\n","  STOPWORDS = set(stopwords.words('english'))\n","  STOPWORDS.add(fillerWord)  \n","  for i in range(len(data)):\n","      review = re.sub(\"#\", \"\", data[i])\n","      review = re.sub(r'^https?:\\/\\/.*[\\r\\n]*', '',review, flags=re.MULTILINE)\n","      html=re.compile(r'<.*?>')\n","      review =html.sub(r\"\", review)\n","      review=re.sub(r'@\\w+',\"\",review)\n","      review=re.sub(r'#\\w+',\"\",review) \n","      review=re.sub(r'\\w+:\\/{2}[\\d\\w-]+(\\.[\\d\\w-]+)*(?:(?:\\/[^\\s/]*))*', '', review) \n","      punctuation = r\"\"\"!\"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~\"\"\"\n","      review.rstrip(string.punctuation)\n","      review=re.sub('[^A-Za-z\\s]+',\"\", review)\n","      review = review.lower()\n","      review = review.split()\n","      review = [lemmatizer.lemmatize(word) for word in review if not word in STOPWORDS]\n","      review = ' '.join(review)\n","      filler=re.compile(fillerword_reg)\n","      review=filler.sub(\"\",review)\n","      corpus.append(review)\n","  return corpus\n","\n"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"XIU4wV99Sp3D"},"source":["# data pre-processing for tranning \n","text_data=df_movies_train['OriginalTweet'].tolist()\n","process_text=data_pre_processing(text_data)\n","#process_text= list(filter(None, process_text))\n","df_train_process_text=pd.DataFrame(process_text)"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"6YPbfMZaii3X","colab":{"base_uri":"https://localhost:8080/","height":204},"executionInfo":{"status":"ok","timestamp":1615031644547,"user_tz":-300,"elapsed":3898,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"759b1cc5-86b1-4d70-8971-386bdcf242eb"},"source":["df_train_process_text.columns=['OriginalTweet']\n","df_train_process_text.head()"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>OriginalTweet</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td></td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>advice talk neighbour family exchange phone nu...</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>coronavirus australia woolworth give elderly d...</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>food stock one empty please dont panic enough ...</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>ready go supermarket covid outbreak im paranoi...</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["                                       OriginalTweet\n","0                                                   \n","1  advice talk neighbour family exchange phone nu...\n","2  coronavirus australia woolworth give elderly d...\n","3  food stock one empty please dont panic enough ...\n","4  ready go supermarket covid outbreak im paranoi..."]},"metadata":{"tags":[]},"execution_count":162}]},{"cell_type":"code","metadata":{"id":"48EGKUioqUYk","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1615026626858,"user_tz":-300,"elapsed":1133,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"9ec31261-23d8-406e-e8f5-382f3c9a1ac5"},"source":["#check dimensions after pre-processing of traning data \n","print(df_movies_train.shape)\n","print(df_train_process_text.shape)"],"execution_count":null,"outputs":[{"output_type":"stream","text":["(41157, 2)\n","(41157, 1)\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"uHurVrkIUtRx"},"source":["# Data concatnate \n","Data_tarning=pd.concat([df_train_process_text,df_movies_train['Sentiment']],axis=1)"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"U-IwKQiIWP8l"},"source":["#drop nun values \n","nan_value = float(\"NaN\")\n","Data_tarning.replace(\"\", nan_value, inplace=True)\n","Data_tarning.dropna(inplace=True)\n"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"_uZvu96Ch-tk"},"source":["dictionary={'Neutral':1,'Positive':2,'Negative':0}\n","Data_tarning['Sentiment']=Data_tarning['Sentiment'].map(dictionary)"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000},"id":"BuCGgbxBv2VI","executionInfo":{"status":"ok","timestamp":1615007542724,"user_tz":-300,"elapsed":929,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"f193bb18-5de4-45b1-fe94-7bd9862c1f97"},"source":["Data_tarning[250:300]\n"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>OriginalTweet</th>\n","      <th>Sentiment</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>254</th>\n","      <td>mpe prepared continue providing safe reliable ...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>255</th>\n","      <td>binge shopping online quarantinelife coronavirus</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>256</th>\n","      <td>long time could dream reid distillery would lo...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>257</th>\n","      <td>way coronavirus changing millennials money hab...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>258</th>\n","      <td>im getting personal im one allow coronavirus h...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>259</th>\n","      <td>panic shopping coronavirus planning doesnt end...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>260</th>\n","      <td>hello brother sister let connect hand support ...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>261</th>\n","      <td>social distancing prevent spread coronavirus c...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>262</th>\n","      <td>woolworth give elderly disabled dedicated shop...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>263</th>\n","      <td>man yelled supermarket today buying yucky garl...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>264</th>\n","      <td>due fear covid people started hoarding item da...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>265</th>\n","      <td>northgate market allowing senior shop first op...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>266</th>\n","      <td>people dying covid people also dying cant affo...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>267</th>\n","      <td>apparently everyone else feel way daves killer...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>268</th>\n","      <td>korea covid case malaysia covid case yet korea...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>269</th>\n","      <td>amazon say coronavirus outbreak caused surge o...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>270</th>\n","      <td>consumer response cannabisindustry covid impor...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>271</th>\n","      <td>got grocery store amp call wife bread want grab</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>272</th>\n","      <td>convid safety measure r taken online shopping ...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>273</th>\n","      <td>shop hiking price shame dont mean people cashi...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>274</th>\n","      <td>gaisss please read thisand please limit go out...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>275</th>\n","      <td>covid impacting supply chain trade workforce m...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>276</th>\n","      <td>love people return something store please let ...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>277</th>\n","      <td>well tracking effect coronavirus retail indust...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>278</th>\n","      <td>consumer alert aware bogus covid home test kit...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>279</th>\n","      <td>say need hire people across u keep crush order...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>280</th>\n","      <td>covid need panic buying food available time pm...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>281</th>\n","      <td>stopthespread research shown faster authority ...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>282</th>\n","      <td>amazes mei go supermarket everything gone peop...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>283</th>\n","      <td>u get covid supermarket letspanic</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>284</th>\n","      <td>going grocery store made panic anything scarci...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>285</th>\n","      <td>worth reading hopefully nothing else public mi...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>286</th>\n","      <td>needle say caused strong feeling coronavirus c...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>287</th>\n","      <td>san francsisco mayor london breed announce loc...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>288</th>\n","      <td>covid pandemic plunging global financial marke...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>289</th>\n","      <td>apple closed retail location outside china mar...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>290</th>\n","      <td>covid need panic buying food available time pm...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>291</th>\n","      <td>full grown as woman county highest number coro...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>292</th>\n","      <td>loving quiet tube atm covid terrible yes get f...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>293</th>\n","      <td>owhealth time turmoil allow formation deeper r...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>294</th>\n","      <td>food supply consumer welfare department odisha...</td>\n","      <td>2</td>\n","    </tr>\n","    <tr>\n","      <th>295</th>\n","      <td>oregon gov kate brown monday afternoon ordered...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>296</th>\n","      <td>coronavirus europe call calm food shortage fea...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>297</th>\n","      <td>even amazon struggling deliver food know stock...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>298</th>\n","      <td>anthony fauci said people may perceive new cor...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>299</th>\n","      <td>breaking center disease control recommended am...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>300</th>\n","      <td>iranprisoners released view coronavirusin ital...</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>301</th>\n","      <td>weve updated covid info page including link la...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>302</th>\n","      <td>pymnts surveyed k consumer covid already chang...</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>303</th>\n","      <td>people arent worried ventilator worried losing...</td>\n","      <td>0</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["                                         OriginalTweet  Sentiment\n","254  mpe prepared continue providing safe reliable ...          2\n","255   binge shopping online quarantinelife coronavirus          1\n","256  long time could dream reid distillery would lo...          2\n","257  way coronavirus changing millennials money hab...          1\n","258  im getting personal im one allow coronavirus h...          0\n","259  panic shopping coronavirus planning doesnt end...          0\n","260  hello brother sister let connect hand support ...          2\n","261  social distancing prevent spread coronavirus c...          0\n","262  woolworth give elderly disabled dedicated shop...          2\n","263  man yelled supermarket today buying yucky garl...          2\n","264  due fear covid people started hoarding item da...          0\n","265  northgate market allowing senior shop first op...          2\n","266  people dying covid people also dying cant affo...          0\n","267  apparently everyone else feel way daves killer...          0\n","268  korea covid case malaysia covid case yet korea...          0\n","269  amazon say coronavirus outbreak caused surge o...          2\n","270  consumer response cannabisindustry covid impor...          2\n","271    got grocery store amp call wife bread want grab          2\n","272  convid safety measure r taken online shopping ...          0\n","273  shop hiking price shame dont mean people cashi...          0\n","274  gaisss please read thisand please limit go out...          2\n","275  covid impacting supply chain trade workforce m...          1\n","276  love people return something store please let ...          2\n","277  well tracking effect coronavirus retail indust...          1\n","278  consumer alert aware bogus covid home test kit...          2\n","279  say need hire people across u keep crush order...          0\n","280  covid need panic buying food available time pm...          0\n","281  stopthespread research shown faster authority ...          2\n","282  amazes mei go supermarket everything gone peop...          2\n","283                  u get covid supermarket letspanic          1\n","284  going grocery store made panic anything scarci...          0\n","285  worth reading hopefully nothing else public mi...          2\n","286  needle say caused strong feeling coronavirus c...          2\n","287  san francsisco mayor london breed announce loc...          1\n","288  covid pandemic plunging global financial marke...          2\n","289  apple closed retail location outside china mar...          2\n","290  covid need panic buying food available time pm...          0\n","291  full grown as woman county highest number coro...          0\n","292  loving quiet tube atm covid terrible yes get f...          2\n","293  owhealth time turmoil allow formation deeper r...          2\n","294  food supply consumer welfare department odisha...          2\n","295  oregon gov kate brown monday afternoon ordered...          0\n","296  coronavirus europe call calm food shortage fea...          0\n","297  even amazon struggling deliver food know stock...          0\n","298  anthony fauci said people may perceive new cor...          0\n","299  breaking center disease control recommended am...          0\n","300  iranprisoners released view coronavirusin ital...          0\n","301  weve updated covid info page including link la...          1\n","302  pymnts surveyed k consumer covid already chang...          1\n","303  people arent worried ventilator worried losing...          0"]},"metadata":{"tags":[]},"execution_count":25}]},{"cell_type":"markdown","metadata":{"id":"zk5-xo_Sfuav"},"source":["## Project1: ML Model"]},{"cell_type":"code","metadata":{"id":"lQ6vrSwzZ-rU"},"source":["from sklearn.feature_extraction.text import CountVectorizer\n","# initialize our Bag of Word Model\n","vectorizer = CountVectorizer(max_features= 3000, min_df = 3, max_df = 0.6)\n","# fit the model for tranning data\n","X_data_train = vectorizer.fit_transform(Data_tarning['OriginalTweet'])\n","y_label_train=Data_tarning['Sentiment']"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"od7H_q-Vf2u3","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1615007554599,"user_tz":-300,"elapsed":1355,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"3bae4f94-21c7-44bc-ff57-d5252424d746"},"source":["from sklearn.feature_extraction.text import TfidfTransformer\n","transformer = TfidfTransformer()\n","X_train_tfid = transformer.fit_transform(X_data_train).toarray()\n","\n","print(X_train_tfid.shape)\n","print(y_label_train.shape)"],"execution_count":null,"outputs":[{"output_type":"stream","text":["(41047, 3000)\n","(41047,)\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"jcb3JA1rRIsW"},"source":["from sklearn.model_selection import train_test_split\n","X_train, X_test, y_train, y_test = train_test_split(X_train_tfid, y_label_train, test_size=0.33, random_state=42)"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"30uF0pozf2RI","executionInfo":{"status":"ok","timestamp":1615007572168,"user_tz":-300,"elapsed":1276,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"82592540-e7f4-4224-8485-d688b5dbb29e"},"source":["#NB\n","from sklearn.naive_bayes import MultinomialNB\n","classifier = MultinomialNB()\n","classifier.fit(X_train,y_train)\n","prediction = classifier.score(X_test,y_test )\n","print(prediction)"],"execution_count":null,"outputs":[{"output_type":"stream","text":["0.6684630149121512\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"ntS2ftuLbOCO","executionInfo":{"status":"ok","timestamp":1613990365927,"user_tz":-300,"elapsed":57362,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"a2624a00-d841-411d-a59c-91a185588070"},"source":["#ada boost \n","from sklearn import model_selection\n","from sklearn.ensemble import AdaBoostClassifier\n","clf = AdaBoostClassifier(n_estimators=100)\n","clf.fit(X_train,y_train)\n","prediction = clf.score(X_test,y_test )\n","print(prediction)\n"],"execution_count":null,"outputs":[{"output_type":"stream","text":["0.6633692602982431\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"Fplgy7IXm0OF","executionInfo":{"status":"ok","timestamp":1613990815380,"user_tz":-300,"elapsed":51091,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"dccb0603-d064-46df-bcc7-954224f87812"},"source":["#Gradient Boosting Classifier\n","from sklearn.ensemble import GradientBoostingClassifier \n","clf = GradientBoostingClassifier(n_estimators=100, learning_rate=1.0, max_depth=1)\n","clf.fit(X_train,y_train)\n","prediction = clf.score(X_test,y_test )\n","print(prediction)"],"execution_count":null,"outputs":[{"output_type":"stream","text":["0.7053742802303263\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"PIcf6EmeXC4B","executionInfo":{"status":"ok","timestamp":1613991074161,"user_tz":-300,"elapsed":11782,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"50cbf1af-9643-4fe7-9c82-5666ad31a683"},"source":["# RF \n","from sklearn.metrics import confusion_matrix,classification_report,accuracy_score\n","from sklearn.ensemble import RandomForestClassifier\n","rf_classifier=RandomForestClassifier(n_estimators=10).fit(X_train,y_train)\n","prediction=rf_classifier.predict(X_test)\n","\n","print(confusion_matrix(y_test,prediction))\n","print(accuracy_score(y_test,prediction))\n","print(classification_report(y_test,prediction))"],"execution_count":null,"outputs":[{"output_type":"stream","text":["[[3568  518  977]\n"," [ 651 1412  504]\n"," [1092  555 4269]]\n","0.6827845858556031\n","              precision    recall  f1-score   support\n","\n","           0       0.67      0.70      0.69      5063\n","           1       0.57      0.55      0.56      2567\n","           2       0.74      0.72      0.73      5916\n","\n","    accuracy                           0.68     13546\n","   macro avg       0.66      0.66      0.66     13546\n","weighted avg       0.68      0.68      0.68     13546\n","\n"],"name":"stdout"}]},{"cell_type":"markdown","metadata":{"id":"cDNKVcNq1oom"},"source":["### Hyper parameters tunning RF"]},{"cell_type":"code","metadata":{"id":"DdG959Ea6MVv"},"source":["import numpy as np\n","from sklearn.ensemble import RandomForestClassifier\n","from sklearn.model_selection import RandomizedSearchCV\n","\n","n_estimators = [int(x) for x in np.linspace(start = 200, stop = 2000, num = 10)]\n","max_features = ['auto', 'sqrt','log2']\n","max_depth = [int(x) for x in np.linspace(10, 1000,10)]\n","min_samples_split = [1,3,4,5,7,9,10]\n","min_samples_leaf = [1, 2, 4,6,8]\n","random_grid = {'n_estimators': n_estimators,\n","               'max_features': max_features,\n","               'max_depth': max_depth,\n","               'min_samples_split': min_samples_split,\n","               'min_samples_leaf': min_samples_leaf,\n","              'criterion':['entropy','gini']}\n","print(random_grid)\n","\n","rf=RandomForestClassifier()\n","rf_randomcv=RandomizedSearchCV(estimator=rf,param_distributions=random_grid,n_iter=50,cv=3,verbose=2,\n","                               random_state=100,n_jobs=-1)\n","### fit the randomized model\n","rf_randomcv.fit(X_train,y_train)\n","print(rf_randomcv.best_params_)\n","print(rf_randomcv)\n","#print(rf_randomcv.best_estimator_)\n","best_random_grid=rf_randomcv.best_estimator_\n","print(best_random_grid)\n"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"zHItctAH-3W2","executionInfo":{"status":"ok","timestamp":1614051994357,"user_tz":-300,"elapsed":45956,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"7c88427e-4292-43ea-f793-90024dc9a644"},"source":["\n","from sklearn.metrics import accuracy_score\n","y_pred=best_random_grid.predict(X_test)\n","print(confusion_matrix(y_test,y_pred))\n","print(\"Accuracy Score {}\".format(accuracy_score(y_test,y_pred)))\n","print(\"Classification report:{}\".format(classification_report(y_test,y_pred)))\n"],"execution_count":null,"outputs":[{"output_type":"stream","text":["[[3681  274 1108]\n"," [ 462 1333  772]\n"," [ 687  241 4988]]\n","Accuracy Score 0.7383729514247749\n","Classification report:              precision    recall  f1-score   support\n","\n","           0       0.76      0.73      0.74      5063\n","           1       0.72      0.52      0.60      2567\n","           2       0.73      0.84      0.78      5916\n","\n","    accuracy                           0.74     13546\n","   macro avg       0.74      0.70      0.71     13546\n","weighted avg       0.74      0.74      0.73     13546\n","\n"],"name":"stdout"}]},{"cell_type":"markdown","metadata":{"id":"ru306UwcMVR_"},"source":["## Project1:Deep learning model (neural network)"]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":204},"id":"c8SSX6l3F0MK","executionInfo":{"status":"ok","timestamp":1615031663816,"user_tz":-300,"elapsed":1547,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"18d895f2-c257-48b8-ca93-b30391d74ff6"},"source":["Data_tarning.head()"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>OriginalTweet</th>\n","      <th>Sentiment</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>1</th>\n","      <td>advice talk neighbour family exchange phone nu...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>coronavirus australia woolworth give elderly d...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>food stock one empty please dont panic enough ...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>ready go supermarket covid outbreak im paranoi...</td>\n","      <td>Negative</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>news region first confirmed covid case came su...</td>\n","      <td>Positive</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["                                       OriginalTweet Sentiment\n","1  advice talk neighbour family exchange phone nu...  Positive\n","2  coronavirus australia woolworth give elderly d...  Positive\n","3  food stock one empty please dont panic enough ...  Positive\n","4  ready go supermarket covid outbreak im paranoi...  Negative\n","5  news region first confirmed covid case came su...  Positive"]},"metadata":{"tags":[]},"execution_count":166}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"Dlc6jqjIYLds","executionInfo":{"status":"ok","timestamp":1615031669287,"user_tz":-300,"elapsed":792,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"47cb24ad-ca46-4514-f317-9e680a298fe4"},"source":["values=Data_tarning['Sentiment'].value_counts()\n","print(values)\n","class_size=values.tolist()"],"execution_count":null,"outputs":[{"output_type":"stream","text":["Positive    18011\n","Negative    15378\n","Neutral      7658\n","Name: Sentiment, dtype: int64\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"YxFfKQjwEZTP","executionInfo":{"status":"ok","timestamp":1615031675576,"user_tz":-300,"elapsed":805,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"56a35cf7-2614-4b19-c78a-5da98b6f5fe3"},"source":["Data_tarning['Sentiment']"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["1        Positive\n","2        Positive\n","3        Positive\n","4        Negative\n","5        Positive\n","           ...   \n","41152     Neutral\n","41153    Negative\n","41154    Positive\n","41155     Neutral\n","41156    Negative\n","Name: Sentiment, Length: 41047, dtype: object"]},"metadata":{"tags":[]},"execution_count":168}]},{"cell_type":"code","metadata":{"id":"tWfQYzTM2AqM"},"source":["# Categorical label encoding \n","#train_labe_nn=pd.get_dummies(Data_tarning['Sentiment'],drop_first=True)\n","train_labe_nn=pd.get_dummies(Data_tarning['Sentiment'])\n","train_labe_nn.head()"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"G4UYAMASAHc-"},"source":["# Count Vectorizer and TFIDF\n","from sklearn.feature_extraction.text import CountVectorizer\n","vectorizer = CountVectorizer(max_features=1000, min_df = 3, max_df = 0.6)\n","X_data_train = vectorizer.fit_transform(Data_tarning['OriginalTweet'])\n","y_label_train=train_labe_nn\n","from sklearn.feature_extraction.text import TfidfTransformer\n","transformer = TfidfTransformer()\n","X_train_tfid = transformer.fit_transform(X_data_train).toarray()"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"cEG08etZHxg5"},"source":["from sklearn.model_selection import train_test_split\n","\n","X_train, X_test, y_train, y_test = train_test_split(X_train_tfid, y_label_train, test_size=0.2, random_state=42)\n"],"execution_count":null,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"tihY5zpJaEFo"},"source":["### class weight for data inbalance "]},{"cell_type":"code","metadata":{"id":"_6KPAFx6ZEJA"},"source":["class_weights = {0:1.5,  \n","                  1:2.5,\n","                 2:1,\n","                 }"],"execution_count":null,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"rXF3pTdTNB38"},"source":["####Model\n","\n"]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"f6IHgrCj2A2_","executionInfo":{"status":"ok","timestamp":1615031911843,"user_tz":-300,"elapsed":75562,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"8fe5bf8c-7559-4ec1-e5b2-0d9f91b44ed9"},"source":["\n","import keras\n","from keras.models import Sequential\n","from keras.layers import Dense\n","from keras.layers import Dropout\n","#adding the input and first hidden layer\n","classifier = Sequential()\n","classifier.add(Dense(50, kernel_initializer='uniform', activation='relu',input_dim=1000))\n","classifier.add(Dropout(0.3))\n","\n","#adding the second hidden layer\n","classifier.add(Dense(30, kernel_initializer='uniform', activation='relu'))\n","classifier.add(Dropout(0.2))\n","classifier.add(Dense(10, kernel_initializer='uniform', activation='relu'))\n","classifier.add(Dropout(0.2))\n","\n","#adding the output layer\n","classifier.add(Dense(3, kernel_initializer='uniform', activation='softmax'))\n","\n","classifier.compile(optimizer=\"Adam\", loss='kullback_leibler_divergence', metrics=['accuracy'])\n","\n","classifier.fit(X_train,y_train, batch_size=30, epochs=30,class_weight=class_weights)\n"],"execution_count":null,"outputs":[{"output_type":"stream","text":["Epoch 1/30\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.4645 - accuracy: 0.5116\n","Epoch 2/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 1.1403 - accuracy: 0.6818\n","Epoch 3/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 1.0343 - accuracy: 0.7110\n","Epoch 4/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.9809 - accuracy: 0.7338\n","Epoch 5/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.9425 - accuracy: 0.7469\n","Epoch 6/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.9010 - accuracy: 0.7625\n","Epoch 7/30\n","1095/1095 [==============================] - 3s 2ms/step - loss: 0.8889 - accuracy: 0.7654\n","Epoch 8/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.8358 - accuracy: 0.7822\n","Epoch 9/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.8135 - accuracy: 0.7886\n","Epoch 10/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.7909 - accuracy: 0.7962\n","Epoch 11/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.7529 - accuracy: 0.8062\n","Epoch 12/30\n","1095/1095 [==============================] - 3s 3ms/step - loss: 0.7385 - accuracy: 0.8106\n","Epoch 13/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.7223 - accuracy: 0.8163\n","Epoch 14/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.7002 - accuracy: 0.8236\n","Epoch 15/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.6727 - accuracy: 0.8280\n","Epoch 16/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.6582 - accuracy: 0.8303\n","Epoch 17/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.6359 - accuracy: 0.8374\n","Epoch 18/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.6274 - accuracy: 0.8390\n","Epoch 19/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.6044 - accuracy: 0.8474\n","Epoch 20/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.5849 - accuracy: 0.8458\n","Epoch 21/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.5820 - accuracy: 0.8491\n","Epoch 22/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.5645 - accuracy: 0.8547\n","Epoch 23/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.5358 - accuracy: 0.8598\n","Epoch 24/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.5362 - accuracy: 0.8635\n","Epoch 25/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.5320 - accuracy: 0.8640\n","Epoch 26/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.5330 - accuracy: 0.8602\n","Epoch 27/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.5198 - accuracy: 0.8662\n","Epoch 28/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.4919 - accuracy: 0.8746\n","Epoch 29/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.4986 - accuracy: 0.8744\n","Epoch 30/30\n","1095/1095 [==============================] - 2s 2ms/step - loss: 0.4755 - accuracy: 0.8741\n"],"name":"stdout"},{"output_type":"execute_result","data":{"text/plain":["<tensorflow.python.keras.callbacks.History at 0x7f5932661c50>"]},"metadata":{"tags":[]},"execution_count":173}]},{"cell_type":"code","metadata":{"id":"Gopzib6G2BGQ","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1615031920724,"user_tz":-300,"elapsed":2840,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"e197bf93-db54-4ee8-9bb7-bbeed9eaeac3"},"source":["y_pred = classifier.predict(X_test)\n","y_pred = (y_pred > 0.5)\n","# Making the Confusion Matrix\n","from sklearn.metrics import classification_report\n","print('\\nClassification Report\\n')\n","print(classification_report(y_test.values.argmax(axis=1), y_pred.argmax(axis=1), target_names=['Class Negative', 'Class Neutral','Class postive']))"],"execution_count":null,"outputs":[{"output_type":"stream","text":["\n","Classification Report\n","\n","                precision    recall  f1-score   support\n","\n","Class Negative       0.69      0.74      0.72      3084\n"," Class Neutral       0.56      0.64      0.59      1549\n"," Class postive       0.81      0.71      0.75      3577\n","\n","      accuracy                           0.71      8210\n","     macro avg       0.69      0.69      0.69      8210\n","  weighted avg       0.72      0.71      0.71      8210\n","\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"FtcVU28l2Bus","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1615031929298,"user_tz":-300,"elapsed":859,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"ae69d07c-1c0c-45cc-de04-89d9cf606cd7"},"source":["from sklearn.metrics import confusion_matrix\n","#cm = confusion_matrix(y_test, y_pred)\n","cm =confusion_matrix(y_test.values.argmax(axis=1), y_pred.argmax(axis=1))\n","cm"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["array([[2290,  382,  412],\n","       [ 374,  984,  191],\n","       [ 644,  404, 2529]])"]},"metadata":{"tags":[]},"execution_count":175}]},{"cell_type":"code","metadata":{"id":"sET6Dcfz1-IG","colab":{"base_uri":"https://localhost:8080/","height":301},"executionInfo":{"status":"ok","timestamp":1615031935086,"user_tz":-300,"elapsed":1055,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"bcb1b0de-3ac0-42bc-cc22-a70adedb7986"},"source":["import seaborn as sns\n","sns.heatmap(cm, annot=True)"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["<matplotlib.axes._subplots.AxesSubplot at 0x7f58e16556d0>"]},"metadata":{"tags":[]},"execution_count":176},{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAZEAAAELCAYAAAAY3LtyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdeVzM+R/A8VclpKJkCusmCRGRc62b3EfEWmzLD+WMRCzWsRTWTe77XAmxNsvKumstNkTumy6FUo6a3x/tTjurQ6NRrffz8ZjHw3yO7/c9pXnP5/h+Ryc2NlaJEEIIoQHdnA5ACCFE3iVJRAghhMYkiQghhNCYJBEhhBAakyQihBBCY5JEhBBCaEySiBBCCI1JEhFCiFxq3rx5NGvWjNKlS1OxYkWcnJwIDQ1Va+Pi4oKJiYnao2XLlmptXr16xdixY6lQoQIlS5akV69ePHz4UK3N/fv3cXJyomTJklSoUAEPDw9ev36daYz5PvxlZq7Apf0f4zTiPVh18s7pEMS/5NfVz+kQxD+cvXEk245V6E10ltq/1DdTe37ixAkGDBhA7dq1USqVzJw5ky5duhAUFISpqamqXdOmTVmxYoXqef78+dWO4+npyYEDB1izZg2mpqZMnDgRJycnfvvtN/T09EhKSsLJyQlTU1MOHDhATEwMLi4uKJVK5syZk2HMHyWJCCHEJyk56YO6+/n5qT1fsWIFZcqU4cyZMzg4OKjKCxQogIWFRZrHePbsGZs2bWLp0qU0a9ZMdRwbGxuOHj1KixYtOHLkCFeuXOHixYuUKlUKgKlTpzJixAgmTZpE4cKF041RprOEEEJblMlZe2QiLi6O5ORkTExM1MpPnz5NpUqVsLOzY8SIEURGRqrqLly4wJs3b2jevLmqrFSpUlhZWREUFARAcHAwVlZWqgQC0KJFC169esWFCxcyjElGIkIIoS3JmSeGrBg/fjw2NjbY29urylq2bEnHjh0pW7Ys9+7dY8aMGXTq1ImjR49SoEABIiIi0NPTw8xMfapMoVAQEREBQEREBAqFQq3ezMwMPT09VZv0SBIRQggtUb7H6OJ9TZgwgTNnzhAQEICenp6qvHv37qp/V6tWDVtbW2xsbDh48CCdOnXKtvOnR6azhBBCW5KTs/ZIh6enJ7t27cLf359y5cpleMoSJUpQsmRJbt26BYC5uTlJSUlER6sv8kdGRmJubq5q888pMIDo6GiSkpJUbdIjSUQIIbQlG9ZExo0bp0oglStXzvSU0dHRPH78WLXQbmtri76+PoGBgao2Dx8+JCwsjHr16gFgb29PWFiY2rbfwMBAChQogK2tbYbnk+ksIYTQlg/cneXu7s6OHTvYvHkzJiYmhIeHA2BoaIiRkRFxcXF4eXnRqVMnLCwsuHfvHtOmTUOhUNChQwcAihQpQt++fZkyZQoKhUK1xbdatWo0bdoUgObNm2Ntbc2QIUOYMWMGMTExTJ48mX79+mW4MwskiQghhPZ84JrI6tWrAejcubNa+bhx4/D09ERPT4/Q0FC2b9/Os2fPsLCw4PPPP2fdunUYGxur2s+aNQs9PT2cnZ1JTEykSZMmLF++XLW2oqenx44dO3B3d6dt27YULFiQHj16MH369Exj1PkY32woFxvmHnKxYe4jFxvmLtl5saFBdFiW2ieYWWXbuT8WGYkIIYSWZOfurNxKkogQQmhL0tucjkDrJIkIIYS2fODCel4gSUQIIbRFprOEEEJoLJtve5IbSRIRQghtkZGIEEIIjclIRAghhKaUSllYF0IIoSmZzhJCCKExmc4SQgihMRmJCCGE0FjSm5yOQOskiQghhLbIdJYQQgiNyXSWEEIIjclIRAghhMYkiQghhNCUXGwohBBCczISEUIIoTFZWBdCCKExGYkIIYTQmIxEhBBCaExGIkIIITSW9DanI9A6SSJCCKEtMhLJvdb8dIJf/7jKnSdR5M+XD5uKnzGiewssS5mn2+fmw0hmbjnArUdRxL1MRGFiTNt61XDp3BT9fHpai/V5fAJeWwP47cI1AL6wrcz4Pg4ULlQwR+PKTv0GOPFl/x6UKlMSgOtXb7L4h5UcOXQ83T5NmjXEbZwLla0r8frVa84GX2DmlHncvnlXa3EWLmLMVK/xtGzbFIDDAUeZMs6L589fAFC/UR0GuPTFtlZ1jAsbcef2fdYu38yPW/doLaaPYfBIZ9y/HcamNTuYNn52mm3yF8jP9LkTqGpThYqVy3Mu+AJfdRms9dgKFzFm0syxtGj7BQC/BvzGNM/ZvHgeB4B9Qzuch3xJjdrVMTY24u7t+2xYuRXfrf5aj+2DfQJrIro5HYCmfr96h57N6rBhwjes8uiLnq4ug+du4llcQrp99PPp0alhTZaP7sPemUPx6N2G3ccvsMTvyAfFMsB7A3tPXEi3fvwKP67cfcIyty9Z5vYlV+4+YeKq3VqP62N6/Cgcr6kLaN/MiY4tenPqeDArNy2gSlXLNNuXLvMZqzYvJPj0Odo17UmfboMoWLAA67cv/aA4tu9dg2PvTunWL17pTfUa1vTv4UL/Hi5Ur2HNfJ+Zqno7e1vCQq/j4jyGVo27sXndj8yaP5nO3dt9UFw5ydauOk59u3Ll0rUM2+np6fIq8RWb1+zg6KET2Xb+zXtW0K1Xx3Tr5634nmo1qvCN0zC+cRpGtRpVmLtsuqq+tn0Nrl25wfBvPGjfpCfb1vsy/YeJdOzWNtti1Jrk5Kw98qD3Gok8fPiQtWvXEhQUREREBADm5ubUr1+fr7/+mlKlSmk1yLQsH/OV2vOZ/+tKo6HenL9xj6a2Vmn2KWNRlDIWRVXPSxYz4fewO5y7fk+t3dELYSzfe4ybDyMoZmJMu3rVGdL5C41GBbceRXLy0k3We35NzUqlAZjUrz3OXuu58ziKciWKvXdcudmhn4+qPZ/z/WK+cu5J7bo1uRp6/Z32NjWroq+fD+/pC0n+649n2YI1bN+7BtOiJsQ8jQWgRZsvcPNwwbJKRSLDo9i76wALZvvw5k3W55orVS5P05aN6e7Qj3NnQwDwHD2dXQc2UKFSOW7duMPS+avV+mxe9yMNGtfFoWNL9u46kOVz5jQjYyN+8Pkez5HTGD72fxm2TXiZyOSxswCwqmZJ4SJGabZr3vpzhnsMxtKqAhHhUezzC2DJnJUa/U4qWpbjixaNcGr/DRfOXgRgkvtMtu9fQ/mKZbl98y7LF6xT67N1vS/1GtehTcfm7PMLyPI5PyoZicDp06epV68ee/bswdraGkdHRxwdHbG2tmbPnj00aNCAM2fOfIxYMxSf+IpkpZLChQzeu8+98KecungTu8plVWUnL91gwsrd9GpeF78ZLkx17sihs1dYtOtXjeL68+YDChXIj+1fCQSglmVpDAroc+HGg/eOKy/R1dWlY9e2FDIsxB/Bf6bZ5s/zl3jz5i29+nZDV1cXQ6NCdO/ViQvnLqoSSJNmDVm4YhYb1myjVaOujB0xGYdOrfD4dqRGcdWuU5O4uHjOBqeOGs8GnSc+7iV29jXT7WdkbMSz2OcanTOnzZg3kYB9hwk6eTZbjte4WQN+WD6DzWt20O7znniOnEbbji0YPXGYRserVbcGcXHxnPvH/5M/gi4QH/+S2vY10u1nZGzIs9gXGp3zo5KRCHh6evLll18ye3ba86jjxo3D09OTwMDAbA8uK2ZvPYhVmeLUrJT5qKjf92u5cvcxr98m0b1JbUZ0b6GqW73/BP3bNqTL57YAlDYvyqgeLZiwajeje7ZCR0cnS3FFP4vD1LiQWj8dHR2KGhsS/dec7/vElRdYWVuyO2ATBQrmJz7+JYP7jSLsyrujEICHDx7zVffBLFs7l+mzJ6Crq8vlkKv0d3JVtRk2+n+sXLyenVv3AnDvzgO8ps5ngc8svp/yQ5bjU1gU42lUzDvl0VFPUZgXS7NP89ZNaNTEnu7t+mf5fDmt51ddKVu+NO4uk7LtmK5u37B6ySZ2bdsHpPxO5kxbzFyf6Xh/tyDLxytmbkZMGr+Tp5FPKZbO76RZq89p8Lk9vdp/k+XzfXSfwEgk0yRy9epVVq1alW79gAED2LBhQ7YGlVVzth/k/PV7rPd0Rk8382We2S7diU94zbX74czbeYh1P59kQPvGAITeecylWw9Z9/NJVXulUkni67dEPYtDYWLM9I0/8dPpEFX9q9dvCbn5gFlbflaV7Z7hSgmzIll6HRnFlRfcunEbh6Y9MC5sRLtOrfhh6QycOg3g2tUb77RVmJsxe+F3+O3Yh7/fzxgaGTJ6vCtL18yhd5eBKJVKbGpWpWbt6gwZkfpmoaurg0EhA8wtihERHsX3c7+la48OqvqCBgWoVacG07wmqMpaNuzCo4dPsvx66tjbsmilF995evPnuUtZ7p+Tylcsy5iJQ+nVYQBv32bfNtNqNaypUasa/xuRmlR1dXQxKFQQhUUxIsOjmDbHk049UteQChYsgK2dDZO9PFRlDo168FiD30lt+5rMWzGDGRPmEHL+8oe9mI8hj44usiLTJGJhYcGZM2ewtEx7gfTMmTNYWFhke2Dva862gwQEX2a1Rz9KmZu+V5/iRVPe3Ct+piBJmczUdfvo37Yh+fR0USqVDOn8Ba3qVH2nn6mxIQCuXZrSv00DVbnnSj9a1rGmRW1rVZnCxBgAsyJGxLx4iVKpVI1GlEolT1/EY1ZYfc45o7jygjdv3nL39n0ALv15hZq1qjPQ5Ss8Rn73Ttt+A3qR8DKBWVPnq8pGuUwg6OIh7OxtORt0Hl1dHRbOWc5Pe395p3/0X59e53ktY+XS1A8xC1d48fO+wwTsP6wqC38SCUBkeBRFi737f8SsWFEiI6LUyurUq8X6HUuZN2sZm9f9mIWfQu5Qq24NihYz5cCJ1Njz5ctH3Qa16d2/OzXLNub166x/dauurg6L567iZ/9D79T9Pcpb4L2c1cs2qcrn+XzPwf2/cvCn1I0iEX/9TqIiojFN43dSVFGUqH/9Tuzq2bJ620IWei9n63rfLMeeIySJwPDhwxk9ejTnzp2jadOmmJunbKGNiIjg6NGjbN26lVmzZmk90LR4bw3g4F8JpHyJtIe+mVEmK0lKTk5Z3NXTxbpscW4/jlJb6P43s8KGmBU2VD0vmF+fosaGafapWbEUL1+95s+bD1TrIn/efEDCqzfYZjD19u+48iJdXV3yF8ifZp2BQUGS/vUHlpyUpOoHcCnkChUty6sSU1qio54SHfVU9TwxIZHoqOg0+5w7+ydGRobY1a3JH7+nzMHb1a2JoZH62o19AzvWbV/CfK9lrF2x+T1fbe5y6EAgFy+EqpV5LZrC3Vv38VmwVqMEAnD54lUqWJbj3u201/MgJZn8c9owMTGR6KiYNPuc/z0EIyNDatWtwfnfU0b3terWwNCwEOeCU0f7dRvUYuXWhSzyXsH6Fds0ij1HKJU5HYHWZZpEBg4cSNGiRVm2bBmbNm0i6a8/dD09PWxtbVm+fDldu3bVeqD/NnPTAfafDmH+cCcKGxoQ9SxlfaFQgfwUKpjyxrXQ91cu3X7IqrH9ANh3KoQC+vmwLGWOvp4el+88YuGuI7SsU5X8+ik/ikGdmjBi4XZKmBWhdd1q5NPT5cbDCC7deohbz1ZZjrNCSQWNqldk+oafmNy/PQDTN/xEk5qWlPsr8b1PXLnduMkjOfLLcR4/fIKhkSGdHR2o36gOzr1SFlw9Jo3AtrYNX3ZN2SF05NBxBrj0ZYT7YPz9fsbIyJCx347g4YPHXPwz5c1v4ZwVrN22mIf3H7F/zy8kJb3FqoolNWtXVxvBvK8b125z9PAJZs6bjKfbVABmzpvM4YDfuHXjDpBynci6bUvZtHYHe3cdQGFuBkBSUjJPo9+du8+tXjyPU11n8beElwnExj7j+tWbAIz5dhg1alWjf3cXVZtKlcujn18f06ImFDIshHX1ygCq7cFL5q5m5ZYFPLr/mAN7D5H0NonK1hWpUasas6ctynKcN6/f4bdfTzJ97kS+HTMDgOlzJ3Lk4DHV9UL2De1YtXUhW9ftZJ9fAMX++p0kJyXxNDo2y+f8qLJxKjG3eq93qG7dutGtWzfevHlDdHQ0AGZmZujr62s1uIzsCEzZbTJozia18iGdmuDSpSkAUc/ieBCR+oefT1eXNT+d4F74U5QoKWFmQq/mdfiqdX1Vm0bVK7F4ZG9W7jvGxoOn0dPVpayFGZ0ap797JzNeg7sxa0sALvO2APCFrRWeXzlkKa7cTmFejAXLZ6IwL8aL53FcDb1G/56uHAs8BYC5hYIy5VJHXqeOBzNi0HgGD/uaIcOdSUxM5PzZi/Tv6ULCy5RrfY4FnsK51zBGuA9m0ND+vE1K4vbNu/hu26txnMMHjWOatycbfZcDKRcbTvZIHUk79u5MIUMDBg//msHDv1aV37/3kMa1HP59uDzN3KKY2u8EYNW2RaoLRgH8A1M+9Vsq7AA4EXiaQV+OxHXMQAa49iUp6S23b97Db/s+jeMYPXgik2d5sO7HJQD8GnCMqeO9VfXde3ekkKEBA4f1Y+CwfqryB/ce0cwu/etPcoVPYGFdJzY2VuvjrQKX9mv7FOI9WXXyzryR+Kjy6+bchzHxrrM3su8i3/x7vbLU/nXn8dl27o8lb8yVCCFEXiRrIkIIITQmu7OEEEJoTJKIEEIIjX0CC+t58wIEIYTIA5TJyiw9/m3evHk0a9aM0qVLU7FiRZycnAgNVb/+R6lUMmvWLKpUqULx4sVp3749V65cUWsTGxvLoEGDKFOmDGXKlGHQoEHExqpvj758+TLt2rWjePHiWFtb4+3tjfI91nQkiQghhLZ84A0YT5w4wYABAzh48CD+/v7ky5ePLl26EBOTeunCwoULWbp0Kd7e3hw5cgSFQkHXrl158SL1BpUDBw4kJCQEX19ffH19CQkJYfDg1O+Kef78OV27dsXc3JwjR47g5eXF4sWLWbJkSaYvUaazhBBCWz5wOsvPz0/t+YoVKyhTpgxnzpzBwcEBpVKJj48Po0aNonPnzgD4+PhgaWmJr68vzs7OhIWFcfjwYQICArC3twdg/vz5ODg4cP36dSwtLdm5cycJCQn4+PhgYGBA1apVuXbtGsuWLWPYsGEZ3nhWRiJCCKEtb5Oy9shEXFwcycnJmJiYAHD37l3Cw8Np3ry5qo2BgQENGzYkKCgIgODgYIyMjKhXr56qTf369TE0NFRr06BBAwwMUr9Ko0WLFjx+/Ji7dzP+plFJIkIIoS3Z/H0i48ePx8bGRjWiCA8PB0ChUKi1UygUqi8QjIiIwMzM7J2voyhWrJham7SO8XddRmQ6SwghtCUbLzacMGECZ86cISAgAD29rH/LqrbISEQIIbQlm0Yinp6e7Nq1C39/f8qVK6cq//trOCIjI9XaR0ZGqu64bm5uTnR0tNpOK6VSSVRUlFqbtI7xd11GJIkIIYS2JCuz9kjDuHHjVAmkcuXKanVly5bFwsJC7ZtlExMTVV9rDmBvb09cXBzBwcGqNsHBwcTHx6u1OX36NImJiao2gYGBlChRgrJlM/6abkkiQgihLcrkrD3+xd3dna1bt7Jq1SpMTEwIDw8nPDycuLiU2/zr6Ojg4uLCwoUL8ff3JzQ0FFdXVwwNDXF0dATAysqKli1b4ubmRnBwMMHBwbi5udGmTRvVlw06OjpiYGCAq6sroaGh+Pv7s2DBAlxdXTP9SnBZExFCCG1JZ3TxvlavXg2g2r77t3HjxuHp6QnAyJEjSUhIYOzYscTGxmJnZ4efnx/GxsZqx/Hw8KB79+4AODg4MHv2bFV9kSJF2L17N+7u7jRr1gwTExOGDh3KsGHDMo1RbgX/iZFbwec+civ43CU7bwWfb9mILLV/65r1L/bKaTISEUIIbfnAkUheIElECCG05RO4AaMkESGE0Jb3uAo9r5MkIoQQ2iLTWUIIITQm01lCCCE0JiMRIYQQmlLK1+MKIYTQmIxEhBBCaEySiBBCCI3JwroQQgiNyUhECCGEppSSRIQQQmhMkogQQgiNyW1PhBBCaExGIkIIITT1z+81/6+SJCKEENoiI5HsId+ml3sMMaqR0yGIf5n0ODCnQxDaIklECCGEpmSLrxBCCM1JEhFCCKGx//5dTySJCCGEtmR1OktHS3FokyQRIYTQFpnOEkIIoSnlWxmJCCGE0JSsiQghhNCUbPEVQgihORmJCCGE0NQn8MWGkkSEEEJrJIkIIYTQlIxEhBBCaE6SiBBCCE3JSEQIIYTGJIkIIYTQmDIpL16DnjWSRIQQQktkJCKEEEJjymQZiQghhNCQjESEEEJoTKmUkYgQQggNfQojEd2cDkAIIf6rlMk6WXqk5eTJk/Tq1Qtra2tMTEzYsmWLWr2LiwsmJiZqj5YtW6q1efXqFWPHjqVChQqULFmSXr168fDhQ7U29+/fx8nJiZIlS1KhQgU8PDx4/fp1pq9RkogQQmiJUpm1R1ri4+OpWrUqXl5eGBgYpNmmadOmhIWFqR47d+5Uq/f09GTfvn2sWbOGAwcO8OLFC5ycnEhKSgIgKSkJJycn4uLiOHDgAGvWrMHf35+JEydm+hplOksIIbQkO3ZntW7dmtatWwPg6uqaZpsCBQpgYWGRZt2zZ8/YtGkTS5cupVmzZgCsWLECGxsbjh49SosWLThy5AhXrlzh4sWLlCpVCoCpU6cyYsQIJk2aROHChdONT0YiQgihJdkxnfU+Tp8+TaVKlbCzs2PEiBFERkaq6i5cuMCbN29o3ry5qqxUqVJYWVkRFBQEQHBwMFZWVqoEAtCiRQtevXrFhQsXMjy3jESEEEJLkj/CFestW7akY8eOlC1blnv37jFjxgw6derE0aNHKVCgABEREejp6WFmZqbWT6FQEBERAUBERAQKhUKt3szMDD09PVWb9EgSEUIILfkYW3y7d++u+ne1atWwtbXFxsaGgwcP0qlTJ62fX6azhBBCS5TJWXtkhxIlSlCyZElu3boFgLm5OUlJSURHR6u1i4yMxNzcXNXmn1NgANHR0SQlJanapEeSiBBCaEmyUidLj+wQHR3N48ePVQvttra26OvrExgYqGrz8OFDwsLCqFevHgD29vaEhYWpbfsNDAykQIEC2NraZng+mc4SQggtyY7prLi4ONWoIjk5mQcPHhASEoKpqSmmpqZ4eXnRqVMnLCwsuHfvHtOmTUOhUNChQwcAihQpQt++fZkyZQoKhQJTU1MmTpxItWrVaNq0KQDNmzfH2tqaIUOGMGPGDGJiYpg8eTL9+vXLcGcWSBIRQgityY4tvufPn6djx46q57NmzWLWrFn07t2befPmERoayvbt23n27BkWFhZ8/vnnrFu3DmNjY7U+enp6ODs7k5iYSJMmTVi+fDl6enoA6OnpsWPHDtzd3Wnbti0FCxakR48eTJ8+PdP4dGJjY9O5xCX71KjwubZPId7TEKMaOR2C+JdJjwMzbyQ+mujI0Gw71qM6vbPUvuTZbdl27o9FRiJCCKEln8Kt4PPswnq/AU4EHPPl0p1TXLpzit0Bm2jeKv0RzygPF+5Gh6T5MCtWVGtx5s+vz1Sv8Zy/9htX7gWxevMiipdMvbLUulplFq305nTIL4Q9COZIkD+Dhzujo5P3/vPlNyxIy8lf4XpyAe5ha+nrN5kSNSpk2Kd8Exv67Z7C6MurGHneh+6r3ChavrhW4yxYuBAd5w/B7eJK3C6upOP8IRQoXEhVX6a+Nd1XuTH89yW4X13DgICZ1OjZRKsxacPnjeux228dd2+f5e3rh/Tr2zPTPo6OHTn7+y88j73BzetBjBk9ROtxmpgUYf26RURHXiE68grr1y2iSJHUefgvmjTAb9da7t89x/PYG5z74xBf93fSelzZIScW1j+2PJtEHj8Kx2vqAto3c6Jji96cOh7Myk0LqFLVMs32K5eup451M7XH6RO/c/p4MNFRTzWOY+6S6YzycEm3fvLMcTh0bMnwQePo0eFrjIwNWbt1Mbq6KT96m5pVeRr9FDeXCbRs1JX5XssY4T4I11EDNI4ppzh4D6T8FzbsH7OCNa09uX3sEr22jMfIwjTN9kVKK3Bc5cb94DDWtfuWbX1moV8wPz3Xj/2gOL7cPhEbx/Q/UHRaPBSL6uX4sf9sfuw/G4vq5eg4P/XN8jM7SyLD7uPnspDVrcZzbvOvOMwaQNXODT4oro/NyMiQy5fDcBszhZcvEzJt37ZNMzZvXMLq1VuoWas5w0dMYOSI/+Hq8vUHxfHroZ0ZJrDNG5dQq1Z12nf4ivYdvqJWrepsWL9IVd+gQR0uXbqKU69B1KzVnBUrNrHcZza9enX5oLg+BqVSJ0uPvCjPTmcd+vmo2vM53y/mK+ee1K5bk6uh199p/zI+gZfxqX9IJUpaYN+gNm4u6jcYs6tbE49JI6lZqxrPnj3nUMBveE2dT9yL+CzHaGxshFOfrowdPokTR88A4OYygVN/HqTxF/U5FniKH7fuUetz/+5DqtewxqFDS5bOX53lc+aUfAX0qeJQF78hC7l35goAJxb4UallLWr3bcGxub7v9CluUw5d/Xwc9d6BMjllae7UMn/6bJ+IgakRCTFxAFRqUYvGbt1QWH5GXGQsl/ee5sQCP5LfJGU5TrNKJanYtCYbu0/l4bkbAAR4rqXvrskUrVCCp7cec3qpv1qf85t/pWyDqlg51CV07+ksnzOn/BxwhJ8DjgCwdvX8TNv36dOd/T8dYvmKDQDcvn0P79lLGOs+lGU+61XtOrRvxaRJo6lWtTKPH0ewfccepk2fx5s3b7IcY5UqlWjbtjlNvujMmaA/AHB1HcdvR/dQuXJFrl27iZf3YrU+K1ZupGnThnTr2o7t2/ekddhcI72bKv6X5NmRyD/p6urSsWtbChkW4o/gP9+rj9NX3XgW+5yf9x1SlVlZW7LJdwWHA47S9oseDO4/mmrVrZizaJpGcdnYViV/fn2OBaa+8Tx+FM6Na7ews09/77WRsSHPnj3X6Jw5RTefHrr59Hj7Sv2N5G3ia0rVsUqzz+M/b5H85i01ezVDR1eH/IYFsen+OY8u3FQlkPJNbOi00JU/NhxiVavx/DR2FVUc7GnqkfnUTFo+q12JV3EJPDyb+nlw6dgAACAASURBVEHjwdlrvI5P5DO7tEexAAWMDEh89lKjc+YVBQrkJzHxlVpZQkIipUuXpGzZlHsqtW71BRs3LGbZsnXUsG3O/waNoVu39nw/fbxG56xfz44XL+I4dfqsquzkqd+Ji4unQX27dPsVNjYiJuaZRuf8mJKSdbP0yIuyJeoHDx4wdOjQ7DhUllhZWxJ69wzXH5/l+x++ZXC/UYRdeXcU8m+6urr07NMFvx/38/p16pve4OFfs39PAKuWbeTOrXtc+OMiE91n0K5TK43WTRTmxXj79i1Po2PUyqMin6IwN0uzT/Ua1jj27szmdT9m+Xw56XV8Ig/+uEaj4V0wsjBFR1eHal0b8VltS4zMTdLs8/xhNNu+8uZzt254XF/P6EsrUVQpzc5vflC1aTisM0Erf+LizmPE3ovg3ukrBHptp1afFhrFaagw4eXTF++Ux0c/x0hRJM0+lZrbUrZRVS5sPaLROfOKX375jU4d29CqZRN0dHSwtKyAm9tgAEoUT1nH8xw/gh/mLWfDxh+5desuR387heeE7xk0qK9G5yxe3JzIyOh3yiMioihePO0rpdu3a0nz5o1ZvXpLmvW5SXbcCj63y5bprJiYGLZt28bSpUuz43Dv7daN2zg07YFxYSPadWrFD0tn4NRpANeu3siwX9MWjfisVAm2b9qlVm5T05qy5cvQoUtbVdnfC9xly5UiOuopQ90GMnTUQFV9/gL6KJUwaGh/VVl/J1d+P3Muy6+nQqVyrNu2hLUrNvPzvsNZ7p/T9o1aTvs5/2N48GKS3ybx5NIdQv1PU9ymXJrtDRVFaD97IJf8ThDqf5r8hgX5fLQjXZYOZ2vvmaBUUtymHCVrVqD+kA6qfjq6OugbFMDQ3IT4iFjafO9M9a6NVPX5Cubns1oVaT0t9XeyquU4nj96980qM5/VsaTToqEc+m4Tj/+8leX+ecnqNVuoWLEsfrvWoq+vz/PnL1i8ZA1TJruTnJxyT47atWtQt64tY91Tb0muq6tLoUIGFC9uzpMnESxd4kWfL7up6g0MClKvXm0WLZyhKrOp2ZT79x9lOcaGDeqwaeMSRrlN5vezGd9dNjfIq4vlWfFeSWTbtoz3Lj948CBbgsmqN2/ecvf2fQAu/XmFmrWqM9DlKzxGfpdhv979HTkbdJ7rYepvCrq6uuzY7Mdqn03v9HnyOOVOlpvX/cj+PQdV5Z5T3HjyOJx1K7e+0zYyIop8+fJR1MxUbTRSTFGU4NPqSaaiZTm271mD/+4AvKctfI9Xn/vE3otgi9P36BsUIL+xAfERsXReMozYe5Fptrfr14rXCa8InLVdVbZvlA/DghZRys6SB2evoaOry4mFu7n6U/A7/V9Gp0z5HZ+3i+CVB1TlnRa6cPXn37kWkDpF8iI85ecfHxlLoaLG/JuhWWHiItWnR0rVqUzP9WM5Ps+X85t/zcJPIu/ynDCTid96qUYILZo3BuDW7bsA6OrqMH3GfHx37X+n798jiu+mzmHe/OWq8k0bFuO3+wC79/ysKnv0KByAJ08iUCjeHZWbmxfjyRP1u8c2aliXff6b+G7qXFas3PiBr/TjyKuL5VnxXknE1dWVQoUKpbvt9O9PKTlNV1eX/AXyZ9jGvLiC5q0+Z9yo796pu/TnFSytKqoSU1qexT7nWWzqekVcXDyxMc/T7HPxQiivX7/h86YN2Lsr5U2ueEkLKlWuwB/BqZ+iLK0qsHX3an7ae5Dp387J7GXmem8SXvEm4RUFCxeiQhMbtSTxT/kM8qNMUv+/8/f/JR3dlP9r4ZfuYFaxJDF3w9M938vo56qEAvAm8Q0vo5+n2efhuRsUMDLgMztLHv6RMvX5mZ0l+Q0Lqp4DlLa3osc6d47P9+P3tQffOc5/WXJyMo8ePQHAyakLp0+fJeqvHYznz1/CyqoSN2/eSbd/ZGS02hRVQkIiERHRafY5E/QHxsZGNKhfh9NnUpJ+g/p1MDIy5PSZP1TtPm9cD/+9G5k67QcWLc47G05kJPKXEiVKqO7PkpaQkBDVPVg+lnGTR3Lkl+M8fvgEQyNDOjs6UL9RHZx7DQPAY9IIbGvb8GXX/6n1c+rThZcvE9i/55d3jumzaC17Dm7m+7nfsnWDL3Fx8VS0LE/LNl8wYUzml///24sXcezYshvP79yIioom9ukzJk1358rla5z4LWW3lqVVRbbtWc3pE8Esnb9aba0kMiLr0y85qXwTG3R0dYi++RjTshY0n9Cb6JuPCdl5DIAvPHpS0rYi276cBcDNIxewH9CWRiO6pExnGRnQdGxPnj2M4snFOwCcWLibHmvH8OxhFFf2B5GclITCqhQla1ZMNzllJPrGI24e/ZO2M78hwHMNAG1nfsP1w+d4eusxkHKdSI91Yzi36VdC957C8K+1kuSkZBLSWE/JrQwNC1GpUnkg5QNWmTIlqVmzGk+fxnD//iO+nzGeunVq0bptyjUXZmamOHbvyG/HTlGgQAH69+uJY/f2NG/hqDrmjO/ns3fPBu7de8BO3328ffuWatWqYF/XlvGe32c5xqtXbxAQcIRly7xwcRkHwLJlXuz/6RDXrt0EUq4T8d+7keXLN7Bt+24sLFK+9yIpKUmV3HKrPLrMkSXvlURq1qxJSEhIuklER0cH5UdeFVKYF2PB8pkozIvx4nkcV0Ov0b+nK8cCTwFgbqGgTLlS7/Rz6tONvb4HSExIfKfuauh1enR0xn3CcHbsW4uerh737j7g4E+aT2VMm+BN0tu3LF09h4IFC3DyWDBurhNVn7jbd26NwtyMTt0c6NTNQa1vWbO8dYuSAsaFaDquJ8bFi5L4LJ6wn4P5bc5Okt+mbMU1MjfBpEzqYundU6HsHbGMeoPbU39IB94mvubh+Rvs6D+bNwkpu4RuH7vITue5NBrRlXqD2pH8Npmntx9z0fe4xnH6D19Kq2n9cdqY8qZ1/fA5fpm8QVVv4/g5+QsVpP7g9tQf3F5VHns/Ep/Gbhqf92OrY1eTXw+nbq3+bspYvpsylg0bf2TAQDeKF7egQoWyan36fuWIt9e36OjocObMH7Ro2UNt7eGXQ7/RqXM/Jk4YxWi3Ibx9+5Zr12+xcaPmG0G+6jeMhQumc+CnlIXyfft/YcTIb1X1/fr1xNCwEGPGuDBmTOo1WXfu3KdS5foan/dj+BRGIu9176xTp04RHx9Pq1at0qyPj4/n/PnzNG7cOM16uXdW7iH3zsp95N5ZuUt23jvrklXWLhquHrYm2879sbzXSKRhw4YZ1hsaGqabQIQQ4lOVO1aLtSvPXrEuhBC5nZL//nSWJBEhhNCSt5/AmogkESGE0BIZiQghhNCYrIkIIYTQmIxEhBBCaExGIkIIITQmSUQIIYTGZDpLCCGExpL/+zlEkogQQmhLsoxEhBBCaEru4iuEEEJjb9P5Dqb/EkkiQgihJTISEUIIoTHZ4iuEEEJjsjtLCCGExmR3lhBCCI3JmogQQgiNyXSWEEIIjcnCuhBCCI3JdJYQQgiNyXSWEEIIjb3N6QA+AkkiQgihJUoZiQghhNCULKwLIYTQmCQRIYQQGvsUdmfp5nQAQgjxX5Wsk7VHWk6ePEmvXr2wtrbGxMSELVu2qNUrlUpmzZpFlSpVKF68OO3bt+fKlStqbWJjYxk0aBBlypShTJkyDBo0iNjYWLU2ly9fpl27dhQvXhxra2u8vb1RKjNPg5JEhBBCS5Kz+EhLfHw8VatWxcvLCwMDg3fqFy5cyNKlS/H29ubIkSMoFAq6du3KixcvVG0GDhxISEgIvr6++Pr6EhISwuDBg1X1z58/p2vXrpibm3PkyBG8vLxYvHgxS5YsyfQ1ynSWEEJoSXasibRu3ZrWrVsD4OrqqlanVCrx8fFh1KhRdO7cGQAfHx8sLS3x9fXF2dmZsLAwDh8+TEBAAPb29gDMnz8fBwcHrl+/jqWlJTt37iQhIQEfHx8MDAyoWrUq165dY9myZQwbNgydDL5cS0YiQgihJcosPrLq7t27hIeH07x5c1WZgYEBDRs2JCgoCIDg4GCMjIyoV6+eqk39+vUxNDRUa9OgQQO1kU6LFi14/Pgxd+/ezTAGSSJCCKEl2bEmkpHw8HAAFAqFWrlCoSAiIgKAiIgIzMzM1EYTOjo6FCtWTK1NWsf4uy4jMp0lhBBaIlt8s0k1w1If4zTiPayOD83pEMS/JDw6ntMhiH94mY3HStLyJl8LCwsAIiMjKV26tKo8MjISc3NzAMzNzYmOjkapVKpGI0qlkqioKLU2kZGRasf++/nfbdIj01lCCKEl2bE7KyNly5bFwsKCwMBAVVliYiKnT59WrYHY29sTFxdHcHCwqk1wcDDx8fFqbU6fPk1iYqKqTWBgICVKlKBs2bIZxiBJRAghtCQ7Ftbj4uIICQkhJCSE5ORkHjx4QEhICPfv30dHRwcXFxcWLlyIv78/oaGhuLq6YmhoiKOjIwBWVla0bNkSNzc3goODCQ4Oxs3NjTZt2mBpaQmAo6MjBgYGuLq6Ehoair+/PwsWLMDV1TXDnVkgayJCCKE12bEmcv78eTp27Kh6PmvWLGbNmkXv3r3x8fFh5MiRJCQkMHbsWGJjY7Gzs8PPzw9jY2NVn9WrV+Ph4UH37t0BcHBwYPbs2ar6IkWKsHv3btzd3WnWrBkmJiYMHTqUYcOGZRqfTmxsrNavzO9t01vbpxDvKezlk5wOQfzL1au+OR2C+IeX+mbZdqwfarpm3ugfxvy5LNvO/bHISEQIIbQk+RO4e5YkESGE0JL/fgqRJCKEEFoj14kIIYTQmExnCSGE0Nh/P4VIEhFCCK15+wmkEUkiQgihJf/9FCJJRAghtEYW1oUQQmhM+QmMRSSJCCGElshIRAghhMZki68QQgiN/fdTiCQRIYTQGhmJCCGE0JisiQghhNCY7M4SQgihMW1/x3puIElECCG0RKazhBBCaCxZKSMRIYQQGvrvpxBJIkIIoTWyxVcIIYTGZHeWEEIIjcnCuhBCCI3JdJYQQgiNyXSWEEIIjcl0lhBCCI0p5ToRIYQQmnor01lCCCE0JWsiQgghNCa7s3I5U3NTnMc7U7dZXQwMDXhy/wlLJizhUtClTPuWLFeSRQcWoaOjQ3fr7lqN06iIEUOmDqFey3oABB0OwmeyD/HP4wGwqW9Dl4FdsLK1opBxIR7fecyeNXs49OMhrcaVE4aMdGbspOFsXL2DqeO9tXYe+4a1mTh9DJZWFQh/EsnKJRvYtn6XWhxtOjSnfKWyvH71hgt/XGTu9MVcu3pTazF9TKs27uDwbye5c+8B+fPrU6NaFUYNccayQrl0+zx8HE4bx6/fKV/+w3Qa16+jtVifPX/BrAXLOXriDABNG9dngpsLhY2NALh5+y4zfljGzTv3iIuPx7yYGW1bfMHQAX3Q19fXWlzZQdZEcjHDwobM9ZtL6O+hTPl6Cs+ePqN4meI8i36Wad98+vkYt3Qcl4IuYVPf5oNj8drhxeGdhznsezjNeo9FHig+UzCp3yQARs4eifsCd6Z+MxUAaztr7l69y67lu3ga8ZTaTWozwmsEb1694ejeox8cX25ha2dDr37duHLp2gcd57PSJTh2/icqFqudZn2pMiVZs20xvlv3MnrIt9Spb8vU2eN5GhXDwf1HAKjXqA6b1+7k4vnLoKOD2/ghbNzlQ5tGjjyLff5B8eUGv58PoVe3DlS3roxSqWTJ6k0MHDkB/y0rKFLYOMO+K+bNwKpSedXzzNpn5uthHnRxaEWX9q3SrB83dTaPn0SwfN4MAKbMWoDn9DksnZ3y96Gvr09nh5ZUqVyRwsaGhF2/zRTvhSQlJTFm6IAPik3bZHdWLuY4xJGYiBh+cPtBVRZ+P/y9+jp7OnPnyh0unrmYZhKxb2lPH7c+lLUsy9OIpxzde5StC7by9s3bLMdZulJp6jSrg3s3d66euwrAYs/FzN01l88qfMbDWw/5cemPan0ObD5AzYY1aeTQ6D+TRIyMjZi3fAbjR05l+NhBanX6+vlw83Slk6MDJiaFuR52i3kzl3E88LRG5/rya0ciwiOZ6jkbgJvXb1PTrjoDh/ZTJRHnnkPV+oxxncSFW8ewq2fLkYPHNDpvbrJy/vdqz70mjaV+G0fOh1ymaeP6GfYtUtiYYmZF060/euIMy9Zu4cbtuyjMitKuVVNcv9FsVHDzzj1OnDnLRp+52Fa3BmCKx3D6uY7l9t0HlC9bijKlSlKmVElVn5LFLfj9fAjn/sx8xiGnfQprIro5HYCmGrRpQNj5MMYvHc/Wc1tZ/PNiOvTvkGm/us3rYt/CHp/JPmnW125SG4+FHuxfv58hrYawYOwCGrdrTH+P/hrFWaV2FV7GvST0bKiqLPT3UBLiE6hqVzXdfgZGBsQ9i9PonLnRzPnfErDvV86cOPtOnffi77BvaIfb4Ik4fN4Tv+37WbllAVWqWWp0rlp1a3A88Ixa2fEjp7GxtSZfvrQ/NxkaGaKnp/efGIWkJf5lAsnJyaopooyMmjiDJu178dWQMfwSeFyt7mTQH4ybOpve3TuyZ9Nypnu6cSjwBAtWrNcorj8vXaGQgQG1bFL/FmrVqIaBQUEuXApNs8+9B484EXSWOrU+fBZB25JRZumRF71XEomNjeXgwYMEBQW9M8cXHx+Pt7f25rbTU7x0cdr3bc+Te0/4tu+3+K/1x3m8c4aJpKhFUUZ4j2DuyLkkvkxMs02v4b3YtWIXh3Ye4sndJ4ScDmHdrHW0+6qdRnGaKkx59vTdKbbY6FhMzU3T7GPfwh7bRrb8vPVnjc6Z2zj17UrZ8qWZN3PZO3VlypWiY7e2DB8wjt9Pn+P+3YdsWrOD3w6foHd/zdaqFOZmREVGq5VFRT5FX18fUzOTNPtM+t6dyyFXOf97iEbnzO28Fi6nimUFav71aT8thQwK4j5sID9M82TZ3GnUt6uJ+2Qv9h08omqzcsN2nL90pGv71pQpVRJ7u5q4uX7Dj3sOaDT/HxUdQ1HTIujo6KjKdHR0MDM1ISo6Rq1tn8Gjqd2sE+2cBlC7RjVGDv46y+f72JRKZZYeeVGm01lXrlyhS5cuREVFkZycTM2aNdm4cSNlypQBUpPIuHHjtB7sP+no6nA95DrrvdcDcOvyLUqWL0mHfh3Yv2F/mn3cF7hzYNMBwi6EpXvcSjaVqFyzMo4ujmrnKmhQEFNzU2IiYhg2cxjNujZT1ecvmJ8qtargMt1FVTakxRAiH0Vm+XVVrVMVj0UeLJ+ynGt/ftjaQW5QvlJZxkwchlOHb3j79t3pwGo1qqCrq8vBk75q5fkL6HP6eOqoJeTOCdW//37D+WfZ2TPn+abXcI1inDB9NHXq18Kp/TckJ//3ZrFnL1rJuT8vs9FnLnp6eum2MzUpwte9UxN3devKxDx7ztotO+nYpjkAoWHXuXgljDVbdqraKZOVJL56RVR0DIpiRZk6ezH7f0lNPK9evSbk8lW+n5/6IcJ/8wpKFDfP0uuYO82Tly9fEnbjNj8sXc2azTv5Xz+nLB3jY8uro4usyDSJTJ06lbp167JixQpevHjB+PHjadu2Lfv27aNixYofI8Y0xUTEcP/6fbWy+zfu0/mbzun2sW1ki009G74c9WVKgQ7o6emx79Y+ln67lICtAejo6rB14VaO7z/+Tv+/F+03/bCJXStSd/p4LPLg5M8nOfnzSVVZdHjKJ+GYyBiKFC3yzrFMzEyIiVD/pFW1blWmrZ/GpnmbOLD5QCY/gbyhdp0amBUzJeBE6ptOvnz5sG9Qmy+/7s7oId+SnJxM11Z930kyiQmpo8WOzXqr/m1RQsE2/9VqZf9sGxkRTTGFmdqxiimK8ubNG2KiY9XKJ84YQ4eurenTeTD37z78sBebC3kvXMHPv/7G2sXelP6sRJb716hqxZ6fUncJJicrcXHuQ5vmn7/T1tQk5f/5sP/1xfnL1GQ0bupsWjVtRMsvGqnKFMVSfj/FzEx5GvMMpVKp+nCgVCqJjomlmJn6SL2EhQKAiuXLkpSczBSvBTh/6Ui+fOknxpz2KayJZJpEzp49y759+zA0NMTQ0JD169czYcIEOnTowL59+yhcuPDHiPMdoWdD+aziZ2pln5X/jIiHEen2cWnpova8fuv6OA13wq2jm+pN/+alm5SqWIrHdx+ne5xn0c/UdoG9SnxFbFRsmn2unrtKIaNCWNtZc+WPK0DKbiwDQwNC/0id861uX53v1n/Hlnlb2LtmbwavPG/55UAgFxurz217L/6OO7fu4TN/LW/evkVXVxeFhVma6yV/u3s79QPD38nmn2X/dP73EFq3b6ZW1rhpfS5euKKWqCZ97077Lq3p02Uwt27cyepLy/VmLVhOwK/HWLvYiwplS2t0jKvXb6EolrrIbm1Vidt376stdP+bmakJZqap04YFCuSnqIlJmn1qVrfmZUICFy5dUa2LXLh0hYSERGyrp79mmJycTFJSEsnJSUDuTSJJeXSKKisyTSKvX79Wm68EmDlzJkqlkg4dOrBq1SqtBZeR3at388PuH3Aa5sSxfceoWL0inZw7sWH2BlWbr8d9TWXbykzoPQGAu9fuqh3DsoYlymSlWvnWBVv5bt13RDyI4Pj+4yQlJVHWqixWtlasnbk2y3Hev3Gfs4FnGT5rOIvGLwJg+KzhBB0O4uGtlE++NvVtmLp+Kvs37ufo3qOYKlI+gSUlJfH8ad5e6H3xPI4Xz9U3CLx8mUBszHPVNRl7dh5g9uKpzJw8j8shVzExLUy9RnW4d+chv/x0JK3DZmjrel/6DnDi2xnubNuwC7t6NenWqyOjBnmq2nznPZ4uPdvh0m8Mz549p5h5yifjl/EveRmf8AGvOHeY8cNS9gX8yiKvyRQxNiIq+ikAhQwMKFTIAID5Puu4dCWMNYu8ANh74BD58uWjSuWK6OrocvTkGbb57We0yzeq47o4f8nQsVMoWdycNi2aoKenx41bd7gYek2j7bYVy5Whcf06TJu9mCnjRgAwbfZivmhkT/mypQDwD/iVAvnzY1mxHPr58nH56nUWLl9Pq6aNyZ8//wf9nLRNprOASpUqcf78eapUqaJWPmvWLJKTk+nTp4/WgsvI9ZDrTP/fdPp79Kf3iN5EPopk0w+b2L8xdT3E1NyUEmWyNoQ/d+wcU76eQu+Rvek2uBvJb5N5ePshh3ZqfuHf7BGzGTJ1CDM2peyDP3PojNrusJY9WlKwUEEchzjiOCR1LSb8fjjOjZw1Pm9eMW74d7iOHsC4KSMpXtKCZzHP+PP85QxHJhl5cO8RA3oPZ+KMMXzp7EjEk0imTZit2t4L0HdATwA2716h1nfh7BUsmq1elhdt90v5OxgwwlOt3OWbPgwd8BUAUdFPuf9QffS8YsM2Hj+JQFdXl7KlSzF9gptqPQSgUT07ls6Zyor121i/zQ89vZR2Xdq11DhW7ykezJzvw2C3iUDKxYYTR7uq6vPp6bF60w7u3n+EEiUlLczp1a0D/Xp11ficH8unkER0YmNjM3yV8+bN49SpU/j6+qZZ7+7uzpo1a4iJiUmzHqC3Te9068THFfbySU6HIP7l6tW0/7ZEznipb5Z5o/fUxjr9Ndq0HLyiPpU9a9asd3a/mpubc+1ayqYbpVKJl5cXGzZsIDY2Fjs7O+bOnYu1deouvNjYWDw8PAgICACgbdu2zJ49GxOTtHcqZlWmW3xHjx6dbgIBmDt3boYJRAghPlXZcZ2IpaUlYWFhqsepU6dUdQsXLmTp0qV4e3tz5MgRFAoFXbt25cWLF6o2AwcOJCQkBF9fX3x9fQkJCWHw4MHZ9hrz7BXrQgiR22XH7qx8+fJhYWHx7rGVSnx8fBg1ahSdO6eMeHx8fLC0tMTX1xdnZ2fCwsI4fPgwAQEB2NvbAzB//nwcHBy4fv06lpaaXdD7T3n2inUhhMjtsuNiwzt37lClShVq1KjBN998w507dwC4e/cu4eHhNG+eumZlYGBAw4YNCQoKAiA4OBgjIyPq1aunalO/fn0MDQ1VbT6UjESEEEJLPnRhvU6dOixbtgxLS0uioqKYM2cOrVu35syZM4SHp9wrUKFQqPVRKBQ8fpyyYSIiIgIzM7N37ghQrFgxIiLSvxwiKySJCCGElnzorUxatVK/83GdOnWwtbVl69at1K1b94OOnV1kOksIIbQku2/AaGRkRJUqVbh165ZqnSQyUv32SpGRkZibp9xSxtzcnOjoaLVkplQqiYqKUrX5UJJEhBBCS5KUyVl6ZCYxMZHr169jYWFB2bJlsbCwIDAwUK3+9OnTqjUQe3t74uLiCA4OVrUJDg4mPj5ebZ3kQ8h0lhBCaMmH7s769ttvadu2LaVKlVKtibx8+ZLevXujo6ODi4sL8+bNw9LSkkqVKjF37lwMDQ1xdEy5aNnKyoqWLVvi5ubGggULAHBzc6NNmzbZsjMLJIkIIYTWJH/gmsijR48YOHAg0dHRFCtWjDp16nDo0CHVXdRHjhxJQkICY8eOVV1s6Ofnh7Fx6rdRrl69Gg8PD7p3T7kppoODA7Nnz/6guP4p0yvWs4NcsZ57yBXruY9csZ67ZOcV6/Uss3Y7mKDraX/Fdm4mIxEhhNCSDx2J5AWSRIQQQkvk+0SEEEJoTEYiQgghNCYjESGEEBpTvse1H3mdJBEhhNCST+FLqSSJCCGElrzPVeh5nSQRIYTQkg+9AWNeIElECCG0RHZnCSGE0JjszhJCCKExmc4SQgihMdmdJYQQQmMyEhFCCKExWVgXQgihMRmJCCGE0JisiQghhNCYjESEEEJoTG57IoQQQmOysC6EEEJjMp0lhBBCY3LbEyGEEBqTkYgQQgiNfQpJRCc2Nva//yqFEEJohW5OByCEECLvkiQihBBCY5JEhBBCaEySiBBCCI1JEhFCCKExSSLvYfXq1dSoUQMLCwu++OILTp06ldMhfbJOnjxJr169sLa2gGaE6QAAAldJREFUxsTEhC1btuR0SJ+0efPm0axZM0qXLk3FihVxcnIiNDQ0p8MSH5EkkUz4+fkxfvx4xowZw7Fjx7C3t6dHjx7cv38/p0P7JMXHx1O1alW8vLwwMDDI6XA+eSdOnGDAgAEcPHgQf39/8uXLR5cuXYiJicnp0MRHIteJZKJFixZUq1aNRYsWqcpq165N586dmTJlSg5GJj777DNmz55Nnz59cjoU8Ze4uDjKlCnDli1bcHBwyOlwxEcgI5EMvH79mgsXLtC8eXO18ubNmxMUFJRDUQmRe8XFxZGcnIyJiUlOhyI+EkkiGYiOjiYpKQmFQqFWrlAoiIiIyKGohMi9xo8fj42NDfb29jkdivhI5N5ZQohsMWHCBM6cOUNAQAB6eno5HY74SCSJZMDMzAw9PT0iIyPVyiMjIzE3N8+hqITIfTw9PfHz82Pfvn2UK1cup8MRH5FMZ2Ugf/782NraEhgYqFYeGBhIvXr1cigqIXKXcePGsWvXLvz9/alcuXJOhyM+MhmJZGLo0KEMHjwYOzs76tWrx9q1a3ny5AnOzs45HdonKS4ujlu3bgGQnJzMgwcPCAkJwdTUlNKlS+dwdJ8ed3d3duzYwebNmzExMSE8PBwAQ0NDjIyMcjg68THIFt/3sHr1ahYuXEh4eDjW1tbMnDmTRo0a5XRYn6Tjx4/TsWPHd8p79+6Nj49PDkT0aft/u3ZoBEAMA0EsXXvc/VfwZEFCpArMFpz/vrBm5uzu5Wt4QUQAyGwiAGQiAkAmIgBkIgJAJiIAZCICQCYiAGQiAkD2AWgQfsDMh/RAAAAAAElFTkSuQmCC\n","text/plain":["<Figure size 432x288 with 2 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"markdown","metadata":{"id":"T1Mzo4OnJb8E"},"source":["## Project2: without merging extremely positive & extremely positive classes "]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":204},"id":"vm_zQI-fHe2D","executionInfo":{"status":"ok","timestamp":1615027811795,"user_tz":-300,"elapsed":977,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"2f03fdd7-7e16-4ad0-fd96-983773e0b14b"},"source":["df_movies_train.head()"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>OriginalTweet</th>\n","      <th>Sentiment</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>@MeNyrbie @Phil_Gahan @Chrisitv https://t.co/i...</td>\n","      <td>Neutral</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>advice Talk to your neighbours family to excha...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Coronavirus Australia: Woolworths to give elde...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>My food stock is not the only one which is emp...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Me, ready to go at supermarket during the #COV...</td>\n","      <td>Extremely Negative</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["                                       OriginalTweet           Sentiment\n","0  @MeNyrbie @Phil_Gahan @Chrisitv https://t.co/i...             Neutral\n","1  advice Talk to your neighbours family to excha...            Positive\n","2  Coronavirus Australia: Woolworths to give elde...            Positive\n","3  My food stock is not the only one which is emp...            Positive\n","4  Me, ready to go at supermarket during the #COV...  Extremely Negative"]},"metadata":{"tags":[]},"execution_count":109}]},{"cell_type":"code","metadata":{"id":"AI7_znMIJvpD"},"source":["#check duplicate values\n","df_movies_train=remove_duplicate(df_movies_train)\n","df_movies_train.head()"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"-VwK7fCHJvWb"},"source":["# data pre-processing for tranning \n","text_data=df_movies_train['OriginalTweet'].tolist()\n","process_text=data_pre_processing(text_data)\n","#process_text= list(filter(None, process_text))\n","df_train_process_text=pd.DataFrame(process_text)"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":204},"id":"b0d32Q-zL--C","executionInfo":{"status":"ok","timestamp":1615028424060,"user_tz":-300,"elapsed":891,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"a6416a49-1d34-46df-d04c-84152576da45"},"source":["df_train_process_text.columns=['OriginalTweet']\n","df_train_process_text.head()"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>OriginalTweet</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td></td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>advice talk neighbour family exchange phone nu...</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>coronavirus australia woolworth give elderly d...</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>food stock one empty please dont panic enough ...</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>ready go supermarket covid outbreak im paranoi...</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["                                       OriginalTweet\n","0                                                   \n","1  advice talk neighbour family exchange phone nu...\n","2  coronavirus australia woolworth give elderly d...\n","3  food stock one empty please dont panic enough ...\n","4  ready go supermarket covid outbreak im paranoi..."]},"metadata":{"tags":[]},"execution_count":113}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"rmauMhgyL_RH","executionInfo":{"status":"ok","timestamp":1615028480426,"user_tz":-300,"elapsed":799,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"e9bbe110-54b4-4f13-e76c-9b9d0831aa24"},"source":["#check dimensions after pre-processing of traning data \n","print(df_movies_train.shape)\n","print(df_train_process_text.shape)"],"execution_count":null,"outputs":[{"output_type":"stream","text":["(41157, 2)\n","(41157, 1)\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"m_FO4sraJvBy"},"source":["# Data concatnate \n","Data_tarning=pd.concat([df_train_process_text,df_movies_train['Sentiment']],axis=1)"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":204},"id":"fjzWt2HfMvzP","executionInfo":{"status":"ok","timestamp":1615028542688,"user_tz":-300,"elapsed":991,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"e6ba836e-ce8a-492d-eea0-cf25691b49c2"},"source":["#drop nun values \n","nan_value = float(\"NaN\")\n","Data_tarning.replace(\"\", nan_value, inplace=True)\n","Data_tarning.dropna(inplace=True)\n","Data_tarning.head()"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>OriginalTweet</th>\n","      <th>Sentiment</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>1</th>\n","      <td>advice talk neighbour family exchange phone nu...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>coronavirus australia woolworth give elderly d...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>food stock one empty please dont panic enough ...</td>\n","      <td>Positive</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>ready go supermarket covid outbreak im paranoi...</td>\n","      <td>Extremely Negative</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>news region first confirmed covid case came su...</td>\n","      <td>Positive</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["                                       OriginalTweet           Sentiment\n","1  advice talk neighbour family exchange phone nu...            Positive\n","2  coronavirus australia woolworth give elderly d...            Positive\n","3  food stock one empty please dont panic enough ...            Positive\n","4  ready go supermarket covid outbreak im paranoi...  Extremely Negative\n","5  news region first confirmed covid case came su...            Positive"]},"metadata":{"tags":[]},"execution_count":117}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":204},"id":"IsJ0CCb5MwFb","executionInfo":{"status":"ok","timestamp":1615028617205,"user_tz":-300,"elapsed":914,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"179c3b14-b8bf-41a4-913f-34cb19ac9ced"},"source":["train_labe_nn=pd.get_dummies(Data_tarning['Sentiment'])\n","train_labe_nn.head()"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Extremely Negative</th>\n","      <th>Extremely Positive</th>\n","      <th>Negative</th>\n","      <th>Neutral</th>\n","      <th>Positive</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>1</th>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>1</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>1</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["   Extremely Negative  Extremely Positive  Negative  Neutral  Positive\n","1                   0                   0         0        0         1\n","2                   0                   0         0        0         1\n","3                   0                   0         0        0         1\n","4                   1                   0         0        0         0\n","5                   0                   0         0        0         1"]},"metadata":{"tags":[]},"execution_count":118}]},{"cell_type":"code","metadata":{"id":"hufkEmzfNPrO"},"source":["# Count Vectorizer and TFIDF\n","from sklearn.feature_extraction.text import CountVectorizer\n","vectorizer = CountVectorizer(max_features=1000, min_df = 3, max_df = 0.6)\n","X_data_train = vectorizer.fit_transform(Data_tarning['OriginalTweet'])\n","y_label_train=train_labe_nn\n","from sklearn.feature_extraction.text import TfidfTransformer\n","transformer = TfidfTransformer()\n","X_train_tfid = transformer.fit_transform(X_data_train).toarray()"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"KRQl98ioNS71"},"source":["from sklearn.model_selection import train_test_split\n","X_train, X_test, y_train, y_test = train_test_split(X_train_tfid, y_label_train, test_size=0.2, random_state=42)"],"execution_count":null,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"L9DvJlSLaRiW"},"source":["## Class weight for data inbalance"]},{"cell_type":"code","metadata":{"id":"7b9dSwxrSk7c"},"source":["class_weights = {0:3,  \n","                  1:2.5,\n","                 2:2,\n","                 3:1.5,\n","                 4:1.0}"],"execution_count":null,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"tZQLAshrab9B"},"source":["## Model"]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"mMp9HotPNTLX","executionInfo":{"status":"ok","timestamp":1615031369992,"user_tz":-300,"elapsed":139437,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"a98787ea-0b59-4fbb-9892-59e4ce78fa5a"},"source":["import keras\n","from keras.models import Sequential\n","from keras.layers import Dense\n","from keras.layers import Dropout\n","#adding the input and first hidden layer\n","classifier = Sequential()\n","classifier.add(Dense(75, kernel_initializer='uniform', activation='relu',input_dim=1000))\n","classifier.add(Dropout(0.3))\n","\n","#adding the second hidden layer\n","classifier.add(Dense(50, kernel_initializer='uniform', activation='relu'))\n","classifier.add(Dropout(0.3))\n","classifier.add(Dense(10, kernel_initializer='uniform', activation='relu'))\n","classifier.add(Dropout(0.3))\n","\n","#adding the output layer\n","classifier.add(Dense(5, kernel_initializer='uniform', activation='softmax'))\n","\n","classifier.compile(optimizer=\"Adam\", loss='kullback_leibler_divergence', metrics=['accuracy'])\n","\n","\n","classifier.fit(X_train,y_train, batch_size=30, epochs=50,class_weight=class_weights)\n"],"execution_count":null,"outputs":[{"output_type":"stream","text":["Epoch 1/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 2.7609 - accuracy: 0.2593\n","Epoch 2/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 2.1709 - accuracy: 0.4003\n","Epoch 3/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 2.0801 - accuracy: 0.4786\n","Epoch 4/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 2.0202 - accuracy: 0.4979\n","Epoch 5/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.9474 - accuracy: 0.5213\n","Epoch 6/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.9059 - accuracy: 0.5308\n","Epoch 7/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.8282 - accuracy: 0.5495\n","Epoch 8/50\n","1095/1095 [==============================] - 3s 3ms/step - loss: 1.7695 - accuracy: 0.5692\n","Epoch 9/50\n","1095/1095 [==============================] - 3s 3ms/step - loss: 1.7111 - accuracy: 0.5851\n","Epoch 10/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.6649 - accuracy: 0.5919\n","Epoch 11/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.6223 - accuracy: 0.6044\n","Epoch 12/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.5733 - accuracy: 0.6194\n","Epoch 13/50\n","1095/1095 [==============================] - 3s 3ms/step - loss: 1.5397 - accuracy: 0.6293\n","Epoch 14/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.4877 - accuracy: 0.6471\n","Epoch 15/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.4843 - accuracy: 0.6497\n","Epoch 16/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.4208 - accuracy: 0.6639\n","Epoch 17/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.4112 - accuracy: 0.6687\n","Epoch 18/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.3708 - accuracy: 0.6779\n","Epoch 19/50\n","1095/1095 [==============================] - 3s 3ms/step - loss: 1.3385 - accuracy: 0.6868\n","Epoch 20/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.3228 - accuracy: 0.6909\n","Epoch 21/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.2919 - accuracy: 0.7005\n","Epoch 22/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.2724 - accuracy: 0.7102\n","Epoch 23/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.2620 - accuracy: 0.7163\n","Epoch 24/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.2210 - accuracy: 0.7220\n","Epoch 25/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.2238 - accuracy: 0.7202\n","Epoch 26/50\n","1095/1095 [==============================] - 3s 3ms/step - loss: 1.2089 - accuracy: 0.7212\n","Epoch 27/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.1840 - accuracy: 0.7296\n","Epoch 28/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.1897 - accuracy: 0.7275\n","Epoch 29/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.1603 - accuracy: 0.7318\n","Epoch 30/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.1386 - accuracy: 0.7374\n","Epoch 31/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.1286 - accuracy: 0.7424\n","Epoch 32/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.1110 - accuracy: 0.7447\n","Epoch 33/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.1033 - accuracy: 0.7453\n","Epoch 34/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.0971 - accuracy: 0.7489\n","Epoch 35/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.1053 - accuracy: 0.7513\n","Epoch 36/50\n","1095/1095 [==============================] - 3s 3ms/step - loss: 1.0601 - accuracy: 0.7604\n","Epoch 37/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.0695 - accuracy: 0.7578\n","Epoch 38/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.0397 - accuracy: 0.7658\n","Epoch 39/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.0583 - accuracy: 0.7639\n","Epoch 40/50\n","1095/1095 [==============================] - 3s 3ms/step - loss: 1.0268 - accuracy: 0.7677\n","Epoch 41/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.0326 - accuracy: 0.7678\n","Epoch 42/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 1.0153 - accuracy: 0.7709\n","Epoch 43/50\n","1095/1095 [==============================] - 3s 3ms/step - loss: 1.0265 - accuracy: 0.7745\n","Epoch 44/50\n","1095/1095 [==============================] - 3s 3ms/step - loss: 1.0198 - accuracy: 0.7752\n","Epoch 45/50\n","1095/1095 [==============================] - 3s 3ms/step - loss: 0.9799 - accuracy: 0.7768\n","Epoch 46/50\n","1095/1095 [==============================] - 3s 3ms/step - loss: 0.9906 - accuracy: 0.7809\n","Epoch 47/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 0.9854 - accuracy: 0.7795\n","Epoch 48/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 0.9598 - accuracy: 0.7865\n","Epoch 49/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 0.9615 - accuracy: 0.7858\n","Epoch 50/50\n","1095/1095 [==============================] - 3s 2ms/step - loss: 0.9646 - accuracy: 0.7844\n"],"name":"stdout"},{"output_type":"execute_result","data":{"text/plain":["<tensorflow.python.keras.callbacks.History at 0x7f58e40bb750>"]},"metadata":{"tags":[]},"execution_count":152}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"U0T0acxZNTb2","executionInfo":{"status":"ok","timestamp":1615031392795,"user_tz":-300,"elapsed":821,"user":{"displayName":"AI Team","photoUrl":"","userId":"17689037188246991143"}},"outputId":"2f783b56-305c-46d4-9160-071efce30949"},"source":["y_pred = classifier.predict(X_test)\n","y_pred = (y_pred > 0.5)\n","# Making the Confusion Matrix\n","from sklearn.metrics import classification_report\n","print('\\nClassification Report\\n')\n","print(classification_report(y_test.values.argmax(axis=1), y_pred.argmax(axis=1), target_names=['class Extremely Negative',\t'class Extremely Positive','Class Negative', 'Class Neutral','Class postive']))"],"execution_count":null,"outputs":[{"output_type":"stream","text":["\n","Classification Report\n","\n","                          precision    recall  f1-score   support\n","\n","class Extremely Negative       0.39      0.49      0.43      1114\n","class Extremely Positive       0.62      0.61      0.62      1283\n","          Class Negative       0.43      0.55      0.48      1970\n","           Class Neutral       0.59      0.53      0.56      1549\n","           Class postive       0.55      0.41      0.47      2294\n","\n","                accuracy                           0.51      8210\n","               macro avg       0.52      0.52      0.51      8210\n","            weighted avg       0.52      0.51      0.51      8210\n","\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"fnEqzGEwNTt2"},"source":[""],"execution_count":null,"outputs":[]}]}