I'm a Freshers and i love in the field of data science, Machine learning and Data Analyst and i had done 6 month of internship in Machine Learning field. I used to have a creative thinking with the real world problem and try to analize it and try to implement it.
This Web app can predict the Email/SMS as Spam or Not Spam.
I had taken dataset called SMSSpamCollection and by the use of Natural Language Processing with Bag of Words i trained the model with Multinomial Naive Byes classifier and after tesing i got with 98% accuracy with also good precision. Then i made a pickle file for bag of words and Multinomial Naive Byes classifier to import into streamlit so that from that pickle file i can run directly of the user entered Email/SMS text data and then creating web app framework by streamlit and finally deployed a web app Machine Learning Model by using heroku cloud platform.
https://sukumars-spam-detector.herokuapp.com/
-
Install jupyter/Spyder Notebook
-
Knowledge of NLP with how Bag of Words or TF-IDF works with implementation
-
Create Pickle files
-
Install Pycharm and Install Streamlit
-
import those pickle files and create all the necessary requirement for web app like button and text box etc
-
Run the user text input and output the preicted result for weather the Message is Spam/ham.
Python, Sklearn, Tensorflow, NLP, Data Visualization Tool(EDA), mySQL, Tableau, Excel, C/C++