- Jupyter Notebook File
- Presentation .PPTX with Recording (Please download the power point and play the presentation under the SlideShow tab).
- Presentation MP4
- Read below for documentation (additional documentation for each step in the Jupyter notebook)
This main goal of this project is to be a model for sentiment analysis of businesses reviews.
The code uses a plethora of libraries such as pandas, numpy, seaborn, matplotlib, and nltk.
The notebook itself has comments and explains each section of the data analysis plus the sentiment analysis run at the and.
The data used in this project is the Yelp Dataset since it has various different businesses and a large number of reviews. This model can be used with other datasets, provided it's given a string input as shown in the notebook.
Open the jupyter notebook file
Click Open in Colab
Follow instructions to sign in using a Google account, if necessary
There are two options for this step. Once inside the Colab
notebook, proceed with either of the following options:
-
(Preferred) Run each snippet, one at a time; read the comments and the code, and understand what each section does.
or
-
Click the
Runtime
tab. Then click on theRun all
option. This will run each code snippet in order.
You may find the model evaluation at the second to last snippet, labeled Analysis
under section #8. Sentiment Analysis.
Please see the Jupyter notebook file for an example of a successful run and its equivalent output.
https://medium.com/analytics-vidhya/restaurant-reviews-sentiment-analysis-and-reccomendation-9bdf31a0b20 https://machinelearninggeek.com/analyzing-sentiment-of-restaurant-reviews/ https://ieeexplore.ieee.org/document/8884282 https://inside-machinelearning.com/en/quickly-upload-public-google-drive-files-on-notebook-and-colab/se https://towardsdatascience.com/a-guide-to-text-classification-and-sentiment-analysis-2ab021796317 https://www.kaggle.com/code/ngyptr/python-nltk-sentiment-analysis