Yelp Reviews

Repository for the Text Mining course at VU Amsterdam 2022

Environment Setup

conda create -n text_mining python=3.7.7
activate text_mining
pip install -r requirements.txt

How to run

To run the application prototype you can either run it locally by setting up the environment or by using google collab link. Either way you would need to get the models with this link https://drive.google.com/drive/folders/1GAP9tuE56pdmuJ7ckQBdOtYKpjBF0QKm?usp=sharing

Structure

.
├── application/
|   ├── application_prototype.ipynb
├── code/
|   ├── fake_classifier/  
|   |   ├── FeatureEngineering.ipynb
|   |   ├── LogisticRegression.ipynb
|   ├── sentiment_analysis/        
|   |   ├── SIEBERT_evaluation.ipynb
|   |   ├── FT_BERT.ipynb
|   |   ├── FT_BERT_evaluation.ipynb
|   ├── topic_modelling/
|   |   ├── BERTopic.ipynb
|   |   ├── LDA.ipynb
|   ├── preprocessing.ipynb
├── models/
|   ├── models.zip                              <-- download the file from drive and unzip the file into models repository
├── datasets/
|   ├── YelpZip/
|   |   ├── YelpZip.zip                         <-- unzip the file into the existing folder (supported using git LFS)
|   ├── production_set.csv
|   ├── sample_production_set.csv
|   ├── processed_yelp.csv                      <-- not included because can be obtained from preprocessing notebook (too large to push)
|   ├── sentiment_sample_25_75.csv              <-- not included because can be obtained from preprocessing notebook (too large to push)
|   ├── sentiment_sample_50_50.csv              <-- not included because can be obtained from preprocessing notebook (too large to push)
|   ├── classifier_sample.csv                   <-- not included because can be obtained from feature engineering notebook (too large to push)
├── report/
|   ├── report.pdf
├── README.md
└── requirements.txt