This project aims to find the confidence level from 1 to 5 for semanitic similarity between two text document.

Group Name: OUTLIER Group Memebers:

1. Arihant Chhajed
2. Diksha Chhabra

Download all the package from the requirement.txt Go to STS Directory

Step #1 Preprocessing

Execute preprocess.py to generate preprocessed data

eg python preprocess.py

Step #2 Feature Generation

Execute generate_feature.py to generate features from data eg python generate_feature.py

Step #3 Model execution

Exceute main.py for model exection argument can be as follows:- 0: Train 1: Train&Test(dev set) 2: Test (test set) eg python main.py

Predictions can be seen in "Predictions Folder" for all the model and average of those model.

For Feature Testing

execute test.py eg python test.py

Contact: [email protected] for more info