Project Uniqueness Checker

Under- Graduate final year students can check the uniqueness of their proposed project in comparison to previous years

Final year project done for award of degree , VTU CSE, 2020

Overview

• Aim: Check uniqueness of a under-graduate students proposed project by taking in students title and abstract and returning a uniqueness score in comparison to previous year projects
• Method: A combination of Siamese Manhattan LSTM + Inception Model is used to compare titles.
• Implementation:

  • Title similarity: checked using Siamese Manhattan LSTM + Inception Model

  • Abstract similarity: checked using spaCy module

  • Uniqueness: calculated as

    *Uniqueness=100-max(similarity scores) X 100*
    

  • Backend: SQLite

  • Frontend: Flask , Flask-admin, Flask-Login , jQuery

Installing

1. Create a conda virtual environment

conda create -n myenv python=3.7

2. Install requirements

pip install -r requirements.txt

3. Run train.py to set up

train.py

Use

Checking similarity score

Webpage where students can enter proposed project ideas title and abstract to get results

python checkscore.py

paste http://127.0.0.1:5000/ in the browser to access the page

Project manager dashboard

Dashboard for project manager to perform CRUD operations.

python projectDashboard.py

paste http://127.0.0.1:5000/ in the browser to access the page

Admin Dashboard

Dashboard where the admin can perform CRUD operations on users and view projects uploaded

python adminDashboard.py

paste http://127.0.0.1:5000/ in the browser to access the page