parthplc / teacheasy

A goto software for teachers and students.

TeachEasy

Levaraging the power of NLP, we have developed TeachEasy with the aim of reducing the manual work for School Teachers in the domain of Question and Answer generation, Text summarization, and Paraphrasing.

Requirements.

pip install numpy
pip install textblob
pip install gensim
pip install scikit-learn
pip install -U transformers==3.0.2
pip install streamlit==0.64.0
pip install -U torch==1.6.0
pip install -U spacy
python -m spacy download en_core_web_sm
pip install streamlit

Technical stuff

The t5-small model for summarisation,t5-base for question generation and distill bert for answer extraction would be automatically installed in your local machine when we run the streamlit app.This might be time taking for first time so keep patience.

Question generation, Summarization and Paraphrasing is achieved using T5 model. Answer generation is developed via DistilBert.

How to use repo

streamlit run app.py or streamlit run main.py

Results and App snapshots

Main template interface

Answer extraction from questions

Summarization with spacy using Text Summarization( Extractive Summarization)

Summarization using T5 (Abstractive Summarization)

Paraphrasing of Question

Boolean Question generation