Summary-Pi

Summarize Huge Books so you can Read and Understand Faster

Comment

Inspiration

During Lock-Down all the students are learning from Home , they are having tough time in reading textbooks or websites. To make their reading easier and faster we have made a Web-App which uses Machine Learning to convert Books into Short Summaries.

And as it is Hack At Home we are hosting the WebApp and Machine Learning Model directly on our Raspberry-Pi at our Home.

What it does

  • It takes PDF files or Text from user and gives the Summary of that book or text.
  • It does this using Natural Language Processing Models and Machine Learning directly from Raspberry-Pi.
  • It also provides a Website GUI for User to easily do this task.

How we built it

We build our front end using flask to interact with back-end server which is raspberry-pi easily

  1. First we created a simple Flask Website to send POST Request and send PDF files and Text to the Server using Flask and HTML
  2. Then we implemented and tested the Machine Learning Model using Natural Language Processing and Extractive Text Summarization on Raspberry Pi
  3. The Machine Learning Model was developed using PyTorch and Transformers Library in Python
  4. Then we tried to host the Local Hosted Flask Server on Raspberry-Pi using NGROK Package.

rpi-nlp

Challenges we ran into

  • Transferring PDF file from front-end to back-end were difficult as we are doing it for first time
  • Running Machine Learning Models on Raspberry Pi was Tricky

Accomplishments that we're proud of

  • Successfully created back end on raspberry pi
  • Users can easy summarize there text and will help them for fast learning

What we learned

  • Flask web app development
  • Domain name hosting
  • Raspberry Pi

What's next for Summary-Pi

  • Handling large files are time consuming we have to look on it
  • Creating Chrome Extension so user can directly summerize website
Share this project:

Updates