Skip to content

Extractive, Abstractive text/document summarizing system with Flask, capable of suggesting topics, detecting language, summary download and text-to-speech.

Notifications You must be signed in to change notification settings

munas-git/text-summarization-webapp

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Project Status: Complete.

Access project Here. Click link or scan QR Code below. (Note: Site may take up to 45 seconds to load due to render free tier RAM allocation, SORRY about that 😥😅).

ABSTRACTIVE SUMMARY FEATURE IS CURRENTLY DOWN.

Project Title: Extractive, Abatractive document/text Summarizing System.

Project Description.

The document summarizing system was done as my final year project in partial fulfilment of the reqirements for the award of Bachelor of Science B.Sc. (HONS) in Software Engineering at Babcock university.

It not just a summarizing system, it is also capable of predicting/suggesting highly accurate subject matters/topics of any body of text entered into it, providing users with the top 2 most likely topics.

The system is also capable of detecting the language of the body of text entered; Model was trained with data from here, and finally, it is equiped with a text-to-speech feature which reads the summary for users and a download feature which allows users to download their summaries as word (.docx) documents.

Tools and Libraries used:

  • CSS
  • Nltk
  • Docx
  • Flask
  • HTML
  • Pandas
  • JavaScript

Snapshots of System.

Landing PageText Summary Page

Document Upload PageSummary Result Page

Demo of System.

Document.summary.system.demo.mp4

Extra Notes

The system UI was a template gotten from free-css however, I made several adjustments in order to suite my taste and my project requirement. I'm currently not very good with HTML and CSS 😅.
There is still room for code refactor and system improvements. New features such as converting summary from one language to another, a broader vocabulary for word synonyms (one word having a list of signle word or phrase synonyms instead of single synonym and synonyms of non-english words), improved UI, better speach synthesizer, multiple download formats, support for more document type summaries, improved system security, improved language model performance (Collecting usser review on predicted review, saving response to db, automate model re-training and re-deploying process.) and more are updates in view however, you are more than welcome to fork this repo and make those changes and more 😉.