This website was designed to help researchers and students help simplify understanding and working with scientific articles and research papers.This tool currently only wokrs for Springr non-mathematical papers only.New versions will aim to add more papers and capabilities, all contributions are openly welcomed, in building a student and research friendly application.
- Automated Text Mining: Utilizes TF-IDF algorithm to extract relevant text segments from scientific papers.
- Content Generation: Employs BART (Bidirectional and Auto-Regressive Transformers) model for generating comprehensive summaries from mined text segments.
- PowerPoint Presentation: Integrates with the Python pptx library to create visually appealing PowerPoint slides containing the summarized content.
- Text-to-Speech: Incorporates the Realistic Text to Speech API by VidLab for converting text summaries into natural-sounding speech, enhancing accessibility and user experience.
- Clone the repository and install the required dependencies.
pip install -r requirements.txt
- To start python backend api
python app.py
- To run React website
cd Frontend
npm start
- Download pre-trained BART models and configure the transformers library accordingly.
- Run the summarization script, by providing the link through the React website.
- View the generated Summary , PowerPoint presentation (summary.pptx) containing the summarized content as well as audio file (audio.mp3).
The Outputs folder will contain all the non-text outputs
- Jafar N
- Sharon T Saju
- Sreedev TS , xreedev@gmail.com
app.py : Flask api to treat requests ppt.py : create ppt based on template.pptx BART.py : Module responsible for summarization using BART data.py :Used to store any data repetitively used dataProcesser.py : Used to remove stop words main.py:To test all backend features pdfextract.py: Used to extract text from pdfs(incomplete) TFIDF.py:Extractive summarization using TF-IDF algorithm tts.py : Test to speech module webcrawler.py:Webcrawl and parse data from springr articles
This project is licensed under the MIT License. See the LICENSE file for details.