Description:
Reading academic papers is a common task for researchers, but the process can often be time-consuming and challenging. Sifting through large amounts of text to extract important information and understand the main points of a paper can be tedious and requires a significant amount of mental effort. In addition, understanding the context and connections between different pieces of information within a paper can be difficult, particularly for papers in fields outside of an individual's expertise.
To address these issues, the goal of this project was to create a web application that would act as a clever PDF reader, aimed at enhancing the reading experience for academics. The project had two main objectives: first, to develop a tool that could parse PDF files and extract crucial data such as figures, tables, and references; and second, to use this extracted material to aid in paper interpretation and generate fresh research concepts. By automating the process of extracting and organising key information from papers, the hope was that researchers could save time and effort and better understand and interpret the content of the papers they were reading.
Features:
- Upload and preview PDF
- Show the reference on reference hover
- Create knowledge graph from references
- Generate summary from the paper
Screenshots:
Upload page
PDF Preview
Reference hovering
Knowledge graph
Summary
Architecture:
Technologies:
- Development
- Frontend: React.js, PDF.js
- Middleware: Sprint Boot 2+, H2 database
- Reference service: Django, Python
- Summary service: Django, Python, OpenAI
- Deployment
- Frontend: Firebase
- Middleware: Render
- Reference service: Render
- Summary service: PythonAnywhere
Project links:
GitHub repositories:
Project specification (requirements definition)
Contact:
Jaime Baquerizo: jimmybaq@gmail.com
Ivan Juričić: ivanjuricic97@gmail.com
Mihael Kampić: mihael.kampic@gmail.com
Leon Kranjčević: leon.kranjcevic@outlook.com
Patrick Niantcho: patrick.niantcho@gmail.com
Santiago Pérez: santiperez3@gmail.com