A prototype that retrieves Wikipedia articles, clusters them based on content similarity, generates summaries, and visualizes the results on a front-end web interface.
This project allows users to:
Enter a search term.
Fetch relevant article titles from Wikipedia.
Group these articles using a clustering algorithm (K-Means).
Generate a short summary for each article.
Display results and visualize clusters on a webpage.
User can type in a keyword to search Wikipedia.
Articles are retrieved via the Wikipedia API.
Simple K-Means clustering groups the articles by content similarity.
Articles are summarized using a text summarization library (e.g., sumy).
Frontend displays article titles, summaries, and indicates cluster labels.
Basic static graph visualization with D3.js (nodes colored by cluster).
Use BART or other advanced models for more sophisticated summarization.
Support dynamic or interactive graph layouts (e.g., force-directed graphs with dragging).
Add topic labeling to clusters (e.g., detect a phrase that describes each cluster).
Implement advanced search and filtering (e.g., search by date, popularity).
Provide interactive tooltips with larger previews or images.