TezHaber — Centroid-based Text Summarization (AçıkHack 2019)
Dec 1, 2019
·
1 min read
TezHaber — NLP Hackathon ProjectProject Overview
TezHaber is a text summarization and aggregation platform, developed as part of the AçıkHack NLP Hackathon (2019) at Bilişim Vadisi. The project focused on adapting centroid-based compositionality of word embeddings to the Turkish language, demonstrating real-world applicability through news scraping and automatic summarization.
Key Features
- Summarization Engine: Centroid-based embeddings to generate concise summaries.
- Turkish Language Adaptation: Custom pre-processing and embedding adjustments for Turkish.
- News Aggregation: Automated scraping and aggregation of trending headlines.
- Hackathon Achievement: Ranked 2nd place in the 2019 AçıkHack NLP Hackathon.
This project highlighted the potential of unsupervised embedding-based methods for multilingual summarization and practical applications in media technology.
