Using co-word network community detection and LDA topic model to extract topics on TED Talks.

Project links

About this project

This project applies two automatic clustering (co-word network analysis with community detection & NLP techniques with LDA topic modeling) on 3 data columns (related tags, video description and transcript) to extract topics from TED Talks, and discerns whether more controlled vocabularies and natural texts that provide different aspects of information of videos reveal similar topics. Also, this projects provides thorough exploratory data analysis of this data collection, such as videos recorded/published over time, popular tags/topics over time, most viewed talks, etc.

Skills