Health news feed: Identifying personally relevant health-related URLs in tweets

Abstract

A common use of micro-blogging systems, such as Twitter, is to `tweet' or `re-tweet' URLs of the latest news articles. The challenge is that with the large number of such micro-blog posts, it is difficult to find and filter to just the most relevant news for an individual. In this paper, we propose and detail the health news feed system which utilises a three-stage filtering and categorisation process with three types of knowledge resources using natural language processing (NLP) technologies for filtering and extracting personally-relevant health-related news articles referred to in tweets. The three stages are term-based filtering, content filtering, and categorization.