Project Newscombinator

News aggregator and news classification in real-time.

Started 2014 and ongoing, one of the biggest topics in the past years for me: working with textual data. News, Blogs, Articles, Comments, name it. My work for my master's thesis included a classification system based on the sentiment of news articles. The labeling of the sentiment was based on the global economic state during the release of the article. Of course this involves a lot of filtering, pre-processing, post-processing, and plenty of coffee.

The initial idea was simple: I am reading a lot of hacker news, marketing news and tutorials. But my own taste is not always the taste of the masses. Sometimes I am after quite "unpopular" topics. As I don't always look at new topics and also couldn't find an Blog which is posting good tutorials or news that interests me, I thought I give a news-classification engine a shot. I already have the knowledge how to crawl and process the news, so having a service like that was only a small step for me.