StackOverflow Tech Trends Demo

Posted by Mark Brandon December 10, 2013

One of the more oft-pointed-out items about our front-end website was the lack of real-time demos showing the power of our favorite open source data exploration and analytics platform. A week ago, we released our StackOverflow Tech Trends demo to accomplish this.

We asked Kerala-based data scientist, Vineeth Mohan, to do an analysis and visualization on the 17 million plus tags on the popular tech Q&A site, StackOverflow. Over time, users can see the trend on the number of times the tag has been cited and the breakdown between questions and answers.

The more interesting graphic is when you compare the trend of two tags. It will come as no surprise to followers of Elasticsearch that, compared to the tag “solr,” you will see a steeper curve indicating that, while both tags are growing in popularity, the “elasticsearch” tag is taking up a greater share of the two. Anybody who appreciates shameless puns would say that “solr” is being rapidly eclipsed. (rim shot)

Some users will want to see WHO are the knowledgeable answerers for a given tag. It is no surprise that the most prolific writers for the “elasticsearch” tag work for the development team, including Boston-based instructor Igor Motov, Amsterdam-based Luca Cavanna, Paris-based Technical Advocate David Pilato, and Czech developer and (Re)Tire maintainer Karel Minařík.

Elasticsearch_tag_answerers.png#asset:37

Finally, the bottom of the page contains the top-most voted answers and questions. Fittingly, the most voted up answer is a face off between Elasticsearch, Sphinx, Lucene, Solr, and Xapian.

Top_Questions.png#asset:406

This demo resides on a 2-node m1.medium cluster with 30GB drives each and 1,000 provisioned IOPS. Click here to see the live demo. As always, we’re interested in your feedback.