Tag: twitter

Feature-pivot methods: did something happen?

Cheer! When something happens, people talk differently. We can use that cue to notice that something happened.

In my previous post, I took a look at document-pivot methods in Topic Detection and Tracking. Document-pivot approaches use clustering to find out what people are talking about, but that’s not the only solution. In this post, I take a look at feature-pivot methods, the second type of Topic Detection and Tracking methods. Instead of looking at what people are talking about, these approaches look at how people are talking to detect that something happened.


Document-pivot methods: what’s happening?

Talk! The way people talk can tell us a lot about what's happening.

Topic Detection and Tracking has been around for more than 20 years, but during this time, there has been a lot of research. When researchers started creating systems, they went off in a few different directions. Earlier, I took a a brief look at Topic Detection and Tracking. In this post I take a look at one of the two main approaches to solving the problem: document-pivot methods.


Topic Detection and Tracking: a brief introduction

Topic Detection and Tracking is not an old research area. It was ‘only’ established in the late 90’s and it didn’t really pick up until a few years later. However, a lot has happened since then and today’s research is almost unrecognizable from back then. In this post, I take a brief look at what Topic Detection and Tracking is and at its history.