Bee Hive at Work: Story Tracking Case Study

Information can change rapidly on the web. For example, news may hint some new story starts to develop. Many more news related to the original event begin to pour in the web. Imagine a person interested in how the story develops. It may be very difficult to trace it by trying to find the most relevant pages with most recent news on it. Our goal is to support user who wants to keep track of a developing story. We propose an approach and a system based on a bee hive model. The problem we focus on in this paper is that it is not possible to download all the pages using e.g. the breadth-first algorithm, nor to constantly revisit all the pages to see if new information were added. We propose to use a focused crawler to download the pages. With a prototype of our system, we performed a case study that shows that the system is able to collect relevant pages, it can monitor the story being developed during the search and it can even reconstruct the story backwards in time.