Communities of Practice
From SSRG Annoki
Supporting wiki-based communities
People Involved
Cleo Espiritu Veselin Ganev Brendan Tansey
Ofer Arazy Stan Ruecker Eleni Stroulia
Wikis and blogs enable people to exchange information and opinions and easily share knowledge and ideas. Giving users the freedom to publish has substantially increased the amount and variety of content available; however, presenting a large amount of content relying only on the traditional hypermedia support (i.e., link navigation) can overwhelm readers, as most hypertext media lack an intuitive navigational structure that effectively guides users through the content.
We have developed a family of components that analyze wikis and blogs, infer semantic structures from their content and interactively visualize these structures to improve the users' access and understanding of their contents. Our experiments with several systems built with these components indicate that, indeed, such semantic support and visualizations help users make better use of Wikis and blogs and more easily perceive the knowledge communicated by their content.
Blog analysis for market analysis
Even more interestingly, blogs seem to compete with authoritative media outlets as a means for delivering news and opinions. The perceived power of blogs to inform and form, public opinion has given rise to several blog-mining services analyzing the blog entries to infer the reputation of their customers. Our work on visualizing and systematically analyzing several blogs in the same domain of interest (movies) shows that the "buzz" that a movie generates in the blogosphere correlates with public opinion and, to a great extent, predicts the movie success at the box office.