MICRO ASPECT MINING IN A COLLABORATIVE ENVIRONMENT USING A NOVEL DISCRIMINATIVE INFINITE HIDDEN MARKOV MODEL

Collaborative environments, which enable companywide global teams to identify the source of the problem and develop a response, are an excellent antidote to a lack of preparedness. Knowledge sharing is an activity through which knowledge is exchanged among people, friends, families, communities or organizations. In order to gain knowledge, user may try to acquire similar information on the web in this collaborative environment. The framework f ormulates tasks from sessions. There is no existing technique for micro aspect mining. A novel discriminative infinite Hidden Markov Model is proposed to mine micro aspects and evolution patterns in each task. The goal is not finding domain experts but a p erson who has the desired specific knowledge. In this project first summarizing web surfing data into fine grained aspects, and then search over these aspects. This strategy is compared with searching advisors directly over sessions both analytically and e mpirically.