Searching People on the Web According to Their Interests
Bing Liu · 2002
Due to lack of structural data and information explosion on the World Wide Web (WWW), searching for useful information is becoming increasingly a difficult task. Traditional search engines on the Web render some form of assistance, but perform sub-optimally when dealing with context sensitive queries. To overcome this, niche search engines serving specific Web communities evolved. These engines index only pages of high quality and relevance to a specific domain and make use of context information for searching. This poster presents BullsI Search (publicly available at http://dm2.comp.nus.edu.sg), a fielded domain specific search engine that helps Web users locate computer science faculty members' homepages and email addresses by specifying a research interest, teaching interest or name. The system is able to automatically extract and index research interests, teaching interests, owners' email addresses and names from a set of discovered homepages. Experimental evaluation shows that the proposed algorithms are very accurate and are independent of structures in different homepages.