Download Now Free registration required
Web search is often an inherently collaborative task and therefore Collaborative Web Search (CWS) has taken on a new meaning. Despite the absence of explicit collaboration features from mainstream search engines, there is clear evidence that users implicitly engage in many different forms of collaboration as they search. Co-located systems offer a collaborative search experience for multiple searchers at a single location, often a single PC. Failed searches are largely due to the mismatch between the query-space of the searcher and the document-space of the search engine index. The result is a significant vocabulary gap that leads to poor search performance. There are many scenarios in which search can be viewed as a community-oriented activity. A search session is considered successful if at least one result is selected by the searcher. The lack of any result selections is a good indication that no relevant results have been noticed. New opportunities have emerged to extend the basic CWS concept. Search communities can be now automatically identified and their promotions can be combined to deliver further improvements in search quality. Profiling result selections according to their snippet terms, with significant improvements in promotion quality accruing to the richer representation format is examined. This extended approach to CWS has led to a novel approach to generating community-based result summaries. These summaries have been shown to offer better precision/recall characteristics than more conventional summarization techniques.
- Format: PDF
- Size: 433.14 KB