Web Search Using Dynamic Keyword Suggestion

Web search has become an essential task for most people. As the Web grows rapidly, effective searches have grown increasingly important. Most of us, however, have experienced frustration in trying to search for something on the Web. In existing keyword-based Web search, the user has to come up with keywords for a query, and a search engine passively retrieves pages based solely on what the user types in. Deciding on effective keywords is vital but often difficult. In this paper, we focus on the problem of how to help users find the right keywords in the first place, and our method is based on the idea that selecting from a suggested keyword list can be a more effective strategy in deciding on keywords for a query. We propose a new search method based on keyword suggestion and selection. The user is guided by a list of active keywords that is suggested dynamically during the search by a search engine. This active keyword list helps the user choose keywords that are more relevant to the search through recognition and guided selection. Our method is based on formal concept analysis and is complementary to the existing user-dependent keyword-based Web and desktop search engines. Based on the proposed method, a prototype system is implemented for directory search. Our method will be especially useful in domain-specific Web search, and can be extended to weighted keywords by applying the conceptual scaling process to a many-valued formal context.