Crowd Aided Web Search: Concept and Implementation

Abstract

Although keyword-based search algorithms usually do their jobs well, they may sometimes yield weird results. Despite of the fact that theWeb is the largest database, comparing to relational databases, the set of search operations for the Web is still primitive. This paper proposes two ways to remedy this: first, advanced information sources should be created. The role of advanced information sources of the Web is analogous to views of relational databases. Second, we propose several data processing tools based on the concept of advanced information sources. With these two mechanisms, the researcher tries to distinguish the data-centric view from the presentation view of the Web. In the paper, both the concept and an implementation are proposed.