A central tenet of classical information retrieval is that the user is driven by an information need. Schneiderman, Byrd, and Croft [SBC97] define information need as "the perceived need for information that leads to someone using an information retrieval system in the first place." But the intent behind a web search is often not informational -- it might be navigational (show me the url of the site I want to reach) or transactional (show me sites where I can perform a certain transaction, e.g., shop, download a file, or find a map). In fact as we show later, informational queries constitute less than 50% of web searches.
The main aim of this paper is to point out this difference and introduce and analyze a taxonomy of web searches. Secondly, we show how search engines evolved to deal with these web-specific needs. The remainder of the paper is organized as follows: in section 2 we discuss the classic model for information retrieval; section 3 introduces a taxonomy of web searches; section 4 presents some statistics extracted from AltaVista surveys and logs regarding the prevalence of various types of searches; section 5 analyzes the evolution of search engines in light of this taxonomy; section 6 discusses some related work; finally, section 7 draws certain conclusions and points to some further directions for research.
2. The classic model for information retrieval We start from the basic model used in many standard information retrieval references textbooks, for instance, van Rijsbergen [R79]. See also [BK94] and references therein for a detailed discussion. Essentially, a user, driven by an information need, constructs a query in some query language. The query is submitted to a system that selects from a collection of documents (corpus), those documents that match the query as indicated by certain matching rules. A query refinement process might be used to create new queries and/or to refine the results. (Figure 1) 数据挖掘研究院
