Search services are now ubiquitously employed in searching for documents on the Internet and on enterprise intranets. Search services may exhibit different behavior depending on the type of information need, the quality of the search service, the ease of filtering results, the user’s domain knowledge and search experience. Users are thus faced with the selection of a search service in order to minimize cost, reduce uncertainty, and maximize the benefits derived for their efforts. This research develops a model of the search process and considers the noise effects of querying, search and filtering of results to derive a benefit measure for evaluating the search service. A methodology for comparing search services based on the benefit measure is presented along with an empirical analysis using three popular search services to validate the methodology. Our analysis revealed that the economic benefit of a search service is determined more by the information need type than by the search service itself. For a particular information need type, the value is determined primarily by the ease of filtering in the search service interface.
Shailaja Venkatsubramanyan and Stephen K. Kwan. "Using information noise to compute the economic benefit of a search service" Journal of Information Technology Management (2009): 1-11.