How incorrect information delivers correct search results: A pragmatic analysis of queries

Abstract

Although the accuracy of the information relied on in formulating known-item search queries is generally regarded as critical to the search success of those queries, it has been observed that in some cases, such as music information retrieval, searching is often successful even when the information the queries are based on is flawed. Resolving this puzzle is important to a comprehensive understanding of information seeking in general and, perhaps, the success of new social indexing strategies in particular. One promising source of insight is the area of linguistics known as pragmatics. We show how a particular concept from pragmatics, the attributive/referential distinction, can illuminate how queries can function effectively independently of the apparent accuracy of the information they seem to rely on. This suggests possible new directions for the development of more effective information retrieval systems.

Notes from the underground city of disinformation: A conceptual investigation

“…In the field of library and information science (LIS), the accuracy of information is generally regarded as of primary importance for successful results in search and reference service (Baker & Lancaster, 1991). Inaccurate information, on the other hand, is typically viewed as information that is simply flawed by mistakes, ambiguity, or vagueness, requiring various kinds of “corrections” in order to achieve better results in search and discovery (Lee & Renear, 2008). Inaccurate information is typically not the main focus in the majority of studies in the LIS field, unless viewed as something that should be identified and corrected in some way.…”