Abstract

Since users formulate their queries based on the common-sense knowledge that they already possess at the time of the search, Web search queries play two roles simultaneously: in addition to requesting new information, they also indirectly convey knowledge.
Inherently noisy search queries turn out to be a highly valuable resource for Web-based information extraction, specifically for the task of assembling lists of prominent class attributes (e.g., "side effects" and "generic equivalent" for drugs) from unstructured text.