Attribute Extraction from Synthetic Web Search Queries

The accuracy and coverage of existing methods for extracting attributes of instances from text in general, and Web search queries in particular, are limited by two main factors: availability of input textual data to which the methods can be applied, and inherent limitations of the underlying assumptions and algorithms being used. This paper proposes a weakly-supervised approach for the acquisition of attributes of instances from input data available in the form of synthetic queries automatically generated from submitted queries. The generated queries allow for the acquisition of additional attributes, leading to extracted lists of attributes of higher quality than with comparable previous methods.