Acquisition of Noncontiguous Class Attributes from Web Search Queries

Proceedings of the 14th Conference of the European Chapter of the Association for Computational Linguistics, pages 386–394, Gothenburg, Sweden, April 26-30 2014. c⃝2014 Association for Computational Linguistics Acquisition of Noncontiguous Class Attributes from Web Search Queries Marius Pas¸ca Google Inc. 1600 Amphitheatre Parkway Mountain View, California 94043 mars@google.com Abstract Previous methods for extracting attributes (e.g., capital, population) of classes (Em- pires) from Web documents or search queries assume that relevant attributes oc- cur verbatim in the source text. The ex- tracted attributes are short phrases that correspond to quantifiable properties of various instances (ottoman empire, ro- man empire, mughal empire) of the class. This paper explores the extraction of non- contiguous class attributes (manner (it) claimed legitimacy of rule), from fact- seeking and explanation-seeking queries. The attributes cover properties that are not always likely to be extracted as short phrases from inherently-noisy queries. 1 Introduction Motivation: Resources such as Wikipedia (Remy, 2002) and Freebase (Bollacker et al., 2008) aim at organizing knowledge around classes (Food in- gredients, Astronomical objects, Religions) and their instances (wheat flower, uranus, hinduism). Due to inherent limitations associated with main- taining and expanding human-curated resources, their content may be incomplete. For example, attributes representing the energy (or energy per 100g) or solubility in water are available in both Wikipedia and Freebase for many instances of Food ingredients (e.g., for olive oil, honey, fennel). But the attributes are missing for some instances (e.g., cornmeal). Moreover, structured informa- tion about how long (it) lasts unopened or manner (it) helps in weight loss is generally missing for Food ingredients, from both resources. Such in- formation is also often absent from among the at- tributes acquired from either documents or queries by previous extraction methods (Pas¸ca et al., 2007; Van Durme et al., 2008). Previously extracted at- tributes tend to be short, often nominal, phrases like nutritional value and taste. Even when ex- tracted attributes are not nominal (Pas¸ca, 2012), they remain relatively short phrases such as good for skin. As such, previous attributes have limited ability to capture the finer-grained properties be- ing asked about in queries such as “how long does olive oil last unopened” and “how does honey help in weight loss”. The presence of such queries suggests that such information is relevant to Web users. Identifying noncontiguous properties, or attributes of interest to Web users, helps filling some of the gaps in existing knowledge resources, which otherwise could not be filled by attributes extracted with previous methods. Contributions: The contributions of this paper are twofold. First, it introduces a method for the acquisition of noncontiguous class attributes, from fact or explanation-seeking Web search queries like “how long does olive oil last unopened” or “how does honey help in weight loss”. The re- sulting attributes are more diverse than, and there- fore subsume, the scope of attributes extracted by previous methods. Indeed, previous meth- ods are unlikely to extract attributes as specific as length/duration (it) lasts unopened and man- ner (it) helps in weight loss, for the …

Content truncated to 500 words