Learning a Compositional Semantics for Freebase with an Open Predicate Vocabulary

Learning a Compositional Semantics for Freebase with an Open Predicate Vocabulary Jayant Krishnamurthy Carnegie Mellon University 5000 Forbes Avenue Pittsburgh, PA 15213 jayantk@cs.cmu.edu Tom M. Mitchell Carnegie Mellon University 5000 Forbes Avenue Pittsburgh, PA 15213 tom.mitchell@cmu.edu Abstract We present an approach to learning a model- theoretic semantics for natural language tied to Freebase. Crucially, our approach uses an open predicate vocabulary, enabling it to produce denotations for phrases such as “Re- publican front-runner from Texas” whose se- mantics cannot be represented using the Free- base schema. Our approach directly converts a sentence’s syntactic CCG parse into a log- ical form containing predicates derived from the words in the sentence, assigning each word a consistent semantics across sentences. This logical form is evaluated against a learned probabilistic database that defines a distribu- tion over denotations for each textual pred- icate. A training phase produces this prob- abilistic database using a corpus of entity- linked text and probabilistic matrix factoriza- tion with a novel ranking objective function. We evaluate our approach on a compositional question answering task where it outperforms several competitive baselines. We also com- pare our approach against manually annotated Freebase queries, finding that our open pred- icate vocabulary enables us to answer many questions that Freebase cannot. 1 Introduction Traditional knowledge representation assumes that world knowledge can be encoded using a closed vocabulary of formal predicates. In recent years, semantic parsing has enabled us to build compo- sitional models of natural language semantics us- ing such a closed predicate vocabulary (Zelle and Mooney, 1996; Zettlemoyer and Collins, 2005). These semantic parsers map natural language state- ments to database queries, enabling applications such as answering questions using a large knowl- edge base (Yahya et al., 2012; Krishnamurthy and Mitchell, 2012; Cai and Yates, 2013; Kwiatkowski et al., 2013; Berant et al., 2013; Berant and Liang, 2014; Reddy et al., 2014). Furthermore, the model- theoretic semantics provided by such parsers have the potential to improve performance on other tasks, such as information extraction and coreference res- olution. However, a closed predicate vocabulary has inher- ent limitations. First, its coverage will be limited, as such vocabularies are typically manually con- structed. Second, it may abstract away potentially relevant semantic differences. For example, the se- mantics of “Republican front-runner” cannot be ad- equately encoded in the Freebase schema because it lacks the concept of a “front-runner.” We could choose to encode this concept as “politician” at the cost of abstracting away the distinction between the two. As this example illustrates, these two problems are prevalent in even the largest knowledge bases. An alternative paradigm is an open predicate vocabulary, where each natural language word or phrase is given its own formal predicate. This paradigm is embodied in both open information ex- traction (Banko et al., 2007) and universal schema (Riedel et al., 2013). Open predicate vocabularies have the potential to capture subtle semantic distinc- tions and achieve high coverage. However, we have yet to develop compelling approaches to composi- tional semantics within this …

Content truncated to 500 words