Paper: Vector Space Semantic Parsing: A Framework for Compositional Vector Space Models

ACL ID W13-3201
Title Vector Space Semantic Parsing: A Framework for Compositional Vector Space Models
Venue Continuous Vector Space Models and their Compositionality
Session
Year 2013
Authors

We present vector space semantic parsing (VSSP), a framework for learning compo- sitional models of vector space semantics. Our framework uses Combinatory Cate- gorial Grammar (CCG) to define a cor- respondence between syntactic categories and semantic representations, which are vectors and functions on vectors. The complete correspondence is a direct con- sequence of minimal assumptions about the semantic representations of basic syn- tactic categories (e.g., nouns are vectors), and CCG?s tight coupling of syntax and semantics. Furthermore, this correspon- dence permits nonuniform semantic repre- sentations and more expressive composi- tion operations than previous work. VSSP builds a CCG semantic parser respecting this correspondence; this semantic parser parses text into lambda calculus...