A Competitve Attachment Model for Resolving Syntactic Ambiguities in Natural Language Parsing

Abstract

Linguistic ambiguity is the greatest obstacle to achieving practical computational systems for natural language understanding. By contrast, people experience surprisingly little difficulty in interpreting ambiguous linguistic input. This dissertation explores distributed computational techniques for mimicking the human ability to resolve syntactic ambiguities efficiently and effectively. The competitive attachment theory of parsing formulates the processing of an ambiguity as a competition for activation within a hybrid connectionist network. Determining the grammaticality of an input relies on a new approach to distributed communication that integrates numeric and symbolic constraints on passing features through the parsing network. The method establishes syntactic relations both incrementally and efficiently, and underlies the ability of the model to establish long-distance syntactic relations using only local communication within a network. The competitive distribution of numeric evidence focuses the activation of the network onto a particular structural interpretation of the input, resolving ambiguities. In contrast to previous approaches to ambiguity resolution, the model makes no use of explicit preference heuristics or revision strategies. Crucially, the structural decisions of the model conform with human preferences, without those preferences having been incorporated explicitly into the parser. Furthermore, the competitive dynamics of the parsing network account for additional on-line processing data that other models of syntactic preferences have left unaddressed. (Also cross-referenced as UMIACS-TR-95-55)