Integrating Context Information for Syntactic Disambiguation

Most natural language utterances are inherently ambiguous. Yet, despite the omnipresence of ambiguity, human communication still succeeds in most cases and even displays a remarkable robustness – quite in contrast to the majority of natural language applications today. The reason for this is that in processing natural language humans also integrate information from sources other than the utterance itself, linguistic or non-linguistic in nature. Humans access additional knowledge to enrich the semantic specification that guides disambiguation. One important source of additional semantic knowledge in humans is context information.

While a wide range of studies has systematically investigated the impact of context upon structural disambiguation in human sentence processing, there are surprisingly few attempts to model the integration of context in natural language processing applications. Motivated by effects during human sentence processing I propose a system architecture which permits to study the integration of context information in syntactic parsing. I hypothesise that integrating context information into syntactic constraint dependency parsing will significantly and substantially improve the accuracy of structural ambiguity resolution.

-- CristinaVertan -- 17 Apr 2007