Publication Date

1994

Abstract

The availability of on-line corpora is rapidly changing the field of natural language processing (NLP) from one dominated by theoretical models of often very specific linguistic phenomena to one guided by computational models that simultaneously account for a wide variety of phenomena that occur in real-world text. Thus far, among the best-performing and most robust systems for reading and summarizing large amounts of real-world text are knowledge-based natural language systems. These systems rely heavily on domain-specific, handcrafted knowledge to handle the myriad syntactic, semantic, and pragmatic ambiguities that pervade virtually all aspects of sentence analysis. Not surprisingly, however, generating this knowledge for new domains is time-consuming, difficult, and error-prone, and requires the expertise of computational linguists familiar with the underlying NLP system. This thesis presents Kenmore, a general framework for domain-specific knowledge acquisition for conceptual sentence analysis. To ease the acquisition of knowledge in new domains, Kenmore exploits an on-line corpus using symbolic machine learning techniques and robust sentence analysis while requiring only minimal human intervention. Unlike most approaches to knowledge acquisition for natural language systems, the framework uniformly addresses a range of subproblems in sentence analysis, each of which traditionally had required a separate computational mechanism. The thesis presents the results of using Kenmore with corpora from two real-world domains (1) to perform part-of-speech tagging, semantic feature tagging, and concept tagging of all open-class words in the corpus; (2) to acquire heuristics for part-ofspeech disambiguation, semantic feature disambiguation, and concept activation; and (3) to find the antecedents of relative pronouns.

Comments

This paper was harvested from CiteSeer

Download

Share

COinS