• Summary
• Related Documents
  • Active Bibliography
  • Co-citation
• Version History

Abstract:

Abstract. The paper is concerned with the role of conceptual representations in access to information, as for example, from the World Wide Web. It contrasts two quite different traditions for doing this: Information Retrieval (IR) and more recently Information Extraction (IE), a development of the natural language processing tradition within Artificial Intelligence (AI). The former has been statistical in nature and largely representation-free (though we discuss exceptions), while the latter has been based on representations making use of ontologies and lexicons in semantics and grammars in syntax. However, this distinction has been eroded by the growth in recent years of machine learning methods in IE, which have attempted to match IE performance but with methods less committed to representations: some have no representations, and some seek to learn them automatically from cases of their assignment. We discuss ways of resolving this division of approaches, a deep and historical issue about the ultimate role of representations in information access. We suggest that modes of use of the Web (e.g. the use of short questions by real users rather than the long artificial `queries ' that statistical methods require) will tend to favour representational methods. We then discuss the crucial example of question-answering in a web environment of information access, as exemplified in the recent TREC competition track on question answering, and suggest that, although indecisive at the moment, this is an ideal forum in which the old issue of conceptual representations may be settled. 1

Citations