Đang chuẩn bị nút TẢI XUỐNG, xin hãy chờ
Tải xuống
A lexical knowledge base is a repository of computational information about concepts intended to be generally useful in many application areas including computational linguistics, artificial intelligence, and information science. It contains information derived from machine-readable dictionaries, the full text of reference books, the results of statistical analyses of text usages, and data manually obtained from human world knowledge. | LEXICAL KNOWLEDGE BASES Robert A. Amsler Natural-Language and Knowledge-Resource Systems SRI International Menlo Park California 94025 USA A lexical knowledge base is a repository of computational information about concepts intended to be generally useful in many application areas including computational linguistics artificial intelligence and information science. It contains information derived from machine-readable dictionaries the full text of reference books the results of statistical analyses of text usages and data manually obtained from human world knowledge. A lexical knowledge base is not intended to serve any one application but to be a general repository of knowledge about lexical concepts and their relationships. Thus natural-language parsers generators or other intelligent processors must be able to interface to the knowledge base and are expected to only extract those portions of its knowledge which they need for specific tasks. Likewise the knowledge base is designed built and maintained primarily as a repository-rather than a tool serving the needs of other computational processors. Just as human memory the knowledge base doesn t distinguish between useful knowledge and information for which it at present doesn t have any functional use. In this manner the knowledge base is a test bed for concept representation mechanisms and data structures rather than an adjunct to other computational processes. Investigations of machine-readable dictionaries over the last decade have shown that they can be computationally useful for tasks such as parsing computer-assisted instruction speech generation and content analysis. Sufficient knowledge of the contents of machine-readable dictionaries now exists to provide meaningful answers to questions concerning what additional information about lexical concepts will be needed to represent many aspects of human world knowledge. Machine-readable dictionaries are seen as providing an index into human knowledge. A .
