PAT-tree-based adaptive keyphrase extraction for intelligent Chinese information retrieval

Considering the urgent need for keyphrase extraction techniques in intelligent information retrieval, in this paper we present a PAT-tree-based adaptive approach, which is critical and fundamental for Chinese and other oriental languages. Compared with conventional dictionary-based approaches, the proposed approach can reduce the reliance on rigid lexicon and sophisticated word segmentation, and compared with conventional statistics-based approaches, it can handle phrases composed of high-frequency words regardless of phrase length. Furthermore, the approach has been designed carefully with Internet utilization in mind. For instance, it can be easily integrated into text retrieval systems to provide automatic term suggestion and is adaptable to changes of the database content. The proposed approach has been successfully used in several information retrieval applications, such as automatic term suggestion, domain-specific lexicon construction, book indexing and document classification. Many Chinese and oriental language processing applications are, therefore, able to move ahead from the character level to the word or phrase level. (C) 1999 Elsevier Science Ltd. All rights reserved.