Construction of a metadata space for a specific field of knowledge from an index of a document

✍ Scribed by Takafumi Nakanishi; Sadaya Kishimoto; Tetsuya Sakurai; Takashi Kitagawa

• Publisher: John Wiley and Sons
• Year: 2007
• Tongue: English
• Weight: 970 KB
• Volume: 38
• Category: Article
• ISSN: 0882-1666
• DOI: 10.1002/scj.20524

✦ Synopsis

Abstract

In this paper, we propose a method for generating a metadata space of associations between words from an index of a book dealing with a specific field of interest. In order to implement a search mechanism using a mathematical model of meaning that deals with a specific field of interest, it is necessary to generate a search space called a metadata space dealing with the field of interest. In the past, the metadata space has been generated by using a terminology dictionary and a highly specialized knowledge base. Also, it is difficult to generate the metadata space if there is no terminology dictionary that deals with the field of interest. The proposed method does not require a terminology dictionary, and it allows the generation of the metadata space without a highly specialized knowledge base. Further, by applying this metadata space generation method to a mathematical model of meaning, a search mechanism based on the associations between words in the search target data, such as documents and media data, in the specific field of interest can be implemented. In this paper, taking examples from the IT field, the effectiveness of the method is verified by comparing the use of a metadata space generated from a terminology dictionary and the use of a metadata space generated from an index to search graphic media data. © 2007 Wiley Periodicals, Inc. Syst Comp Jpn, 38(5): 30– 42, 2007; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/scj.20524