• DocumentCode
    1903324
  • Title

    A Common Concept Description of Natural Language Texts as the Foundation of Semantic Computing on the Web

  • Author

    Ishizuka, Mitsuru

  • Author_Institution
    Sch. of Inf. Sci. & Technol., Tokyo Univ., Tokyo
  • fYear
    2008
  • fDate
    11-13 June 2008
  • Firstpage
    385
  • Lastpage
    385
  • Abstract
    Summary form only given. In order to intelligently process vast information on the Web, we need to make computers understand the meaning of the Web contents and manipulate them taking account of their semantics. Since text is the major medium conveying information, it is thus natural and reasonable to set it as the immediate target that the computer understands the meaning, while there are other types of media such as picture, movie, etc. Toward this direction, the activity of the semantic Web is going on. It aims to establish a standardized machine-readable description format of meta-data. However, the meta-data are only fragments of the Web contents. Unlike the semantic Web, we aim to describe the concept meaning expressed in the whole natural language texts with a common format that the computer can understand. We have designed concept description language (CDL) as a vehicle for this end, and started its standardization activity in W3C. There are several levels of the meaning of the texts, ranging from shallow level to deep one. While it is still difficult to make a consensus on how to describe the deep meaning, we think that a certain consensus can be attained on a way of describing the shallow meaning of the texts, based on the research results accumulated in the field of natural language processing such as machine translation over the last several decades. In CDL, besides lexicons, 45 relations are predefined as being necessary and sufficient for denoting every semantic relation between entities (lexicons in a simple case). These CDL relations can be used universally, while the ontologies in the semantic Web are domain dependent and thus cause some problematic situations. Current issues of CDL are, among others, an easy semi-automatic way of converting natural language texts into the CDL description, and an effective mechanism of executing semantic retrieval on the CDL database. We believe that CDL contributes to build a framework of next-generation Web which - provides the foundation for a variety of semantic computing. Also, CDL may contribute to overcome the language barrier among nations.
  • Keywords
    meta data; natural language processing; semantic Web; text analysis; Web content; concept description language; lexicons; machine-readable description format; metadata; natural language text; semantic Web; semantic computing; Computer networks; Information science; Intelligent networks; Intelligent sensors; Motion pictures; Natural languages; Pervasive computing; Semantic Web; Sensor phenomena and characterization; Vehicles; Common Concept Description; Natural Language Texts; Semantic Computing; Web;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Sensor Networks, Ubiquitous and Trustworthy Computing, 2008. SUTC '08. IEEE International Conference on
  • Conference_Location
    Taichung
  • Print_ISBN
    978-0-7695-3158-8
  • Electronic_ISBN
    978-0-7695-3158-8
  • Type

    conf

  • DOI
    10.1109/SUTC.2008.91
  • Filename
    4545789