본문 바로가기
  • Home

Meaning-Flow Based Clustering for Document Retrieval in a Large Document Set

  • Journal of Knowledge Information Technology and Systems
  • Abbr : JKITS
  • 2013, 8(4), pp.37-42
  • Publisher : Korea Knowledge Information Technology Society
  • Research Area : Interdisciplinary Studies > Interdisciplinary Research
  • Published : August 31, 2013

SAJOON PARK 1

1대구한의대학교

Accredited

ABSTRACT

It is inconvenient that users retrieve information from documents efficiently and precisely. Clustering documents is a function for efficient information retrieval from massive document set. In this paper, we use meaning particle unit rather than document unit for document retrieval. We present a method with using an ontology that makes a document into paragraphs based on meaning flow. As the process of classification is done based on meaning paragraph, it is possible to achieve meaning-based clustering. The processing unit of clustering is shrunk from a document to a paragraph. Therefore, paragraph-based retrieval makes it possible for user to retrieve information in a document. We performed some experiments by using Reuter-21578 document set and the results showed the performance of meaning-flow based clustering was better than the performance of documents-based clustering.

Citation status

* References for papers published after 2023 are currently being built.