본문 바로가기
  • Home

Clustering of XML Documents using Tag Information with Kohonen Map

  • Journal of Knowledge Information Technology and Systems
  • Abbr : JKITS
  • 2010, 5(2), pp.35-41
  • Publisher : Korea Knowledge Information Technology Society
  • Research Area : Interdisciplinary Studies > Interdisciplinary Research
  • Published : April 30, 2010

SAJOON PARK 1 박현근 2

1대구한의대학교
2숭실대학교

Candidate

ABSTRACT

One of the important features for the XML document is the creation of arbitrary tags. In this paper, we make use of it for clustering XML documents. Tag feature vector and word feature vector are separately created . Clustering was performed by applying a Kohonen map. Because tags are necessary keywords, we utilized binary method for them. TF / IDF technique was used for word feature vector. Reuter-21578 collections are experimented. The results of experimentation is almost rate of 50% in recall and precision rate. In addition, the traditional classification algorithm, SVM and K-NN was also compared with our system. Performance of our results were 10% more than SVM, K_NN system.

Citation status

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