본문 바로가기
  • Home

Clustering method for similar user with Miexed Data in SNS

  • Journal of The Korea Society of Computer and Information
  • Abbr : JKSCI
  • 2015, 20(11), pp.25-30
  • Publisher : The Korean Society Of Computer And Information
  • Research Area : Engineering > Computer Science

송형민 1 LEE SANG JOON 1 Ho-Young Kwak 1

1제주대학교

Accredited

ABSTRACT

The enormous increase of data with the development of the information technology make internet users to be hard to find suitable information tailored to their needs. In the face of changing environment, the information filtering method, which provide sorted-out information to users, is becoming important. The data on the internet exists as various type. However, similarity calculation algorithm frequently used in existing collaborative filtering method is tend to be suitable to the numeric data. In addition, in the case of the categorical data, it shows the extreme similarity like Boolean Algebra. In this paper, We get the similarity in SNS user's information which consist of the mixed data using the Gower's similarity coefficient. And we suggest a method that is softer than radical expression such as 0 or 1 in categorical data. The clustering method using this algorithm can be utilized in SNS or various recommendation system.

Citation status

* References for papers published after 2022 are currently being built.