본문 바로가기
HOME> 논문 > 논문 검색상세

학위논문 상세정보

Swarm collaborative filtering through fish school search 원문보기

  • 저자

    Andri Fachrur Rozie

  • 학위수여기관

    Graduate School, Korea University

  • 학위구분

    국내석사

  • 학과

    컴퓨터ㆍ전파통신공학과

  • 지도교수

    印浩

  • 발행년도

    2014

  • 총페이지

    v, 47장

  • 키워드

    Collaborative Filtering Swarm Algorithm Fish School Search;

  • 언어

    eng

  • 원문 URL

    http://www.riss.kr/link?id=T13541927&outLink=K  

  • 초록

    Collaborative filtering is one of the most popular techniques for building recommendation systems. The basic assumption of collaborative filtering is that users who had similar preferences in the past are likely to have similar preferences in the future. Traditional collaborative filtering calculates the similarity between users based on rating of co-rated item. However, current methods often lead to inaccurate prediction because they not enough to determine user similarity, especially in the case of sparse data when user only rates a small number of items. This paper concentrated on enhancing the prediction accuracy by including user's demographic information and item interests besides rating as similarity features. Nevertheless, the weights of each feature are not equal between each other and need to be estimated before building prediction. The proposed approach is adaptive feature weight learner, which applying the Fish School Search (FSS) algorithm to learn the best combination of feature weight for each user to generate better prediction. The research work of this paper proves that this approach is able to increase the accuracy of the predictions. To the best of my knowledge, this is the first time Fish School Search applied in recommendation system domain.


 활용도 분석

  • 상세보기

    amChart 영역
  • 원문보기

    amChart 영역