Title page for etd-0113117-153742


[Back to Results | New Search]

URN etd-0113117-153742
Author Tsung-Lin Yang
Author's Email Address No Public.
Statistics This thesis had been viewed 5347 times. Download 0 times.
Department Information Management
Year 2016
Semester 1
Degree Master
Type of Document
Language zh-TW.Big5 Chinese
Title Combining Trust Relationships and Super Users for Collaborative Recommendation
Date of Defense 2017-01-23
Page Count 55
Keyword
  • Collaborative Filtering
  • Recommendation System
  • Cold Start
  • Matrix Factorization
  • Trust
  • Abstract The network and mobile communication develop very quickly in recent years. We could get the data easily online. However, the speed of data transformation is growing rapidly and the quantity of data is increasing. When user search for the data, they don’t know how to choose the suitable information. Without appropriate method, users will waste more time to eliminate from unnecessary data.
    To solve this problem, this thesis proposes a method to build the recommendation system by matrix factorization. It combine trust relationships and super users to improve the efficiency of the matrix factorization recommendation system. Through the method of machine learning, it helps users screen the massive data and recommends the suitable information to users. In addition to the whole date, this thesis will analyze the different rating number users. And this thesis compares the result with traditional matrix factorization recommendation system. According to the experiment of this thesis, it discovers that when incoming users join the system. This research could provide the recommendation effectively, and it is the best contribute for this thesis.
    Advisory Committee
  • Yuh-Jiuan Tsay - chair
  • Bing-Chiang Jeng - co-chair
  • Wei-Po Lee - advisor
  • Files
  • etd-0113117-153742.pdf
  • Indicate in-campus at 3 year and off-campus access at 3 year.
    Date of Submission 2017-02-13

    [Back to Results | New Search]


    Browse | Search All Available ETDs

    If you have more questions or technical problems, please contact eThesys