Live streaming channel recommendation using HITS algorithm

Yu Wen Liu, Chen Yi Lin, Jiun-Long Huang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

Live streaming services have been developed prosperously in recent years. With live streaming services, broadcasters broadcast their videos to attract large numbers of viewers to watch. Some live streaming services also provide a platform for viewers gathering together to watch channels and interact with others. Over hundreds of videos broadcast every day, recommending channels are necessary to help viewers finding channels of interests. In this paper, we propose the hybrid HITS algorithm for Channel Recommendation (HyHITS-CR) which adopts the HITS algorithm with modification to utilize the follower-followee relationship on channel recommendation. Besides, we also observe from real logs that some viewers may focus on few particular channels. Thus, HyHITS-CR also takes advantage on such observation to further improve the accuracy of recommendation. Experimental results show that the hit rate of HyHITS-CR is around 10% higher than other prior algorithms.

Original languageEnglish
Title of host publication2015 IEEE International Conference on Consumer Electronics - Taiwan, ICCE-TW 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages118-119
Number of pages2
ISBN (Electronic)9781479987443
DOIs
StatePublished - 20 Aug 2015
Event2nd IEEE International Conference on Consumer Electronics - Taiwan, ICCE-TW 2015 - Taipei, Taiwan
Duration: 6 Jun 20158 Jun 2015

Publication series

Name2015 IEEE International Conference on Consumer Electronics - Taiwan, ICCE-TW 2015

Conference

Conference2nd IEEE International Conference on Consumer Electronics - Taiwan, ICCE-TW 2015
CountryTaiwan
CityTaipei
Period6/06/158/06/15

Keywords

  • Accuracy
  • Clustering algorithms
  • Information management
  • TV
  • Testing
  • Videos
  • Watches

Fingerprint Dive into the research topics of 'Live streaming channel recommendation using HITS algorithm'. Together they form a unique fingerprint.

Cite this