IoT as a Service. Third International Conference, IoTaaS 2017, Taichung, Taiwan, September 20–22, 2017, Proceedings

Research Article

IoT Service Provider Recommender Model Using Trust Strength

Download
164 downloads
  • @INPROCEEDINGS{10.1007/978-3-030-00410-1_34,
        author={Weiwei Yuan and Chenliang Li and Donghai Guan and Guangjie Han and Feng Wang},
        title={IoT Service Provider Recommender Model Using Trust Strength},
        proceedings={IoT as a Service. Third International Conference, IoTaaS 2017, Taichung, Taiwan, September 20--22, 2017, Proceedings},
        proceedings_a={IOTAAS},
        year={2018},
        month={10},
        keywords={Service provider recommendation Trust-aware recommendation algorithm Recommender systems},
        doi={10.1007/978-3-030-00410-1_34}
    }
    
  • Weiwei Yuan
    Chenliang Li
    Donghai Guan
    Guangjie Han
    Feng Wang
    Year: 2018
    IoT Service Provider Recommender Model Using Trust Strength
    IOTAAS
    Springer
    DOI: 10.1007/978-3-030-00410-1_34
Weiwei Yuan,*, Chenliang Li1,*, Donghai Guan,*, Guangjie Han2,*, Feng Wang3,*
  • 1: Nanjing University of Aeronautics and Astronautics
  • 2: Hohai University
  • 3: Changzhou University
*Contact email: yuanweiwei@nuaa.edu.cn, lcljoric@gmail.com, dhguan@nuaa.edu.cn, hanguangjie@gmail.com, wfeng@cczu.edu.cn

Abstract

Recommendation algorithms predict users’ opinions towards IoT service providers, helping users finding things that might be of their interests. With the rapid development of IoT applications, various recommender models have been proposed for usage, trust-aware recommender models have been verified to have reasonable recommendation performances even in case of data sparseness. However, existing works did not consider the influence of distrust between users. They recommend items only base on the trust relations between users. We therefore propose a novel trust strength based IoT service provider recommender model which predicts ratings with recommendations given by recommenders with both trust and distrust relations with the active users. The trust strength also merges both local and structural information of users in the trust network. The experimental results show that the proposed method has better prediction accuracy and prediction coverage than the existing works. In addition, the proposed method is computational less expensive.