10th EAI International Conference on Communications and Networking in China

Research Article

Weighted Association Rule Mining based on PCA Algorithm in Wireless Communication Network

  • @INPROCEEDINGS{10.4108/eai.15-8-2015.2260795,
        author={Panfeng Zhang and Fangping Liu and Shilian Wang and Eryang Zhang},
        title={Weighted Association Rule Mining based on PCA Algorithm in Wireless Communication Network},
        proceedings={10th EAI International Conference on Communications and Networking in China},
        publisher={IEEE},
        proceedings_a={CHINACOM},
        year={2015},
        month={9},
        keywords={weighted association rule mining (warm) wireless communication network pca},
        doi={10.4108/eai.15-8-2015.2260795}
    }
    
  • Panfeng Zhang
    Fangping Liu
    Shilian Wang
    Eryang Zhang
    Year: 2015
    Weighted Association Rule Mining based on PCA Algorithm in Wireless Communication Network
    CHINACOM
    IEEE
    DOI: 10.4108/eai.15-8-2015.2260795
Panfeng Zhang1,*, Fangping Liu2, Shilian Wang1, Eryang Zhang1
  • 1: National University of Defense Technology
  • 2: 65067 Troops, PLA
*Contact email: ryekeey@163.com

Abstract

The concept of finding the significant communication relation in the wireless network only with the behavioral data is of great importance for non-authorized monitor. A valuable technology to solve the problem is the weighted association rules mining algorithm. However, classical models ignore the difference between items, and many weighted association rule mining algorithms do not work without a pre-assigned weight or work with a high complexity. In this paper, we introduce a novel measure p-weight considering the property of the behavioral database and propose a weighted association rule mining algorithm based on the primary components analysis (PCA). Performance of the proposed algorithm are compared with HITS algorithm and the other existing algorithm. It is observed that for the behavioral database of the wireless network, there is drastic reduction in the computational time for the proposed algorithm.