FINDING TARGET GROUPS FOR MUTUAL INSURANCE: WITH COMMUNITY EXTRACTION FROM SOCIAL NETWORKS
Community extraction techniques have been used in many areas, such as collaborative recommendation, knowledge sharing, information spreading and so on.But there has been no published works to apply these methods in mutual insurance.Potential mutual insurance groups should meet two conditions: sharing common interests and having relatively close ties.To fill up this gap, we propose a weighted community extraction approach based on content and links simultaneously.We first define our interested topics and pick the related users.Then, relationship based on content and links is calculated respectively.After combining the two kinds of relationship by variable weighting parameters, links analysis is conducted for community extraction.Experiments on the dataset collected from Delicious.com show that our method can be both flexible and effective.As we can choose weighting parameters according to practical problems, this method is well suited for mutual insurance case.
Mutual Insurance Community Extraction Social Network Target Marketing
Yuan Wei Lele Huang Huiwen Wang
School of Economics and Management, Beihang University, Beijing 100191, China;Research Center of Complex Data Analysis, Beihang University, Beijing 100191, China
国际会议
The 12th International Conference on Industrial Management(第十二届工业管理国际会议)
成都
英文
321-326
2014-09-03（万方平台首次上网日期，不代表论文的发表时间）