Adverting is one of the most important profit models in internet world. With more than ten years development, internet advertisement becomes smarter than ever and RTB (Real Time Bidding) is becoming the major share in the whole internet advertisement. Under RTB environment, advertisers need more accurate and efficient advertising technology than before. Targeted advertising integrates game theory, big data analysis, data mining and advertising technology and helps to publish advertisement to audiences precisely. Social network is the reflection of real world on internet built on six degrees of separation theory, and it has massive users and enormous access every day and thus collects massive personal information which can help to improve targeted advertising. This paper presents a framework to use social network analysis to improve targeted advertising and introduces clustering and cosine similarity as specified algorithm in the framework.
A study on frame transition of personal information leakage, 1984-2014: social network analysis approach