|visits||member for||2 years, 6 months|
|seen||Jul 6 '12 at 8:54|
|How to implement a system to hide spammed user content?|
Using Machine Learning algorithms, you could improve your click-fraud detection. Support Vector Machines are probably the best bet in your case. After a quick search on Google I could find this paper: dl.acm.org/citation.cfm?id=2020471, from the abstract sounds good.