Explore Courses Blog Tutorials Interview Questions
0 votes
in AI and Deep Learning by (50.2k points)

My day job is for an online browser based game, one that is small, with a very small staff. In fact, the majority of our staff are volunteers.

I am focused today on one aspect. I want to create an artificial intelligence system that will analyze our users database and report back on accounts that may possibly be run by the same user--which is clearly against our terms and conditions. This "duping" is the major time drain for our staff, and if I can speed it up by giving them a short list of names to check FIRST, I would.

The problem is, I'm not well versed in artificial intelligence. I understand very very basics, but have not successfully implemented a solution currently. I have been reading up on heuristic searches, specifically A* searches, and I "think" it may be appropriate for what I'm looking for, but I can't be sure.

So my question here is: Using an A* search, would it be possible to accurately analyze two user accounts data such as username, password, email, interactions between accounts, interactions between others, login times, activity times, etc. And if not, do you know of a system that would make it possible to analyze this amount of data, and give a "probability" that two accounts may be run by the same person?

1 Answer

0 votes
by (108k points)

If you have access to the user's game-movements log you could use clustering to group users who play 'similar'. Once you have the clusters you could use the IP to filter users inside each cluster. 

Coming to your question, as you know that A* algorithm is used for pathfinding, if you can detect similar paths of two different accounts then you can easily identify the fraud accounts.

A* is commonly used for the common pathfinding problem in applications such as video games but was originally designed as a general graph traversal algorithm. It finds applications to diverse problems, including the problem of parsing using stochastic grammars in NLP. So yes it is possible to analyze this amount of data and give a "probability" that two accounts may be run by the same person.

Browse Categories