This is my first question on stack overflow. So, excuse the mistakes I make. I have a situation where I have a dataset which consists of 3 fields:Database useame, Source program and count(of the number times the source program has been accessed). I need an automated method of tracking down database user names accessing a certain source program more than a threshold. I don't have access to timestamps.
I had hit upon this [link] :http://www.slideshare.net/erikbe/collaborative-filtering-at-spotify-16182818
Spotify's collaborative filtering system uses a dataset with the fields (user,track,count). I was hoping I could use something similar (user-item implicit feedback)in my case. I could figure out which DB useames would access a certain source program more number of times than a certain threshold. I could raise alerts based on that.
Does this make sense? Could somebody help me out?
Thanks