This really sounds like a machine learning problem--why hand engineer all these rules yourself (which can be error prone in any case) when a program could do it for you? If you can figure out how to get a large data set together and come up with a good general representation of the data, you're half way there.
I'd consider taking a look at recurrent neural networks if you're looking at your data as a time series or if its more of a static problem that doesn't take into account the last N things that came before, you might consider tree based methods or even DBNs if you can get a lot of data together (say 50000-100000 samples)
Potential pitfall: if you're using a NN based approach, you will judge results based on test performance, but you won't get as much insight into the "rules" your network has learned.
I'd consider taking a look at recurrent neural networks if you're looking at your data as a time series or if its more of a static problem that doesn't take into account the last N things that came before, you might consider tree based methods or even DBNs if you can get a lot of data together (say 50000-100000 samples)
Potential pitfall: if you're using a NN based approach, you will judge results based on test performance, but you won't get as much insight into the "rules" your network has learned.