Interactive HMM construction based on interesting sequences

The paper presents a method of interactive construction of global Hidden Markov Models based on local patterns discovered in se- quence data. The method works by finding interesting sequences whose probability in data differs from that predicted by the model. The patterns are then presented to the user who updates the model using their un- derstanding of the modelled domain. It is argued that such an approach leads to more understandable models than automated approaches. An application to modelling webpage visitors behavior is presented, showing the strengths of the proposed approach.