Abstract: Hidden Markov Model, as a statistical model, can get the probability of hidden status by calculating the sequcnce of observed status. In this paper, a recognition approach based on HMM was proposed to infer situation in smart space. The approach infers situation by calculating partly contexts of system-related, using HMM to model the hidden situations. We designed the recognition algorithm based on HMM. Our experimental results show that this method can make a good performs and get a higher efficiency.

Key words: Hidden markov model,Situation recognition,Smart space