Research prediction based on hidden Markov model
Zuo Wang · 2005
Research prediction of based on Latent Markov model, only from to research conversation itself set out to predict user's next chaining user, can not catch users' real interest. This text use Hidden Markov Model, analyse users browse through the content of the webpage, catch users' interest of research further, do next chaining to predict, When the browsing sequence length increases gradually, the system capture user browsing information are more and more many, this time can refract the user the interest to be at, forecast the rate of accuracy also gradually increases. When the browsing sequence length is bigger than or was equal to 8, forecast the rate of accuracy already arrived 80%, have improved the rate of accuracy of predicting.