Web page recommendation based on Markov Logic Network

Recent research initiatives have addressed the need for improved performance of Web page recommendation accuracy that would profit many applications, e-business in particular. Different Web usage mining frameworks have been implemented for this purpose specifically Association rules and clustering. Each of these frameworks has its own strengths and weaknesses. An approach based on Markov Logic Network is proposed to do Web page recommendation, which take full use of four kinds of characteristics. The experiments show that the approach can achieve high accuracy on Web page recommendation.