Title :
Application of Collaborative Filtering Algorithm in Education Platform Based on Pattern Recognition
Author_Institution :
Sch. of Math. & Comput. Sci., Jianghan Univ., Wuhan, China
Abstract :
In order to improve the accuracy of personalized recommendation technology, first give a definition for "empty state" in the state space of multidimensional semi Markov process, to obtain the extended multidimensional Markov process, which is combined with social network analysis theory to obtain social network information flow model. this model describes the process flow process of information between members in the network society. Then, based on social network information flow model, collaborative filtering algorithm SMRR (Semi-Markov and reward renewal) is put forward. Experimental results show that due to the comprehensive consideration of user\´s preferences and the effect of other members in the social network, the prediction accuracy of SMRR is obviously higher than that of the original algorithm.
Keywords :
Markov processes; collaborative filtering; educational administrative data processing; pattern recognition; recommender systems; social networking (online); SMRR; collaborative filtering algorithm; education platform; empty state definition; extended multidimensional Markov process; information process flow; multidimensional semi Markov process; network society; pattern recognition; personalized recommendation technology; semiMarkov and reward renewal; social network analysis theory; social network information flow model; state space; user preferences; Accuracy; Collaboration; Educational institutions; Filtering algorithms; Markov processes; Random variables; Social network services; Collaborative Filtering; Communication Technology; Electronic Commerce; Multidimensional Semi-Markov Process; Paid Semi-Markov Model; Social Network;
Conference_Titel :
Measuring Technology and Mechatronics Automation (ICMTMA), 2014 Sixth International Conference on
Conference_Location :
Zhangjiajie
Print_ISBN :
978-1-4799-3434-8
DOI :
10.1109/ICMTMA.2014.33