tailieunhanh - Handbook of Multimedia for Digital Entertainment and Arts- P2

Handbook of Multimedia for Digital Entertainment and Arts- P2: The advances in computer entertainment, multi-player and online games, technology-enabled art, culture and performance have created a new form of entertainment and art, which attracts and absorbs their participants. The fantastic success of this new field has influenced the development of the new digital entertainment industry and related products and services, which has impacted every aspect of our lives. | 16 G. Lekakos et al. ADMINISTRATOR Set Recommendation Method Dataset _Collaborative Filtering Content-Based Recommendations Fig. 4 Method selection in MoRe Hybrid Substitute Recommendations Hybrid Switclüng Recommendations Fig. 5 Ranked list of movie recommendations Recommendation Algorithms Pure Collaborative Filtering Our collaborative filtering engine applies the typical neighbourhood-based algorithm 8 divided into three steps a computation of similarities between the target and the remaining of the users b neighborhood development and c computation of prediction based on weighted average of the neighbors ratings on the target item. 1 Personalized Movie Recommendation 17 For the first step as formula 1 illustrates the Pearson correlation coefficient is used. E X - X Yt - Y r_ i Je Xi - X 2 E Yi - Y 2 ii 1 where X and Y are the ratings of users X and Y for movie I while X Y refer to the mean values of the available ratings for the users X and Y. However in the MoRe implementation we used formula 2 given below which is equivalent to formula 1 but it computes similarities faster since it does not need to compute the mean rating values. n represents the number of commonly rated movies by users X and Y. n XiYi Xi Yi i i i 2 n p Xi2 - 2 n Yi2 - Note that in the above formulas if either user has evaluated all movies with identical ratings the result is a divide by zero error and therefore we decided to ignore users with such ratings. In addition we devaluate the contribution of neighbors with less than 50 commonly rated movies by applying a significance weight of n o where n is the number of ratings in common 32 . At the neighborhood development step of the collaborative filtering process we select neighbors with positive correlation to the target user. In order to increase the accuracy of the recommendations prediction for a movie is produced only if the neighbourhood consists of at least 5 neighbors. To compute an arithmetic prediction for a movie the weighted