A personalized recommendation algorithm based on approximating the singular value decomposition (ApproSVD)

Personalized recommendation is, according to the user's interest characteristics and purchasing behavior, to recommend information and goods to users in which they may be interested. With the rapid development of Internet technology, we have entered the era of information explosion, where huge amounts of information are presented at the same time. On one hand, it is difficult for the user to discover information in which he is most interested, on the other hand, general users experience difficult in obtaining information which very few people browse. In order to extract information in which the user is interested from a massive amount of data, we propose a personalized recommendation algorithm based on approximating the singular value decomposition (SVD) in this paper. SVD is a powerful technique for dimensionality reduction. However, due to its expensive computational requirements and weak performance for large sparse matrices, it has been considered inappropriate for practical applications involving massive data. Finally, we present an empirical study to compare the prediction accuracy of our proposed algorithm with that of Drineas's LINEARTIMESVD algorithm and the standard SVD algorithm on the Movie Lens dataset, and show that our method has the best prediction quality. © 2012 IEEE.

Investor attention, analyst recommendation revisions, and stock prices

We provide evidence that investors underreact after analysts' recommendation upgrades; however, price reactions are faster after downgrades. We measure individual investors' attention using Google's search volume index. Our findings indicate that, after upgrades, stocks that enjoy greater individual investors' attention underreact significantly more compared to stocks that receive high level of attention from institutional investors. On the other hand, after recommendation downgrades, stocks with higher levels of prior attention from individual investors overreact and show a significantly greater price reversal compared to stocks that received high level of attention from institutional investors. Our results suggest that attentive individual investors may not be rational; hence investor attention and investor sophistication are important for price discovery in the market.

FUIR: Fusing user and item information to deal with data sparsity by using side information in recommendation systems

Recommendation systems adopt various techniques to recommend ranked lists of items to help users in identifying items that fit their personal tastes best. Among various recommendation algorithms, user and item-based collaborative filtering methods have been very successful in both industry and academia. More recently, the rapid growth of the Internet and E-commerce applications results in great challenges for recommendation systems as the number of users and the amount of available online information have been growing too fast. These challenges include performing high quality recommendations per second for millions of users and items, achieving high coverage under the circumstance of data sparsity and increasing the scalability of recommendation systems. To obtain higher quality recommendations under the circumstance of data sparsity, in this paper, we propose a novel method to compute the similarity of different users based on the side information which is beyond user-item rating information from various online recommendation and review sites. Furthermore, we take the special interests of users into consideration and combine three types of information (users, items, user-items) to predict the ratings of items. Then FUIR, a novel recommendation algorithm which fuses user and item information, is proposed to generate recommendation results for target users. We evaluate our proposed FUIR algorithm on three data sets and the experimental results demonstrate that our FUIR algorithm is effective against sparse rating data and can produce higher quality recommendations.

Exploring the effectiveness of true abnormal data elimination in context-aware web services recommendation

Recent years have witnessed a growing interest in context-aware recommender system (CARS), which explores the impact of context factors on personalized Web services recommendation. Basically, the general idea of CARS methods is to mine historical service invocation records through the process of context-aware similarity computation. It is observed that traditional similarity mining process would very likely generate relatively big deviations of QoS values, due to the dynamic change of contexts. As a consequence, including a considerable amount of deviated QoS values in the similarity calculation would probably result in a poor accuracy for predicting unknown QoS values. In allusion to this problem, this paper first distinguishes two definitions of Abnormal Data and True Abnormal Data, the latter of which should be eliminated. Second, we propose a novel CASR-TADE method by incorporating the effectiveness of True Abnormal Data Elimination into context-aware Web services recommendation. Finally, the experimental evaluations on a real-world Web services dataset show that the proposed CASR-TADE method significantly outperforms other existing approaches.