Personalized smart TV program recommender based on collaborative filtering and a novel similarity method

Towards comprehensive approaches for the rating prediction phase in memory-based collaborative filtering recommender systems

Collaborative filtering (CF), the most successful information filtering technique for recommender systems, is either memory-based or model-based. While the former is more accurate, its scalability compared to model-based is poor. Moreover, the similarity functions used by most recommender systems are compensatory and allow very high (pros) and very low (cons) scores to compensate each other. This paper presents a hybrid movie recommender system that retains memory-based CF accuracy, model-based CF scalability, and alleviates the compensation problem of similarity functions. The proposed recommender system relies on a compact user model and fuzzy concordance/discordance principle. The user model speeds up the online process of generating a set of like-minded users within which a memory-based CF is carried out. The inter users comparison is done by using fuzzy concordance/discordance principle to alleviate the similarity compensation problem. The pros and cons between users are measured separately and then the overall statement about them is obtained by balancing the pros and cons within the set of criteria. Besides our approach is fast and compact, computational results reveal that it outperforms the classical one.

Memory-based collaborative filtering (CF) aims at predicting the rating of a certain item for a particular user based on the previous ratings from similar users and/or similar items. Previous studies in finding similar users and items have several drawbacks. First, they are based on user-defined similarity measurements, such as Pearson Correlation Coefficient (PCC) or Vector Space Similarity (VSS), which are, for the most part, not adaptive and optimized for specific applications and data. Second, these similarity measures are restricted to symmetric ones such that the similarity between A and B is the same as that for B and A, although symmetry may not always hold in many real world applications. Third, they typically treat the similarity functions between users and functions between items separately. However, in reality, the similarities between users and between items are inter-related. In this paper, we propose a novel unified model for users and items, known as Similarity Learning based Collaborative Filtering (SLCF) , based on a novel adaptive bidirectional asymmetric similarity measurement. Our proposed model automatically learns asymmetric similarities between users and items at the same time through matrix factorization. Theoretical analysis shows that our model is a novel generalization of singular value decomposition (SVD). We show that, once the similarity relation is learned, it can be used flexibly in many ways for rating prediction. To take full advantage of the model, we propose several strategies to make the best use of the proposed similarity function for rating prediction. The similarity can be used either to improve the memory-based approaches or directly in a model based CF approaches. In addition, we also propose an online version of the rating prediction method to incorporate new users and new items. We evaluate SLCF using three benchmark datasets, including MovieLens, EachMovie and Netflix, through which we show that our methods can outperform many state-of-the-art baselines.

Summary form only given. As one of the most successful approaches to building recommender systems, collaborative filtering (CF) uses the known preferences of a group of users to make recommendations or predictions of the unknown preferences for other users. In this paper, we first introduce CF tasks and their main challenges, such as data sparsity, scalability, synonymy, gray sheep, shilling attacks, privacy protection, etc., and their possible solutions. We then present three main categories of CF techniques: memory-based, model-based, and hybrid CF algorithms (that combine CF with other recommendation techniques), with examples for representative algorithms of each category, and analysis of their predictive performance and their ability to address the challenges. From basic techniques to the state-of-the-art, we attempt to present a comprehensive survey for CF techniques, which can be served as a roadmap for research and practice in this area.

In online environments, a vast amount of data is explored daily on the internet, such as news, movies, audios, and books. The interest of the target user is the greatest demand for recommender systems, and getting suitable information is a challenge. To develop a recommender system, collaborative filtering (CF) approach considers users who have similar ratings. Therefore, can compute the similarity of users when there are enough rated by users to items. A significant challenge of the collaborative filtering approach is cold start problem, which is how to make a recommendation for users who have few ratings than others. This work proposes a collaborative filtering model based on applying fuzzy c-means clustering on user's truthfulness information. A new fuzzy user-based similarity measure formula is suggested which combines user's rating with fuzzy truthfulness information using a combination coefficient. The experimental results using Movie Lens data set have shown an improvement in recommendation under sparsity and cold start condition.

Memory-based collaborative filtering (CF) techniques have been widely implemented for predicting ratings to unseen items by aggregating ratings of similar users or items in recommender systems (RS). Usually, sufficient ratings from similar users or similar items are not available in the rating matrix, due to the data sparsity problem. Further, these techniques suffer from correlation based problems inherent in used similarity measures. Consequently, higher prediction accuracy cannot be achieved. In this paper, we propose the use of fuzzy Naive Bayesian (FNB) classifier for user based CF and item based CF for implicitly computing similarity between users as well as items on the basis of conditional probabilities and develop fuzzy Naive Bayesian classifier to user based CF (FNB-UB-CF) and item based CF (FNB-IB-CF). We further develop a hybrid RS (FNB-UB-IB-CF) by combining the proposed FNB-UB-CF and FNB-IB-CF. Their combinations would be helpful in alleviating the sparsity because both user ratings and item ratings are employed. Experimental results demonstrate that the proposed methods are indeed more robust against data sparsity and give better recommendation quality using a popular MovieLens dataset.

Collaborative Filtering, which plays an important role in the recommendation, is based on the way that users rate the items throughout history. Although the CF recommendation system is used widely, the performance of recommendation still needs improving. In order to improve the accuracy of CF, the weight of item and the factor of time are considered in this paper. A new similarity method is proposed by improving the traditional similarity function with the weight of items. When putting the similarity of items and the factor of time as the weight of the target item, the neighbors of a user is not identical for all his items. Experimental results from MAE, precision, recall, f1 represent that the algorithm proposed can improve the performance of recommendation system. Keyword: collaborative filtering; similarity calculation; weight of item; time factor

Recommender Systems (RS) have been implemented in the Technology enhanced Learning (TeL) field for facilitating, among others, Learning Object (LO) selection by teachers to support their daily teaching practice. In particular, memory-based collaborative filtering (CF) approaches have demonstrated promising results for real-life implementations of web-based Learning Object Repositories (LOR). Building on this, the contribution of this paper is an enhancement to existing memory-based CF RS methods, by adaptively selecting the teacher neighbors based on their co-rated LOs and the attribute similarity of the latter to the LO to be recommended. The evaluation results show a significant increase in the predictive accuracy of the adaptive RS approaches compared to their traditional benchmarks, signifying the proposed approach's capacity to enhance the accuracy of existing memory-based CF approaches.

Due to extensive growth of e-commerce industries, recommender systems are widely being used by many sites to provide better services to their customers by assisting them to finding their items of interest. However, there is still scope of improvement in finding the best technique to implement recommender systems. Many techniques have been developed to implement the recommender systems. Both in academic research and commercial applications, collaborative filtering algorithms are very popular and widely used techniques. Neighborhood-based or memory based methods are classic approaches to collaborative filtering methods, while model-based methods, notably matrix factorization techniques, are more modern approaches. In this work we perform the comparative analysis of various approaches of memory and model-based approaches with different similarity measures using MovieLens dataset with three variations 100 k, 1 M and ML small. The results indicate that the item-based approach with adjusted cosine similarity performs better in memory based and in model-based SVD performs better than other algorithms.KeywordsRecommender systemContent-basedCollaborative filteringSimilarity measureMemory basedModel-based

User-based collaborative filtering is one of the most-used methods for the recommender systems. However, it takes time to perform the method because it requires a full scan of the entire data to find the neighboring users of each active user, who have similar rating patterns. It also requires time-consuming computations because of the complexity of the algorithms. Furthermore, the amount of rating data in the recommender systems grows rapidly, as the number of users, items, and their rating activities tend to increase. Thus, a big data framework with parallel processing, such as Hadoop, is needed for the recommender systems. There are already many research studies on the MapReduce-based parallel processing method for collaborative filtering. However, most of the research studies have not considered the sequential-access restriction for executing MapReduce jobs and the minimization of the required full scan on the entire data on the Hadoop Distributed File System (HDFS), because HDFS sequentially access data on the disk. In this paper, we introduce an efficient MapReduce-based parallel processing framework for collaborative filtering method that requires only a one-time parallelized full scan, while adhering to the sequential access patterns on Hadoop data nodes. Our proposed framework contains a novel MapReduce framework, including a partial computation framework for calculating the predictions and finding the recommended items for an active user with such a one-way parallelized scan. Lastly, we have used the MovieLens dataset to show the validity of our proposed method, mainly in terms of the efficiency of the parallelized method.

Collaborative Filtering is one of the most successful techniques of Recommender Systems. Despite its success, similarity-based Collaborative Filtering methods suffer from inherent weakness: users tend to rate few items. As a result, the similarity is not easily computed. This paper aims to solve the above problem by introducing the trust metric into Collaborative Filtering. We develop a novel computation model of trust by incorporating the tastes of users. Then we propagate trust throughout the trust relationship network, and more potential neighbors can be found. At last, we make recommendations based on trust-based Collaborative Filtering. Experimental results on a real extremely sparse dataset have shown best performance of our method in terms of MAE and Coverage when compared with similarity-based Collaborative Filtering methods.

Nowadays, a recommendation system is an important technique in the development of electronic-commerce systems and the most popular approaches that use in many recommendation systems are a collaborative filtering algorithm. However, it still has problems such as scalability and sparse data. There are several previous methods used to deal with the weakness of collaborative filtering techniques such as a hybrid user model, but the results show their disadvantages in practical use. In this paper, we proposed a hybrid recommender system with review helpfulness features, which we have used to construct the hybrid model. In the experiment, the results of three recommendation techniques are compared: collaborative filtering based on hybrid user model, user-based collaborative filtering and our proposed method. The experimental results show that our proposed method is more efficient than other methods with the same dataset, in terms of accuracy. In addition, our proposed algorithm can decrease time consuming by constructing the hybrid model in the off-line phase and calculates the recommendation results in on-line phase.

Highly Recommended: Collaborative Filtering Gives Customers What They Want

Majority of the e-commerce sites implement Recommender Systems (RS) to help users navigate through the large search space and assist their decision making process by suggesting products that the user may like. Collaborative Filtering (CF) is the most successful and widely used algorithm in the domain of RS. However, due to the exponential growth of the web in terms of both content and number of users, CF based RS face serious scalability issues. To alleviate this problem, we propose a clustering based CF approach using two hierarchical space partitioning data structures — K-d tree and Quadtree. We cluster or partition the users’ space of the system on the basis of user location and then use the resultant clusters for predicting ratings of a target user. Since the CF based recommendation algorithm is applied separately to the clusters and not on the entire rating data, it helps in bringing down the runtime of the algorithm substantially. We further measure spatial autocorrelation indices in the clusters to justify our clustering method. However, our objective is not only to reduce the runtime but also to maintain an acceptable recommendation quality. This requirement is rightly addressed by the proposed method which assures scalability, by processing very large datasets using the same computing resource. Moreover our proposed clustering scheme is oblivious of the underlying CF algorithm. Results from the extensive experiments conducted, show that our hierarchical clustering based recommendation approach reduces runtime of the standard CF algorithms by about 88%, 82%, 79% and 85% for MovieLens-100K, MovieLens-1M, Book-Crossing and TripAdvisor data respectively, while maintaining good recommendation quality.

Collaborative filtering methods are utilized to provide personalized recommendations for users in order to alleviate information overload problem in different domains. Traditional collaborative filtering methods operate on a user-item matrix in which each user reveal her admiration about an item based on a single criterion. However, recent studies indicate that recommender systems depending on multi-criteria can improve accuracy level of referrals. Since multi-criteria rating-based collaborative filtering systems consider users in multi-aspects of items, they are more successful at forming correlation-based user neighborhoods. Although, proposed multi-criteria user-based collaborative filtering algorithms' accuracy results are very promising, they have online scalability issues. In this paper, we propose an item-based multi-criteria collaborative filtering framework. In order to determine appropriate neighbor selection method, we compare traditional correlation approaches with multi-dimensional distance metrics. Also, we investigate accuracy performance of statistical regression-based predictions. According to real data-based experiments, it is possible to produce more accurate recommendations by utilizing multi-criteria item-based collaborative filtering algorithm instead of a single criterion rating-based algorithm.