top-N Recommendation Research Articles

Mixture Matrix Approximation for Collaborative Filtering

Matrix approximation (MA) methods are integral parts of today's recommender systems. In standard MA methods, only one feature vector is learned for each user/item, which may not be accurate enough to characterize the diverse interests of users/items. For instance, users could have different opinions on a given item, so that they may need different feature vectors for the item to represent their unique interests. To this end, this article proposes a mixture matrix approximation (MMA) method, in which we assume that the user-item ratings follow mixture distributions and the user/item feature vectors vary among different stars to better characterize the diverse interests of users/items. Furthermore, we show that the proposed method can tackle both rating prediction and the top-N recommendation problems. Empirical studies on MovieLens, Netflix and Amazon datasets demonstrate that the proposed method can outperform state-of-the-art MA-based collaborative filtering methods in both rating prediction and top-N recommendation tasks.

Smart Campus Microservice Recommendation Algorithm Based on Neighborhood Clustering and Time Collaboration

The basic algorithm research of smart campus construction is the basic path for education to benefit from the development of science and technology. Based on the application environment of higher education teaching and management, according to the teaching periodicity and the closed-loop characteristics of the business system in universities, a neighborhood-based user attribute clustering algorithm is used in this paper to form user clusters, which are coupled micro-application clustering with periodic time. Finally, through normalized conversion and Pearson correlation coefficient analysis, the Top-N recommendation list is obtained, and a recommendation algorithm suitable for smart campus microservice system is constructed. In this paper, matrix sparsity and cold start are discussed, and data compensation is carried out by analyzing the general behavior patterns of teachers and students, which effectively solves the problem of insufficient recommendation list caused by matrix sparsity and cold start. Upon data analysis before and after the algorithm deployment, the algorithm effectively reduces the number of times that users use the search function.

Scalable and Explainable 1-Bit Matrix Completion via Graph Signal Learning

One-bit matrix completion is an important class of positive-unlabeled (PU) learning problems where the observations consist of only positive examples, e.g., in top-N recommender systems. For the first time, we show that 1-bit matrix completion can be formulated as the problem of recovering clean graph signals from noise-corrupted signals in hypergraphs. This makes it possible to enjoy recent advances in graph signal learning. Then, we propose the spectral graph matrix completion (SGMC) method, which can recover the underlying matrix in distributed systems by filtering the noisy data in the graph frequency domain. Meanwhile, it can provide micro- and macro-level explanations by following vertex-frequency analysis. To tackle the computational and memory issue of performing graph signal operations on large graphs, we construct a scalable Nystrom algorithm which can efficiently compute orthonormal eigenvectors. Furthermore, we also develop polynomial and sparse frequency filters to remedy the accuracy loss caused by the approximations. We demonstrate the effectiveness of our algorithms on top-N recommendation tasks, and the results on three large-scale real-world datasets show that SGMC can outperform state-of-the-art top-N recommendation algorithms in accuracy while only requiring a small fraction of training time compared to the baselines.

Deep Collaborative Filtering with Multi-Aspect Information in Heterogeneous Networks

Recently, recommender systems play a pivotal role in alleviating the problem of information overload. Latent factor models have been widely used for recommendation. Most existing latent factor models mainly utilize the interaction information between users and items, although some recently extended models utilize some auxiliary information to learn a unified latent factor for users and items. The unified latent factor only represents the characteristics of users and the properties of items from the aspect of purchase history. However, the characteristics of users and the properties of items may stem from different aspects, e.g., the brand-aspect and category-aspect of items. Moreover, the latent factor models usually use the shallow projection, which cannot capture the characteristics of users and items well. Deep neural network has shown tremendous potential to model the non-linearity relationship between users and items. It can be used to replace shallow projection to model the complex correlation between users and items. In this paper, we propose a Neural network based Aspect-level Collaborative Filtering model (NeuACF) to exploit different aspect latent factors. Through modelling the rich object properties and relations in recommender system as a heterogeneous information network, NeuACF first extracts different aspect-level similarity matrices of users and items, respectively, through different meta-paths, and then feeds an elaborately designed deep neural network with these matrices to learn aspect-level latent factors. Finally, the aspect-level latent factors are fused for the top-N recommendation. Moreover, to fuse information from different aspects more effectively, we further propose NeuACF++ to fuse aspect-level latent factors with self-attention mechanism. Extensive experiments on three real world datasets show that NeuACF and NeuACF++ significantly outperform both existing latent factor models and recent neural network models.

A New Collaborative Filtering Approach Based on Game Theory for Recommendation Systems

Recommendation systems can help internet users to find interesting things that match more with their profile. With the development of the digital age, recommendation systems have become indispensable in our lives. On the one hand, most of recommendation systems of the actual generation are based on Collaborative Filtering (CF) and their effectiveness is proved in several real applications. The main objective of this paper is to improve the recommendations provided by collaborative filtering using clustering. Nevertheless, taking into account the intrinsic relationship between users can enhance the recommendations performances. On the other hand, cooperative game theory techniques such as Shapley Value, take into consideration the intrinsic relationship among users when creating communities. With that in mind, we have used SV for the creation of user communities. Indeed, our proposed algorithm preforms into two steps, the first one consists to generate communities user based on Shapley Value, all taking into account the intrinsic properties between users. It applies in the second step a classical collaborative filtering process on each community to provide the Top-N recommendation. Experimental results show that the proposed approach significantly enhances the recommendation compared to the classical collaborative filtering and k-means based collaborative filtering. The cooperative game theory contributes to the improvement of the clustering based CF process because the quality of the users communities obtained is better.

Enhancing Top-N Recommendation Using Stacked Autoencoder in Context-Aware Recommender System

Context-aware recommender systems (CARS) are a vital module of many corporate, especially within the online commerce domain, where consumers are provided with recommendations about products potentially relevant for them. A traditional CARS, which utilizes deep learning models considers that user’s preferences can be predicted by ratings, reviews, demographics, etc. However, the feedback given by the users is often conflicting when comparing the rating score and the sentiment behind the reviews. Therefore, a model that utilizes either ratings or reviews for predicting items for top-N recommendation may generate unsatisfactory recommendations in many cases. In order to address this problem, this paper proposes an effective context-specific sentiment based stacked autoencoder (CSSAE) to learn the concrete preference of the user by merging the rating and reviews for a context-specific item into a stacked autoencoder. Hence, the user's preferences are consistently predicted to enhance the Top-N recommendation quality, by adapting the recommended list to the exact context where an active user is operating. Experiments on four Amazon (5-core) datasets demonstrated that the proposed CSSAE model persistently outperforms state-of-the-art top-N recommendation methods on various effectiveness metrics.

Holistic Transfer to Rank for Top-N Recommendation

Recommender systems have been a valuable component in various online services such as e-commerce and entertainment. To provide an accurate top-N recommendation list of items for each target user, we have to answer a very basic question of how to model users’ feedback effectively. In this article, we focus on studying users’ explicit feedback, which is usually assumed to contain more preference information than the counterpart, i.e., implicit feedback. In particular, we follow two very recent transfer to rank algorithms by converting the original feedback to three different but related views of examinations, scores, and purchases, and then propose a novel solution called holistic transfer to rank (HoToR), which is able to address the uncertainty challenge and the inconvenience challenge in the existing works. More specifically, we take the rating scores as a weighting strategy to alleviate the uncertainty of the examinations, and we design a holistic one-stage solution to address the inconvenience of the two/three-stage training and prediction procedures in previous works. We then conduct extensive empirical studies in a direct comparison with the two closely related transfer learning algorithms and some very competitive factorization- and neighborhood-based methods on three public datasets and find that our HoToR performs significantly better than the other methods in terms of several ranking-oriented evaluation metrics.

Collaborative filtering based on multiple attribute decision making

ABSTRACT To address the sparsity problem, a novel collaborative filtering approach based on multiple attribute decision making (MADM-CF) is proposed. In MADM-CF, users in collaborative filtering are treated as decision alternatives, items are treated as attributes. The weight of each item is determined, and the preference similarities between the active user and other users are computed. The preference similarity means that how the users’ preferences are similar on positive ratings and negative ratings. According to the preference similarities, the candidate neighbourhood of the active user is determined. A method to compute overall assessment value is designed, the overall assessment value of each user in the candidate neighbourhood is computed, and users with the smallest overall assessment values are selected as the active user’s nearest neighbours. Finally, the most frequent item recommendation method (MFIR) is used to provide top-N recommendations to the active user. Experimental results based on MovieLens and Netflix datasets show that the proposed approach is superior to existing alternatives.

Applying the Affective Aware Pseudo Association Method to Enhance the Top-N Recommendations Distribution to Users in Group Emotion Recommender Systems

Recommender Systems are a subclass of information retrieval systems, or more succinctly, a class of information filtering systems that seeks to predict how close is the match of the user’s preference to a recommended item. A common approach for making recommendations for a user group is to extend Personalized Recommender Systems’ capability. This approach gives the impression that group recommendations are retrofits of the Personalized Recommender Systems. Moreover, such an approach not taken the dynamics of group emotion and individual emotion into the consideration in making top-N recommendations. Recommending items to a group of two or more users has certainly raised unique challenges in group behaviors that influence group decision-making that researchers only partially understand. This study applies the Affective Aware Pseudo Association Method in studying group formation and dynamics in group decision making. The method shows its adaptability to group's moods change when making recommendations.

A privacy-aware deep learning framework for health recommendation system on analysis of big data

In recent technological advancement, the health recommendation system is gaining attention among the public to acquire health care services online. Traditional health recommendations are insecure due to the lack of security constraints caused by the intruders and not suitable to suggest appropriate recommendations. Thus, it creates hesitation in the minds of the people to share sensitive medical information. Hence, it is essential to design a privacy-preserving health recommendation system that should guarantee privacy and also suggest top-N recommendation to the user based on their preferences and earlier feedback. To cope with these issues, we propose a stacked discriminative de-noising convolution auto-encoder–decoder with a two-way recommendation scheme that provides secure and efficient health data to the end-users. In this scheme, privacy is assured to users through the modified blowfish algorithm. For structuring the big data collected from the patient, the Hadoop transform is used. Here, the two-way system analyzes and learns more effective features from the explicit and implicit information of the patient individually, and finally, all the learned features are fused to provide an efficient recommendation. The performance of the proposed system is analyzed with different statistical metrics and compared with recent approaches. From the result analysis, it is evident that the proposed system performs better than the earlier approaches.

Effects of Personalized and Aggregate Top-N Recommendation Lists on User Preference Ratings

Prior research has shown a robust effect of personalized product recommendations on user preference judgments for items. Specifically, the display of system-predicted preference ratings as item recommendations has been shown in multiple studies to bias users’ preference ratings after item consumption in the direction of the predicted rating. Top-N lists represent another common approach for presenting item recommendations in recommender systems. Through three controlled laboratory experiments, we show that top-N lists do not induce a discernible bias in user preference judgments. This result is robust, holding for both lists of personalized item recommendations and lists of items that are top-rated based on averages of aggregate user ratings. Adding numerical ratings to the list items does generate a bias, consistent with earlier studies. Thus, in contexts where preference biases are of concern to an online retailer or platform, top-N lists, without numerical predicted ratings, would be a promising format for displaying item recommendations.

A Course Teacher Recommendation Algorithm Based on Improved Latent Factor Model and PersonalRank

To scientifically and accurately recommend suitable teachers for university courses and improve teaching quality, designing an effective recommendation algorithm is necessary. Therefore, we construct quantitative models of teacher characteristics, course characteristics, and teaching evaluations under the theories and methods of education and build a sparse experimental data matrix based on the quantified data. On this basis, we propose a teacher recommendation algorithm (PRLFM) based on the improved latent factor model (LFM) and the improved PersonalRank algorithm. Firstly, the improved LFM is used to predict the evaluation scores of those courses that teachers have not taught. The scores which are higher than the specified threshold are used to fill the corresponding missing items in the sparse matrix to reduce the matrix's sparsity. Then, the bipartite graph model based on the teacher set and course set is constructed according to the filled experimental data matrix. The weight of edges in the bipartite graph is replaced by the teacher and course's evaluation score multiplied by the course difficulty, which reflects the correlation between course and evaluation score. Next, an improved probability transition matrix based on the bipartite graph is constructed. The access probability in the matrix is replaced by the node's out degree's reciprocal multiplied by the edge's weight. The correlation degree between the course and all teachers is quickly calculated using the matrix algorithm of PersonalRank. Finally, a teacher recommendation model is constructed to realize teachers' top-N recommendation by combining the correlation degree with teachers' characteristics. Experiments show that the PRLFM algorithm can effectively improve the accuracy of prediction and top-N recommendation. It solves the problem of lack of scientific basis in recommending suitable teachers for university courses and improving the teaching quality.

Real Estate Recommendation Approach for Solving the Item Cold-Start Problem

The item cold-start problem occurs when a recommendation system cannot recommend new items owing to record deficiencies and new listing omissions. When searching for real estate, users can register a concurrent interest in recent and prior projects. Thus, an approach to recommend cold-start and warm-start items simultaneously must be determined. Furthermore, unrequired membership and stop-by behavior cause real estate recommendations to have many cold-start and new users. This characteristic encourages the use of a content-based approach and a session-based recommendation system. Herein, we propose a real estate recommendation approach for solving the item cold-start problem with acceptable warm-start item recommendations in the many-cold-start-users scenario. We modify a session-based recommendation system and employ existing mechanisms to efficiently deal with sequential and context information for the next-interacted item’s encoded attribute prediction. Subsequently, we use the nearest-neighbors approach using weighted cosine similarity to determine conforming candidates. We use Recall@K and MRR@K with the top-n recommendation to evaluate warm-start and cold-start item recommendations among different applied mechanisms and against the baselines. The results demonstrate the effectiveness of efficiently integrating the information and the difficulty in performing well in warm-start and cold-start item recommendations simultaneously. Our proposed approach illustrates the capability of solving the item cold-start problem while yielding promising results in both recommendations although neither result is the best. We believe that our approach provides a suitable compromise between both recommendations and that it will benefit recommendation tasks focusing on both recommendations.

Speed up Differential Evolution for ranking of items in recommendation systems

Recommendation systems can suggest users list of items they have not yet seen but might be interested in. To improve the quality of the generated recommendations, different techniques are often used which try to personalize recommendations. Usually user preferences are stored in the form of a vector in which individual values describe to what extent a given feature is desired by the user. To find this vector, metaheuristic algorithms can be used, however their main drawback is their computational complexity. Therefore, in this paper, a modification of the Differential Evolution algorithm is proposed to enable faster computation of the ranking score for each item in the system, which is used to create a recommendation list. Experiments have been performed on the current MovieLens 25m database and they show that our modification can significantly speed up the process of finding a preference vector, without losing their quality for the top-N recommendation task. We will also address the vulnerability of recommendation systems to profile injection attacks, as a result of which an attacker can influence the generated recommendations.

APPLYING THE AFFECTIVE AWARE PSEUDO ASSOCIATION METHOD TO ENHANCE THE TOP-N RECOMMENDATIONS DISTRIBUTION TO USERS IN GROUP EMOTION RECOMMENDER SYSTEMS

Recommender Systems are a subclass of information retrieval systems, or more succinctly, a class of information filtering systems that seeks to predict how close is the match of the user’s preference to a recommended item. A common approach for making recommendations for a user group is to extend Personalized Recommender Systems’ capability. This approach gives the impression that group recommendations are retrofits of the Personalized Recommender Systems. Moreover, such an approach not taken the dynamics of group emotion and individual emotion into the consideration in making top-N recommendations. Recommending items to a group of two or more users has certainly raised unique challenges in group behaviors that influence group decision-making that researchers only partially understand. This study applies the Affective Aware Pseudo Association Method in studying group formation and dynamics in group decision making. The method shows its adaptability to group's moods change when making recommendations.

Content boosted hybrid filtering for solving pessimistic user problem in recommendation systems

The main goal of a recommendation system is to recommend items of interest to users by analyzing their historical data. Content-based and collaborative filtering are the traditional recommendation strategies, each with its own strengths and weaknesses. Some of their weaknesses can be overcome by combining the two strategies. The resulting hybrid system performs qualitatively better than the traditional recommendation systems. However, historical data of some users may consist largely of only likes or only dislikes. Those users are termed as optimistic or pessimistic users respectively. On an average there are around 10 to 20% of pessimistic users present in a given dataset. For pessimistic users, whose profiles have mostly dislikes and very few likes, content-based filtering can hardly recommend any items of interest. In content-based filtering technique pessimistic users get poor recommendations of either uninteresting movies or no recommendations at all. This can be alleviated by boosting the content profiles of pessimistic users using the top-n recommendations of collaborative filtering. This content boosted hybrid filtering system provides a novel list of recommendations even for pessimistic users, with predictive accuracy better than that of a traditional content-based filtering system.

Transfer to Rank for Top-N Recommendation

In this paper, we study top-N recommendation by exploiting users’ explicit feedback such as 5-star numerical ratings, which has been overlooked to some extent in the past decade. As a response, we design a novel and generic transfer learning based recommendation framework coarse-to-fine transfer to rank (CoFiToR), which is a significant extension of a very recent work called transfer to rank (ToR). The key idea of our solution is modeling users’ behaviors by simulating users’ shopping processes. Therefore, we convert the studied ranking problem to three subtasks corresponding to three specific questions, including (i) whether an item will be examined by a user, (ii) how an item will be scored by a user, and (iii) whether an item will finally be purchased by a user. Based on this new conversion, we then develop a three-staged solution that progressively models users’ preferences from a coarse granularity to a fine granularity. In each stage, we adopt an appropriate recommendation algorithm with pointwise or pairwise preference assumption to answer each question in order to seek an effective and efficient overall solution. Empirical studies on two large and public datasets showcase the merits of our solution in comparison with the state-of-the-art methods.

Hybrid crow search and uniform crossover algorithm-based clustering for top-N recommendation system

Recommender systems (RSs) have gained immense popularity due to their capability of dealing with a huge amount of information available in various domains. They are considered to be information filtering systems that make predictions or recommendations to users based on their interests. One of the most common recommender system techniques is user-based collaborative filtering. In this paper, we follow this technique by proposing a new algorithm which is called hybrid crow search and uniform crossover algorithm (HCSUC) to find a set of feasible clusters of similar users to enhance the recommendation process. Invoking the genetic uniform crossover operator in the standard crow search algorithm can increase the diversity of the search and help the algorithm to escape from trapping in local minima. The top-N recommendations are presented for the corresponding user according to the most feasible cluster’s members. The performance of the HCSUC algorithm is evaluated using the Jester dataset. A set of experiments have been conducted to validate the solution quality and accuracy of the HCSUC algorithm against the standard particle swarm optimization (PSO), African buffalo optimization (ABO), and the crow search algorithm (CSA). In addition, the proposed algorithm and the other meta-heuristic algorithms are compared against the collaborative filtering recommendation technique (CF). The results indicate that the HCSUC algorithm has obtained superior results in terms of mean absolute error, root means square errors and in minimization of the objective function.

Attribute-aware deep attentive recommendation

Since the rich semantics of attribute information has become a great supplement to the ratings data in designing recommender systems, fusing attributes information into ratings has shown promising performance for many recommendation tasks. However, the use of attribute information is not easy, because different attributes are often: (1) multi-source, that is, attributes may come from many different fields, (2) unstructured, (3) unbalanced, (4) heterogeneous. In this paper, we explore effective fusion of user-item ratings and item attributes to improve recommendations, we propose an attribute-aware deep attentive recommendation model, which embeds attribute information into the latent semantic space of items through the attention mechanism, forming more accurate item representations. Extensive experiments show that our method is superior to the existing methods on both rating prediction and Top-N Recommendation tasks.

Dynamic clustering collaborative filtering recommendation algorithm based on double-layer network

With the rapid development of internet economy, personal recommender system plays an increasingly important role in e-commerce. In order to improve the quality of recommendation, a variety of scholars and engineers devoted themselves in developing the recommendation algorithms. Traditional collaborative filtering algorithms are only dependent on rating information or attribute information. Most of them were considered in perspective of a single-layer network, which destroyed the original hierarchy of data and resulted in sparse matrix and poor timeliness. In order to address these problems and improve the accuracy of recommendation, dynamic clustering collaborative filtering recommendation algorithm based on double-layer network is put forward in this paper. Firstly, attribute information of users and items are respectively used to construct the user layer network and the item layer network. Secondly, new hierarchical clustering method is further presented, which separates users into different communities according to dynamic evolutionary clustering. Finally, score prediction and top-N recommendation lists are obtained by similarity between users in each community. Extensive experiments are conducted with three real datasets, and the effectiveness of our algorithm is verified by different metrics.