HERS: Modeling Influential Contexts with Heterogeneous Relations for Sparse and Cold-Start Recommendation

Taxonomies are ubiquitous in many real-world recommendation scenarios where each item is classified into a category of a predefined hierarchical taxonomy and provide important auxiliary information for inferring user preferences. However, traditional collaborative filtering approaches have focused on user-item interactions (e.g., ratings) and neglected the impact of taxonomy information on recommendation. In this paper, we present a taxonomy-aware denoising autoencoder based model which incorporates taxonomy-aware side information into denoising autoencoder based recommendation models to enhance recommendation accuracy and alleviate data sparsity and cold start problems in recommendation systems. We propose two types of taxonomic side information, namely the topological representation of tree-structured taxonomy and the statistical properties of the taxonomy. By integrating taxonomic side information, our model can learn more effective user latent vectors which are not only determined by user ratings but also rely on the taxonomy information. We conduct a comprehensive set of experiments on two real-world datasets which provide several outcomes: first, our proposed taxonomy-aware method outperforms the baseline method on RMSE metric. Next, information extracted from taxonomy can help alleviate data sparsity and cold start problems. Moreover, we conduct supplementary experiments to explore the reason why our proposed taxonomic side information improves recommendation performance.

Cross Domain Recommendation Systems using Deep Learning: A Systematic Literature Review

Social recommendation with Graph Neural Networks(GNNs) learns to represent cold users by fusing user-user social relations with user-item interactions, thereby alleviating the cold-start problem associated with recommender systems. Despite being well adapted to social relations and user-item interactions, these supervised models are still susceptible to popularity bias. Contrastive learning helps resolve this dilemma by identifying the properties that distinguish positive from negative samples. In its previous combinations with recommender systems, social relationships and cold-start cases in this context are not considered. Also, they primarily focus on collaborative features between users and items, leaving the similarity between items under-utilized. In this work, we propose socially-aware dual contrastive learning for cold-start recommendation, where cold users can be modeled in the same way as warm users. To take full advantage of social relations, we create dynamic node embeddings for each user by aggregating information from different neighbors according to each different query item, in the form of user-item pairs. We further design a dual-branch self-supervised contrastive objective to account for user-item collaborative features and item-item mutual information, respectively. On one hand, our framework eliminates popularity bias with proper negative sampling in contrastive learning, without extra ground-truth supervision. On the other hand, we extend previous contrastive learning methods to provide a solution to cold-start problem with social relations included. Extensive experiments on two real-world social recommendation datasets demonstrate its effectiveness.

Meta-learning has been proven to be effective for the cold-start problem of recommender systems. Many meta-learning recommender systems that are designed for the user cold-start problem are gradient-based. They use a global parameter learned from existing users to initialize the recommender system parameter for new users that provides a personalized recommendation with limited user-item interactions. Many systems require users’ demographic information to learn the global parameter. This requirement raises privacy concerns, and demographic information is not always available. In addition, some of the gradient-based systems need dozens or even hundreds of user-item interactions from existing users to learn the global parameter. This requirement is difficult to satisfy in specific scenarios in which active users and their item interactions are rare. Moreover, gradient-based meta-learning systems rarely capture user preferences over different item attributes, as updating the corresponding weights requires many optimization steps. We proposed HyperRS, a Hypernetwork-based Recommender System, for the user cold-start problem. Our system does not rely on demographic information to provide personalized recommendations. In our system, a hypernetwork generates all weights in the underlying recommender system. The hypernetwork enables the weights to adapt quickly to capture user interest in both item attributes and the contents of the item attributes. The experimental results show that our method outperforms several state-of-the-art meta-learning recommender systems for the user cold-start problem.

In the current era of online information overload, recommendation systems are very useful for helping users locate content that may be of interest to them. A personalized recommendation system presents content based on information such as a user’s browsing history and the videos watched. However, information filtering-based recommendation systems are vulnerable to data sparsity and cold-start problems. Additionally, existing recommendation systems suffer from the large overhead incurred in learning regression models used for preference prediction or in selecting groups of similar users. In this study, we propose a preference-tree-based real-time recommendation system that uses various tree models to predict user preferences with a fast runtime. The proposed system predicts preferences based on two balance constants and one similarity threshold to recommend content with a high accuracy while balancing generalized and personalized preferences. The results of comparative experiments and ablation studies confirm that the proposed system can accurately recommend content to users. Specifically, we confirmed that the accuracy and novelty of the recommended content were, respectively, improved by 12.1% and 27.2% compared to existing systems. Furthermore, we verified that the proposed system satisfies real-time requirements and mitigates both cold-start and overfitting problems.

Factorization Machines (FMs) are effective in incorporating side information to overcome the cold-start and data sparsity problems in recommender systems. Traditional FMs adopt the inner product to model the second-order interactions between different attributes, which are represented via feature vectors. The problem is that the inner product violates the triangle inequality property of feature vectors. As a result, it cannot well capture fine-grained attribute interactions, resulting in sub-optimal performance. Recently, the euclidean distance is exploited in FMs to replace the inner product and has delivered better performance. However, previous FM methods including the ones equipped with the euclidean distance all focus on the attribute-level interaction modeling, ignoring the critical intrinsic feature correlations inside attributes. Thereby, they fail to model the complex and rich interactions exhibited in the real-world data. To tackle this problem, in this paper, we propose a FM framework equipped with generalized metric learning techniques to better capture these feature correlations. In particular, based on this framework, we present a Mahalanobis distance and a deep neural network (DNN) methods, which can effectively model the linear and non-linear correlations between features, respectively. Besides, we design an efficient approach for simplifying the model functions. Experiments on several benchmark datasets demonstrate that our proposed framework outperforms several state-of-the-art baselines by a large margin. Moreover, we collect a new large-scale dataset on second-hand trading to justify the effectiveness of our method over cold-start and data sparsity problems in recommender systems.

Social recommendation leverages social information to solve data sparsity and cold-start problems in traditional collaborative filtering methods. However, most existing models assume that social effects from friend users are static and under the forms of constant weights or fixed constraints. To relax this strong assumption, in this paper, we propose dual graph attention networks to collaboratively learn representations for two-fold social effects, where one is modeled by a user-specific attention weight and the other is modeled by a dynamic and context-aware attention weight. We also extend the social effects in user domain to item domain, so that information from related items can be leveraged to further alleviate the data sparsity problem. Furthermore, considering that different social effects in two domains could interact with each other and jointly influence user preferences for items, we propose a new policy-based fusion strategy based on contextual multi-armed bandit to weigh interactions of various social effects. Experiments on one benchmark dataset and a commercial dataset verify the efficacy of the key components in our model. The results show that our model achieves great improvement for recommendation accuracy compared with other state-of-the-art social recommendation methods.

The traditional matrix factorization algorithm has the problems of cold start, data sparsity and high predictive time complexity, which is not perfect in most recommendation systems. Especially in the sparse matrix problem, the recommendation accuracy cannot be guaranteed. This paper will be based on the matrix factorization algorithm modeling and create optimization algorithm, this method will be graded dataset preprocessing, the user and item's score matrix by embedding process at the same time and introducing BatchNorm sparse matrix algorithm for training normalization processing parameters do to speed up the convergence speed, increase the stability of the training, respectively add item bias that more able to show the user's true score, then the user matrix to join implicit feedback, the final score projections for the user. The optimization algorithm performs well in solving cold startup and data sparse problems of matrix factorization. The Results of the Movielens-1M experiment show that it has great advantages over traditional matrix factorization algorithms in terms of prediction accuracy, root mean square error and square absolute error.

Matrix Factorization (MF) is a well-known method which is to predict users’ preference in recommender systems. This method has the advantage of reducing data sparsity and cold start problem. Probabilistic matrix factorization (PMF) is an improved method of matrix factorization. Recommender system can efficiently reduce the information overload problem. However, it is bound by the recommended accuracy. In this paper we introduce Bayesian distribution to improve accuracy and design our novel probabilistic matrix factorization (NPMF) proved to have better performance. In addition, we also use K nearest neighbor method and stochastic gradient descent to further improve our method. The experimental studies using two different datasets including MovieLens 100k and MovieLens 1M, show that the proposed methods already have comparable and better performance.

A many objective commercial recommendation algorithm via Game-Based core node extraction

The main task of recommender systems is to recommend items or any other information which the particular user is interested in accurately and efficiently. However, traditional recommender systems usually suffer from the data sparsity problem and cold start problem. To alleviate these problems, social recommender systems usually combine the social network data and rating data together to improve the accuracy of prediction results. In this paper, we propose SimilarMF, a novel and accurate social recommender system trying to improve the performance by means of utilizing the implicit similarity information among users and items. Firstly, SimilarMF integrates social network with rating data to construct a user similarity graph and an item similarity graph. Secondly, it adopts an embedding method to acquire enough similar user-user pairs and item-item pairs. It also performs negative sampling for dissimilar pairs. Lastly, it incorporates the extracted information into matrix factorization framework to predict ratings more accurately. The experimental results show that our method offers better recommendations than its competitors.

To solve the information explosion problem and enhance user experience in various online applications, recommender systems have been developed to model users’ preferences. Although numerous efforts have been made toward more personalized recommendations, recommender systems still suffer from several challenges, such as data sparsity and cold-start problems. In recent years, generating recommendations with the knowledge graph as side information has attracted considerable interest. Such an approach can not only alleviate the above mentioned issues for a more accurate recommendation, but also provide explanations for recommended items. In this paper, we conduct a systematical survey of knowledge graph-based recommender systems. We collect recently published papers in this field, and group them into three categories, i.e., embedding-based methods, connection-based methods, and propagation-based methods. Also, we further subdivide each category according to the characteristics of these approaches. Moreover, we investigate the proposed algorithms by focusing on how the papers utilize the knowledge graph for accurate and explainable recommendation. Finally, we propose several potential research directions in this field.

Recommendation systems (RSs) are tools for interacting with large and complex information spaces. They provide a personalized view of such spaces, prioritizing items likely to be of interest to the user. The main objective of RSs is to tool up users with desired items that meet their preferences. A major problem in RSs is called: “cold-start”; it is a potential problem called so in computer-based information systems which comprises a degree of automated data modeling. Particularly, it concerns the issue in which the system cannot draw any inferences nor have it yet gathered sufficient information about users or items. Since RSs performance is substantially limited by cold-start users and cold-start items problems; this research study takes the route for a major aim to attenuate users’ cold-start problem. Still in the process of researching, sundry studies have been conducted to tackle this issue by using clustering techniques to group users according to their social relations, their ratings or both. However, a clustering technique disregards a variety of users’ tastes. In this case, the researcher has adopted the overlapping technique as a tool to deal with the clustering technique’s defects. The advantage of the overlapping technique excels over others by allowing users to belong to multi-clusters at the same time according to their behavior in the social network and ratings feedback. On that account, a novel overlapping method is presented and applied. This latter is executed by using the partitioning around medoids (PAM) algorithm to implement the clustering, which is achieved by means of exploiting social relations and confidence values. After acquiring users’ clusters, the average distances are computed in each cluster. Thereafter, a content comparison is made regarding the distances between every user and the computed distances of the clusters. If the comparison result is less than or equal to the average distance of a cluster, a new user is added to this cluster. The singular value decomposition plus (SVD[Formula: see text]) method is then applied to every cluster to compute predictions values. The outcome is calculated by computing the average of mean absolute error (MAE) and root mean square error (RMSE) for every cluster. The model is tested by two real world datasets: Ciao and FilmTrust. Ultimately, findings have exhibited a great deal of insights on how the proposed model outperformed a number of the state-of-the-art studies in terms of prediction accuracy.

Structured, semi-structured and unstructured data is growing exponentially. The exponential increase in the data resulted in providing a plethora of alternatives for items, services and products. In e-commerce, recommender systems have become the most powerful tool to avoid choice overload. People who lack effective proficiency to evaluate the various alternatives can use Recommender systems. Poor decision making is due to the exponential growth of information and the convention of new e-business services like comparison of products, auction, purchase of products etc. This paper outline the key concepts used in recommendation systems and elucidate different recommendation approaches using content based, collaborative and hybrid methods. The paper reviews the drawbacks of various recommendation methods and the possible future scope to enhance the performance of recommendation systems. Several approaches are addressed in this paper to deal with the common cold start and data sparsity problems. The paper also proposes a hybrid recommendation system, which provides a recommendation based on the score for each scenario.

In the context of Big data, large-scale open online courses increase learning paths for learners, but in the face of countless high-quality curriculum resources, it is easy for derivative learners to face the dilemma of rich curriculum resources but difficult to choose resources, which leads to information maze for learners. How to help learners quickly and accurately find their own learning resources in the explosive growth of MOOC resources is an urgent problem in the field of education Big data. However, the traditional Collaborative filtering recommendation technology does not perform well when dealing with sparse data and cold start. The recommendation content is repeated and can not effectively deal with high-dimensional and nonlinear data of online learning users, resulting in low efficiency of resource recommendation. Therefore, the study adopts a deep belief network (DBN) to construct a personalized resource recommendation model. The model combines the learner behavior characteristics with the curriculum resource content attribute characteristics to form the learner feature vector. The parameters of the model are adjusted according to the characteristics of learners. Through experiments, the proposed model has shown good performance. The experiment explored the effects of training set size, learner characteristics, and GPU on model performance. The experimental results show that when the training set proportion is 100%, the RMSE, Accuracy, Recall, and F1 values of the model are 0.76, 0.946, 0.957, and 0.951, respectively. When the model is trained using a training set containing learner features, the RMSE, Accuracy, Recall, and F1 values of the model are 0.75, 0.962, 0.908, and 0.958, respectively. After using GPU to accelerate the model, the running time of the model decreased from 360 minutes to 90 minutes. The results indicate that the model cannot effectively mine data information when the degree of correlation between sample information is low. The richer the relationships between samples, the better the performance of the model. Simultaneously learning hunger feature vectors and learner behavior feature vectors for training can significantly improve the recommendation accuracy of the model. The main contribution of this study is to propose a recommendation method based on DBN classification to replace traditional similarity calculation methods, using DBN's efficient feature abstraction and feature extraction capabilities to fully explore learners' interest and preference for course resources. In addition, in view of the common problems of cold start and data sparsity in traditional Collaborative filtering recommendation methods, the research deeply mines the characteristics of learners' Demographics and curriculum resources' content attributes, and constructs a learner interest model based on DBN combined with learners' behavior characteristics, which effectively solves the problems of cold start and data sparsity, as well as the inaccurate expression of learners' interest preferences for curriculum resources.