Search

With the rapid development of science and technology, especially the rapid progress of the Internet, the e-commerce industry has ushered in unprecedented development opportunities. However, existing e-commerce platforms still have many problems in terms of user experience and functionality, such as the lack of accuracy in the recommendation system, inconsistency in product descriptions, and lack of reliability in user reviews. In this paper, we design and implement an online shopping mall system based on Spring Boot and Vue framework, aiming to improve user experience through advanced information technology and optimization algorithms. The system is developed in Java language, and the architecture is divided into Filter/Interceptor Layer, Controller Layer, Service Layer, and Dao Layer, which ensures the efficiency and maintainability of the system. The introduction of collaborative filtering algorithms significantly improves the accuracy of product recommendations, provides personalized services, and improves the user shopping experience. Meanwhile, the dynamically updated sensitive word filtering mechanism safeguards the healthy environment of the comment area and improves user trust. The experimental results show that the system achieves the expected goals in terms of functionality and performance, which provides a reference and reference for the development of future e-commerce platforms. [ABSTRACT FROM AUTHOR]

The primary goal of recommender systems is to identify and propose items that users might find appealing. A large number of these systems are heavily dependent on explicit interactions between the user and the item, which can often be infrequent. In this work, we introduce a unique model known as Context-Aware Metric Learning (CAML), designed to enhance the effectiveness of recommendations. The CAML model utilizes an attentive autoencoder to extract latent features from contextual context and incorporates these features into a metric learning framework. In particular, these extracted features act as a Gaussian prior for the embeddings of the items, thereby enhancing the precision of their positioning in the latent space. This integration not only boosts the precision of the recommendations but also increases computational efficiency, rendering CAML appropriate for both offline and online application scenarios. Our model's evaluation on two real-world datasets reveals that it outperforms several existing baseline models, including those that do not incorporate contextual information such as CML and CPE, as well as other contextual recommendation models like CDL, CATA, and CML+F. [ABSTRACT FROM AUTHOR]

Watching movies has been our source of entertainment for decades. In this modern world, watching movies has never been more convenient with the existence of video streaming services. This convenience also comes with new problems because there are a lot of movies and series on the platform that exploring it on our own will take time and effort. To solve that problem, companies use recommendation systems to improve user experience. In the field of recommendation systems, there are two famous approaches, content-based filtering, and collaborative filtering. This research aims to compare both methods and find the best possible method to use in a video streaming service platform. We found that the difference between the two is located on the data that they need for the model to work properly. Content-based recommendation systems have higher accuracy when using a limited amount of data while collaborative filtering needs more data about user's behavior. Other than that, both methods give different recommendations because content-based only gives recommendations that are like the content that the user likes while collaborative filtering recommends contents that other users like. [ABSTRACT FROM AUTHOR]

Travelers usually check information about the destination before they decide to go. However, sometimes the information is too much to handle and causes confusion. A filtering mechanism is needed to help them make their choices. The solution is to implement a recommendation system; some common techniques that are used for this are Collaborative Filtering and content- based filtering, but both techniques have their flaws. Therefore, we propose a solution to combine those two methods through Graph Neural Network architectures to help fix their flaws. The proposed method will be used for recommending hotels in Indonesia using the HotelRec dataset as the base dataset and hotel features acquired from the TripAdvisor API. We use several GNN architectures, such as GCN, LightGCN, GraphSAGE, and GAT, to implement our approach. Then, we compare it to several baseline models, such as SVD and NeuMF, which are common baselines in recommendation systems. The performance of these models is calculated based on three metrics: HitRate@K, NDCG@K, and Recall@K. Based on the result of this experiment, we found that incorporating hotel features into the learning process of the model, accompanied by user-item interactions, boosted the model's performance by at least 20% in all metrics, especially in datasets with a smaller number of interactions. [ABSTRACT FROM AUTHOR]

Matrix factorization (MF) is a simple collaborative filtering technique that achieves superior recommendation accuracy by decomposing the user–item interaction matrix into user and item latent matrices. Because the model typically learns each interaction independently, it may overlook the underlying shared dependencies between users and items, resulting in less stable and interpretable recommendations. Based on these insights, we propose "Hierarchical Matrix Factorization" (HMF), which incorporates clustering concepts to capture the hierarchy, where leaf nodes and other nodes correspond to users/items and clusters, respectively. Central to our approach, called hierarchical embeddings, is the additional decomposition of the latent matrices (embeddings) into probabilistic connection matrices, which link the hierarchy, and a root cluster latent matrix. The embeddings are differentiable, allowing simultaneous learning of interactions and clustering using a single gradient descent method. Furthermore, the obtained cluster-specific interactions naturally summarize user–item interactions and provide interpretability. Experimental results on ratings and ranking predictions show that HMF outperforms existing MF methods, in particular achieving a 1.37 point improvement in RMSE for sparse interactions. Additionally, it was confirmed that the clustering integration of HMF has the potential for faster learning convergence and mitigation of overfitting compared to MF, and also provides interpretability through a cluster-centered case study. [Display omitted] • Recommender model to capture user/item hierarchies simultaneously in factorization. • Hierarchical embeddings are optimized with a single gradient descent method. • Obtained clusters capable of summarizing of interactions and providing interpretation. • Possibility of being robust to sparse interactions due to its hierarchical structure. • Presentation of a case study summarizing preferences on a cluster-by-cluster basis. [ABSTRACT FROM AUTHOR]

To give users personalised recommendations, the collaborative filtering technique is widely used. In this technique, similarity computation plays a crucial role. Once the similarities among users or items are determined, the system can predict how a user might rate or interact with items that they have not yet encountered. The traditional approaches or measures mainly considered the user-item historical ratings to compute the similarity, while user preferences may change with time. Considering this, the objective of this research is to create an efficient recommendation system that utilized the temporal data. For this, few time decay functions, i.e., exponential, linear, logistic, and power applied to the ratings to give more weightage to the most recent ratings. Since, Gower's coefficient is suitable for handling the missing data, it is applied to calculate the similarity and compared its results with popular traditional similarity measures i.e., Cosine and Pearson correlation coefficient. Experimental findings on the ML-100k dataset in terms of Root Mean Squared Error (RMSE) and Mean Average Error (MAE) demonstrate that performance of RS improved when we applied a power function on ratings. With comparison to most recent methods PIP, NHSM, RJaccard etc., proposed approach gives almost 10% better results in MAE and 5% better results in RMSE comparison. [ABSTRACT FROM AUTHOR]

In recent times, recommender systems have acquired significant popularity as a solution to the issue of information overload. These systems offer personalised recommendations to users. The superiority of multi-criteria recommender systems over their single-criterion counterparts has been demonstrated, as the former are able to provide more precise recommendations by taking into account multiple dimensions of user preferences when rating items. The prevalent recommendation technique, matrix factorization of collaborative filtering, is hindered by the data sparsity problem of the user-item matrix. On the other hand, it is noteworthy that deep learning techniques have demonstrated significant potential in various research domains, including but not limited to image processing, pattern recognition, computer vision, and natural language processing. In recent times, there has been a surge in the utilisation of deep learning techniques in recommender systems, yielding promising outcomes. This study presents a novel approach to multi-criteria recommender systems through the utilisation of deep learning algorithms to mitigate the data sparsity issue. Specifically, deep autoencoders are utilised to uncover complex, non-linear, and latent relationships between users' multi-criteria preferences followed by matrix factorization technique, ultimately leading to more precise recommendations. The proposed model is evaluated by conducting the experiments on the multi-criteria dataset of Yahoo! Movies. According to the outcomes, the proposed approach outperforms the state of the art recommendation methods by generating more accurate and personalized recommendations. Also, it reduces the data sparsity up to 11% from the multi-criteria dataset. [ABSTRACT FROM AUTHOR]

A filtering method is indispensable in a data-flooded environment. Recommended systems have made a massive step towards this aim, speeding up internet-based customer experience. Most of today's examples of artificial marketing intelligence are known as supervised learning, which varies from offering personalized specific products identifying the most valuable marketing strategies, to forecasting customer churn rate or customer life value, and building up a positive client base. Generally, different types of stored information are used to customize various dimensions or search results, demonstrate the most targeted advertising on the homepage, etc. n our research we have also incorporated the benefits and drawbacks of every approach. Finally, this paper also presents numerous difficulties and problems confronting recommenders in their application systems algorithms. In this paper, we initially present multiple best-known types of recommended systems and concentrate on one part of the e- commerce recommendation and afterwards make their quantitative comparison. Recommender systems have taken a huge step towards this goal, greatly improving the user experience in the online environment. [ABSTRACT FROM AUTHOR]

The performance of recommendation engines is challenged by the cold-start issues. Classical recommendation techniques have limited capability to address this problem because the underlying machine learning models are data-thirsty. Conversely, the recent success of meta-learning attracted a lot of attention due to few-shot learning capabilities which can be employed to accommodate new tasks even when there are few data samples for training. Most of the meta-learning recommendation models adopt model-agnostic meta-learning to initialize parameters that may lead to stuck into local optima instead of global optima for some users. To leverage the learning process, we propose a Contextually Augmented Meta-Learning recommender system (CAML). The proposed method augments the contextual features into a meta-learning model which considerably improves the tasks adaptation capability. We constructed a Data Augmentation Unit (DAU) that used a hybrid similarity to augment data samples of similar neighbors. The augmented samples are then forwarded to a Meta-Learner (MetaL) to learn user preferences and generate relevant recommendations. Additionally, we highlighted that the proposed method is generalized and can be adapted for various meta-learning-based recommendation models. We validate our proposal on three benchmark datasets for both ranking and rating prediction perspectives. The experimental outcomes prove the significance of our proposal over the state-of-the-art recommendation methods. [ABSTRACT FROM AUTHOR]

The Collaborative Filtering (CF) technique is the most common neighbourhood-based recommendation strategy, that provides personalized recommendation to a user for the items using a similarity measure. Hence, the selection of the appropriate similarity measure becomes crucial in the CF based recommendation system. The traditional similarity measures merely focus only on the historical ratings provided by the users to compute the similarity, completely ignoring the fact that preferences change over a period of time. Considering this, the paper aims to develop an effective Recommendation System that uses temporal information to capture the changes in the preferences over a period of time. For this, the existing exponential and power time decay functions are integrated with Cosine, Pearson Correlation, and Gower's similarity measures to compute similarity. The similarity is computed at the similarity computation and prediction levels of recommendation processes. Experimental findings in terms of MAE and RMSE on the MovieLens-100k demonstrate that performance of Gower's coefficient is better when applied with the exponential function at the similarity computation level of the recommendation process. [ABSTRACT FROM AUTHOR]

Nowadays, recommender systems play a vital role in every human being's life due to the time retrieving the items. The matrix factorization (MF) technique is one of the main methods among collaborative filtering (CF) techniques that have been widely used after the Netflix competition. Traditional MF techniques are static in nature. However, the perception and popularity of products are constantly changing with time. Similarly, the users' tastes are changed with time. Hence, traditional MF cannot handle the dynamic effect of the user-item interaction. To tackle the temporal and dynamic effect of user-item interaction, we proposed a collaborative filtering model for movie recommendations that include temporal effects. To justify the significance of the proposed technique, we evaluated our model on a standard dataset (Movielens) and compared it with state-of-art models. The exploratory outcomes signify that the proposed technique obtains a better result than a state-of-art model with an improvement of 1.35% and 1.28% on ML-100K and 1M datasets, respectively. [ABSTRACT FROM AUTHOR]

Recommender System is a information filtering tool to filter the relevant information from given information in the present era of big data. Movie Recommender System is machine learning based autonomous tool that filters the movies from big movie database like netflix, amazon etc according to user preferences. The main focus of this paper is Partitional Weighted co-clustering for Movie Recommender System. The primary objective of this research article is to fine tune the parameters of user and movie neighborhoods by setting different values for row clusters number and column clutsers number parameters of co-clustering. Test results obtained from the Movie database show that the proposed method can bring more accurate personalized recommendations for the movie as compared to existing methods of the order of 7.91%. [ABSTRACT FROM AUTHOR]

Existing works in recommender system have widely explored extracting reviews as explanations beyond user–item interactions, and formulated the explanation generation as a ranking task to enhance item recommendation performance. To associate explanations with users and items, graph neural networks (GNN) are usually employed to learn node representations on the heterogeneous user–item–explanation interaction graph. However, modeling heterogeneous graph convolution poses limitations in both message passing styles and computational efficiency, resulting in sub-optimal recommendation performance. To address the limitations, we propose an Explanation-aware Graph Convolution Network (ExpGCN). In particular, the heterogeneous interaction graph is divided to subgraphs regard to the edge types in ExpGCN. By aggregating information from distinct subgraphs, ExpGCN is capable of generating node representations for explanation ranking task and item recommendation task respectively. Task-oriented graph convolution can not only reduce the complexity of heterogeneous node aggregation, but also alleviate the performance degeneration caused by the conflicts between task learning objectives, which has been neglected in current studies. Extensive experiments on four public datasets show that ExpGCN significantly outperforms state-of-the-art baselines with high efficiency, demonstrating the effectiveness of ExpGCN in explainable recommendations. [ABSTRACT FROM AUTHOR]

Typical targets in SAR observing scenes are characterized by anisotropy, which represents backscattering information changing with the observing angle. Anisotropic scattering can be used to study the structural characteristics of targets and the response differences between structures. In actual situations, due to flight conditions and other reasons, only partial observation information can be obtained, so it is important to use limited data to deduce remaining backscattering information for target analysis. In our article, the scattering deduction model based on matrix completion is established, including scattering characteristic analysis. The multi-aspect data is transferred from range and azimuth in conventional SAR image to pixel dimension and angle dimension for analysis. At the same time, the pixel similarity matrix is used as a feature in pixel dimension, and one-dimension back-amplitude scattering curve with angle is established as a feature matrix in angle dimension. In the process of modeling existing data, two corresponding decomposition matrices are generated with the help of collaborative filtering idea. Finally, a four-matrix combination for anisotropy deduction is formed. Ku-band multi-aspect data from autonomous experimental flights is used for algorithm. In the experiment, the scattering deduction of the remaining aspect sequence was carried out based on obtaining partial aspect sequence images, and the strong scattering regions of the target point and sub-aperture details were compared. The edge information and structure map of the target were reconstructed using scattering deduction results. [ABSTRACT FROM AUTHOR]

Food Recommendation System (FRS) assists individuals in making healthier dietary choices. However, current FRS uses collaborative filtering algorithms for one-step recommendations. Although these systems can recommend foods based on users' historical preferences, they lack the adaptability to real-time changes in users' health requirements and, as a result, the dynamic adjustment of recommendation strategies. This study introduces a groundbreaking approach by incorporating the dynamic and adaptive nature of reinforcement learning algorithms (RL) into FRS. The proposed multi-step recommendation framework, RecipeRL, leverages RL's continuous decision-making and sustained interaction capabilities. To more accurately recommend foods aligned with user preferences, we introduce an effective method for expressing users' real-time state through fused state representation. We also introduce an interactive environment to simulate authentic interactions between users and the recommendation system, enabling the system to handle multi-step recommendations. Our approach was evaluated using publicly available real-world datasets and compared to ten state-of-the-art methods. The results of the Top@10 analysis show that our method outperforms other algorithms significantly, achieving 94.68% and 95.67% for traditional Precision and the recommendation system metric NDCG, respectively. Our method also exhibits adaptability in scenarios where user preferences change, achieving 93.2% and 95.71%, respectively. [ABSTRACT FROM AUTHOR]

Well log data provide key subsurface information, which is crucial for lithology evaluation and reservoir characterization. However, due to technical issues, well log data may contain missing values at certain depth intervals, which can be detrimental for data analysis. The best method is to reacquire the missing data by relogging, but this increases operational costs. Thus, a cost-efficient method for restoring the lost data is needed to overcome this issue. We propose an imputation method for missing well log data using collaborative filtering, a widely used algorithm for making new item recommendations to users. Although collaborative filtering is mainly used in recommendation systems, its fundamental principle allows us to utilize it to help make predictions for missing log data. The method is applied to a well log dataset obtained from the North Sea near Norway. The results show that the collaborative filtering algorithm has the potential to be a powerful imputation method for missing well log data, but there are some limitations that need to be addressed. • K-nearest neighbor collaborative filtering is used as an imputation method for missing well log data. • RMSE is used as a rough estimate to find the optimum k value. • Many limitations need to be overcome to utilize the full potential of collaborative filtering. [ABSTRACT FROM AUTHOR]

Collaborative filtering (CF) algorithms provide personalized recommendations based on user preferences and they are widely applied in various domains including social media and video platforms. Recently, the broad learning system (BLS) has been incorporated into CF. The combination of BLS and CF produces more accurate recommendation results due to the fact it can efficiently learn the non-linear relationship between items and users. However, the BLS-based CF adopts cosine similarity to seek several nearest neighbors when constructing features. This feature construction process generally leads to the difference of user rating scales and thus affects the calculation accuracy of the nearest neighbors. Furthermore, like many other CF algorithms, the BLS-based version cannot effectively handle with the data sparsity problem. The above two issues frequently result in poor recommendation performance for the BLS-based CF. To address these problems, a broad collaborative filtering with adjusted cosine similarity (ACOS), named MC-ABCF, is proposed by fusing matrix completion (MC). The matrix completion technique is first used to complete the rating matrix in order to relieve the phenomenon of data sparsity. Subsequently, the adjusted cosine similarity is utilized to find the nearest neighbors of a given user or item. As a result, the problem of rating scales difference between different users is avoided to some extent. The BLS is finally employed to establish complex nonlinear relationships between users and items. Extensive experiments on three benchmark datasets demonstrate that the proposed MC-ABCF model can mitigate the difficulties of data sparsity and user rating scales difference to a certain extent. Taking RMSE as an example, the MC-ABCF algorithm achieved an average improvement of 8.48 % compared to the BCF model across three datasets. In addition, ablation studies have shown the application of MC or ACOS is beneficial for the recommendation performance. • Matrix completion is used to complete the user-item rating matrix to overcome the data sparsity problem effectively. • The proposed model adopts adjusted cosine similarity to measure user/item proximity to alleviate the rating scale difference. • Ablation experiments validate that both matrix completion and adjusted cosine similarity can improve the prediction accuracy. • Extensive experiments prove that the proposed model has better recommendation accuracy and efficiency. [ABSTRACT FROM AUTHOR]

Recommender systems filter information to meet users' personalized interests actively. Existing graph-based models typically extract users' interests from a heterogeneous interaction graph. They do not distinguish learning between users and items, ignoring the heterogeneous property. In addition, the interaction sparsity and long-tail bias issues still limit the recommendation performance significantly. Fortunately, hidden homogeneous correlations that have a considerable volume can entangle abundant CF signals. In this paper, we propose a light dual hypergraph convolution (LDHC) for collaborative filtering, which designs a hypergraph to involve heterogeneous and homogeneous correlations with more CF signals confronting the challenges. Over the integrated hypergraph, a two-level interest propagation is performed within the heterogeneous interaction graph and between the homogeneous user/item graphs to model users' interests, where learning on users and items is distinguished and collaborated by the homogeneous propagation. Specifically, hypergraph convolution is lightened by removing unnecessary parameters to propagate users' interests. Extensive experiments on publicly available datasets demonstrate that the proposed LDHC outperforms the state-of-the-art baselines. • Homogeneous user/item correlations bring entangled CF signals beyond the single CF signal in heterogeneous user-item interactions, which enables hybrid learning with much more reliable CF signals to alleviate interaction sparsity and confront long-tail bias. • This work proposes a light dual hypergraph convolution (LDHC) model by hybrid learning with a hypergraph convolution to predict interests, which performs a two-level interest propagation within the heterogeneous correlations and between the homogeneous ones. • A light convolution is introduced in hypergraph-based interest propagation to lighten the LDHC model for reducing the burden and difficulty in training. [ABSTRACT FROM AUTHOR]

Recently, contrastive learning has been shown to be effective in improving sequential recommendation, alleviating noise and long-tail issues caused by real-world noise and data sparsity, and enabling models to learn high-quality user representations. The data augmentation approach of contrastive learning mainly includes randomly changing the sequence data and adding randomly perturbations, and by pulling the distance between the original sample and the positive sample and pushing the distance between the original sample and the negative sample to enhance alignment and to optimize the uniformity of representation space. However, these approaches may use false positive/negative samples, learn biased user representations, and compromise the uniformity of the representation space, leading to second-best recommendation results. To solve these issues, we propose a generic Simple Debiased Contrastive Learning for Sequential Recommendation framework (SimDCL), to alleviate these biased samples, optimize the uniformity of the representation space, and enhance anti-noise ability. Specifically, we augment the samples based on noise, and optimize the uniformity of the representation space through gradient algorithms, and design an object filtering approach to punish false positive and false negative samples in user interaction sequences. Finally, we combine contrastive learning and sequential recommendation for multi-task collaborative training to improve the model's anti-noise capability and the quality of recommendation. The empirical studies on four benchmark datasets prove the superiority of our approach, and the code and datasets are available at URL. [ABSTRACT FROM AUTHOR]

An intelligent recommendation system facilitates users by recommending items based on the user's preferences. Most of the existing recommender systems employ collaborative filtering which consider homogenous data only and does not exploit different knowledge sources. Such approaches encounter many challenges such as Cold Start problem, Sparsity, and Scalability. To overcome these setbacks, a novel deep learning & semantic fusion-based recommendation system named DLSF is proposed in this paper which collaborates multiple knowledge sources by employing three modules to generate top-N recommendations. The first module precludes sparsity by using latent representations and deep learning for understanding user-item interactions. The second module exploits semantic information provided to similar items by the users for dealing with cold start problem. And the third module works for new users and items which don't have any prior information. A comprehensive experimental evaluation is conducted over benchmark datasets. The proposed DLSF showed on average 16.02% and 16.71% more efficient over the baselines recommendation systems taken into account in terms of MAE and RMSE respectively. DLSF has found to have significant efficiency in terms of Precision, Recall and overall accuracy. [ABSTRACT FROM AUTHOR]

The purpose of the recommendation systems (RS) is to provide personalized information filtering for users by analyzing their interactions with items. However, the sparsity of data in RS, which means that the number of negative samples is much larger than the number of positive samples, causes the necessity of the negative sampling method. Previous work has shown that an appropriate negative sampling method, rather than simple random negative sampling, can significantly improve the performance of recommendation algorithms. Unfortunately, the existing works have significant shortcomings in explainability and computational complexity. To remedy these shortcomings, we conducted a theoretical analysis of the gradient of the Bayesian Personalized Ranking (BPR) loss function and propose an explainable negative sampling method. Moreover, we conducted experiments on real-world datasets and find that the proposed negative sampling method improves the recommendation performance by an average of 10.23% over uniform negative sampling. Finally, the effectiveness of the proposed method is demonstrated through comparison with other methods and hyperparameter sensitivity analysis. • Proposed method has the advantages of explainability and high computation speed. • We propose the theoretical and empirical basis of the proposed negative sampling method. • Proposed method can also be generalized to deep learning recommendation method. [ABSTRACT FROM AUTHOR]

Similarity is a vital component for neighborhood-based collaborative filtering (CF). To improve the quality of recommendation, many similarity methods have been proposed and analyzed in recent decades. However, nearly all traditional similarity methods and many advanced similarity methods only utilize corated items among users to compute their similarity, which provides limited information in cold-start/sparse scenarios and yields misleading results. In addition, although a few advanced hybrid similarity models consider items beyond corated items, which can partly mitigate the above limitation, they still have drawbacks, such as disregarding penalizing noncorated items that have many disadvantages. In this paper, we explore a new robust hybrid similarity model, namely Wasserstein distance-based CF (WCF) model, for mitigating the cold-start problem of CF in sparse data. Specifically, we measure item similarity via the Wasserstein distance, which can help circumvent the drawbacks in the Bhattacharyya coefficient and KL divergence that are used in the literature, and is thus more robust in a cold-start/sparse scenario. Besides, we further design a new multiplicative user similarity formula which identifies all noncorated items as a whole to prioritize the importance of corated items and impair the negative effects of noncorated items, which will also play an important role in a cold-start/sparse scenario. In addition, we also propose two novel heuristic similarity factors to impair the negative effects of popular users and items as supplements. We conduct extensive experiments on five real-world benchmark recommendation datasets to test WCF. The experimental results show the superiority of WCF over other existing similarity methods in cold-start/sparse scenarios. • A novel formula is proposed to compute user similarity. • The Wasserstein distance is adopted to compute item similarity. • Two novel factors are proposed to enhance the reliability of user similarity. • Results show that our proposed model outperforms existing similarity methods. [ABSTRACT FROM AUTHOR]

Recommender systems are developed to personalize services for each user. The focus of recommender systems is to accurately discover the unknown preferences of users. To address this issue, neighbor-based recommender systems compute a weighted average of neighbors' known preferences to predict the unknown preferences of an active user. Its weights represent the influence of neighbors' preferences on the active user's ones. Previous research estimates these weights solely based on users' observed preferences. However, preferences are commonly expressed through numerical ratings, presenting a challenge for users to grasp the rating scale and assign the most precise numerical values. Therefore, the observed rating data is characterized by sparsity, a lack of precision, and insufficient detail. In modern recommender systems, besides collecting ratings, it's entirely possible to gather various types of information. One valuable data source proven to enhance recommender systems significantly is textual reviews authored by users following their experiences with items. In this study, we utilize Bert models to derive user vectors from observed textual reviews. These vectors are then employed to estimate weights between the active user and each neighbor in a neighbor-based recommender system. In contrast to numerical ratings, textual reviews provide a more detailed and precise representation of users' preferences. This contributes to improving the performance of neighbor-based recommender systems. [ABSTRACT FROM AUTHOR]

In recommendation systems, users often click on some items that are distinct from historically clicked items. This verifies the existence of interest gaps between the historical interests reflected by historical behaviors and current interests. The gaps pose key challenges for interest modeling in recommendation systems, i.e., learn user current interests from historical behaviors. To bridge interest gaps, various solutions have been proposed from a model perspective, and mainly focus on capturing interest diversity and interest evolution process (the changes in user interests over time). Different from existing solutions, our key idea is smoothing the historical interests from a data perspective, so that interest gaps can be smoothed and bridged. To realize this idea, we propose two types of data-driven smoothing approaches based on the principle of collaborative filtering, i.e., click-centric smoothing and user-centric smoothing. The approaches can be easily applied to a number of existing solutions and enhance their performances. To ensure that the smoothed interests completely and accurately cover user current interests (coverage and accuracy), the click-centric smoothing leverages the relevant clicks of a click to smooth the interests on the click, and the user-centric smoothing leverages the relevant clicks of a user click sequence to smooth the interests learned from the click sequence. The "smooth" is to enhance the coverage, and "relevant" is to enhance the accuracy. To search the relevant clicks, we develop an iterative Bayesian inference over a click-flow graph, i.e., a new implementation of collaborative filtering. To further enhance the accuracy, we propose new constrained attention networks to generate high-quality smoothing factors from these relevant clicks. Both offline and online experiments show the data-driven smoothing approaches significantly outperform the baselines. The user-centric smoothing has been deployed on the e-commerce platform of Meituan and achieved 1.51% elevation of Click-through & Conversion Rate and 0.53% elevation of Click-through Rate. • We are the first to apply the smoothing mechanism to bridge interest gaps in the interest modeling task. • We propose two types of data-driven smoothing approaches based on the assumption of collaborative filtering. • A novel iterative Bayesian inference framework on a click-flow graph is designed to search relevant clicks. • New constrained attention networks are proposed to generate high-quality smoothing factors. • Both offline and online experiments show that our models significantly outperform the SOTA baselines. [ABSTRACT FROM AUTHOR]

• This article collects and analyzes learners' historical learning data, establishes learners' learning models. • This article also designs a comprehensive set of social functions to promote interaction and knowledge sharing among learners. • Effectiveness and feasibility of the designed English learning social platform were verified through analysis and experimentation of actual user data. The role of social networks in distance education cannot be ignored. Through social networking platforms, students can communicate and interact with other classmates, share learning experiences and problem-solving methods, and provide learning support and encouragement to each other. However, traditional English learning platforms have problems such as limited content and lack of interactivity. This article collects and analyzes learners' historical learning data, establishes learners' learning models, and learns about their learning habits, preferences, and other information. Subsequently, a collaborative filtering algorithm is used to match learners with other learners with similar learning behaviors and interest preferences, identify similar groups of learners, refer to the choices and preferences of learners who are similar to them, and recommend suitable learning resources for them. In addition to personalized recommendations, this article also designs a comprehensive set of social functions to promote interaction and knowledge sharing among learners. The results showed that the effectiveness and feasibility of the designed English learning social platform were verified through analysis and experimentation of actual user data. User feedback shows that the personalized recommendations and social functions provided by the platform have significantly improved learning outcomes and communication between learners. [ABSTRACT FROM AUTHOR]

• This study suggests a unique method for recommending films based on analysis of user preference data that combines ALS collaborative filtering with deep learning techniques. • The input in this case is gathered as web data based on previously performed user searches and then processed for noise reduction and normalisation. • Convolutional multimodal auto multilayer graph with ALS collaborative filtering (CMAMG_ALSCF) was used to classify this processed data according to user evaluations and interests. • Movies that are related to the interests of users are recommended by examining the similarity between users and other users or the similarity between movies and other movies. • For several movie recommendation datasets, experimental analysis is done in terms of training accuracy, validation accuracy, RMSE, and average precision. Development of recommender systems has recently emerged as a prominent study field that has drawn the attention of several scientists and researchers worldwide. Various fields, such as music, movies, books, news, search queries, and commercial goods, employ recommender systems. One of the well-liked and effective RS strategies is the collaborative filtering algorithm, which seeks out users who are quite similar to active one to propose products. This study suggests a unique method for recommending films based on analysis of user preference data that combines ALS collaborative filtering with deep learning techniques. The input in this case is gathered as web data based on previously performed user searches and then processed for noise reduction and normalisation. Convolutional multimodal auto multilayer graph with ALS collaborative filtering (CMAMG_ALSCF) was used to classify this processed data according to user evaluations and interests. Movies that are related to the interests of users are recommended by examining the similarity between users and other users or the similarity between movies and other movies. For several movie recommendation datasets, experimental analysis is done in terms of training accuracy, validation accuracy, RMSE, and average precision. [ABSTRACT FROM AUTHOR]

The tremendous growth in information over the last decade leads to information overwhelming problems for accessing personalized products. The recommender framework that retrieves user preferences on past interactions is known as collaborative filtering (CF). Although, CF is a prevalent technique amongst the techniques applied in the recommender environment. However, it suffers from many problems like information sparsity, scalability, cold-start, etc. Many investigations have been made to tackle some of these issues with the help of matrix factorization (MF) approaches. However, MF cannot handle the nonlinearity among the data. Deep learning is an advanced learning technique that has shown success in many applications such as image classification, computer vision, natural language processing, etc. Little work has been reported on deep learning techniques in the recommender domain. We propose an efficient deep collaborative recommender system that embeds item metadata to handle the nonlinearity in data and sparsity. The model consists of two stages, wherein the first stage, a neural network, is used to retrieve the data's nonlinear features through embedding vectors. These vectors are concatenated together and fed as input to the second stage of the model. The output of the model yields a partially observed rating. The input and the parameters are simultaneously optimized and updated to minimize errors. The proposed strategy is evaluated against the benchmark techniques on two well-known datasets. The exploratory outcomes signify our approach's exactitude and efficiency. Moreover, the missing values can also be recovered by propagating the embedding vectors from the input to the output layers. [ABSTRACT FROM AUTHOR]

Recently, cross-domain recommendation systems have been very helpful in improving the quality of recommendation and solving the problem of cold start and data sparsity. Cross-domain recommender systems allow the transfer of knowledge from one domain with dense ratings to other domain with sparse ratings. Such transfer of knowledge helps in addressing the data sparseness and cold start issues in traditional recommender systems. Although cross-domain recommendations have evolved significantly, yet employment of various factors, such as time, trust, and location remains a challenge. Most of the existing approaches ignore the important fact that at what specific time the user may be interested in the recommended item. Moreover, a person's trust level and sentiments may be influenced by the variation in the location and time, thereby affecting the decision making. In this paper, we propose a cross-domain recommender system that not only takes into account the time at finer granularity levels (e.g., hours, days, weeks, etc.), but also considers a persons location, trust level, and sentiment analysis while computing recommendations. Our proposed model, named as, Trust-Aware Spatial-Temporal Activity based Denoising Autoencoder (TSTDAE), employs autoencoder-based deep-learning models to generate top-N recommendations for a given user and addresses the cold-start problem in the cross-domain scenario of 'User Overlap'. The proposed work is fivefold: i) Filter out the users' biased profiles based on sentiment analysis. ii) Learn the features using autoencoder and then perform clustering among the users of source and target domains to discover the best neighbors. iii) Compute the trust and ratio of preference bias between active user (the user to whom top-N items are recommended) and their neighbors and grade the neighbors based on unbiased preferences iv) Project the best time for recommending the items to an active user and generate the top-N recommendations. We have evaluated the proposed model on a public dataset of e-commerce retail service 'AliExpress' for the evaluation metrics: Precision, Mean Absolute Precision (MAP), Normalized Discounted Cumulative Gain (NDCG), and Hit Ratio (HR). The experimental results showed improved performance of the proposed system over the existing models. [ABSTRACT FROM AUTHOR]

A Collaborative Filtering (CF) method predicts an unknown overall rating of a target user towards an item based on the known overall ratings of the users that are similar to the target user. The similarity between two users is generally found based on their overall ratings toward items that both have reviewed. Two users may have similar overall ratings towards a given item, but different sentiments towards various aspects of the item. Understanding the effect of user sentiment towards specific aspects on overall ratings will sharpen estimates of user similarity as well as provide an rationale for making specific recommendations. We propose an Aspect-Sentiments based Multilevel Clustering of Users (ASMCU) approach that finds the multiple clusters of users similar to a specific user where similarity between users is based on various aspect sentiments. The proposed ASMCU CF approach can be used to predict both the overall ratings and the aspect-sentiments. The ASMCU based CF approach performed mostly better than and sometimes comparable to the eight well-established CF methods that rely only on the overall ratings or a particular aspect-sentiments. Note however that the ASMCU can also explicitly justify the recommendation in terms of aspect sentiments. We evaluated our approach using three datasets: One Hotel dataset and Two Beer datasets. The Hotel dataset involved six aspects and each Beer dataset has four aspects. Each dataset has one overall rating matrix and one sentiment tensor. [ABSTRACT FROM AUTHOR]

Collaborative Filtering (CF) filters the flow of data that can be recommended, by a Recommender System (RS), to a target user according to his taste and his preferences. The target user's profile is built based on his similarity with other users. For this reason, CF technique is very sensitive to the similarity measure used to quantify the dependency strength between two users (or two items). In this paper we provide an in-depth review on similarity measures used for CF-based RS. For each measure, we outline its fundamental background and we test its performance through an experimental study. Experiments are carried out on three standard datasets (MovieLens100k, MovieLens1M and Jester) and reveal many important conclusions. In fact, results show that ITR and IPWR are the most suitable similarity measures for a user-based RS while AMI is the best choice for an item-based RS. Evaluation metrics show that under the user-based approach, ITR obtains an MAE equal to 0.786 and 0.731 on MovieLens100k and MovieLens1M, respectively. Whereas, IPWR reach an MAE equal to 3.256 on Jester. Also, AMI gets under the item-based approach an MAE equal to 0.745, 0.724 and 3.281 on MovieLens100k, MovieLens1M and Jester, respectively. [ABSTRACT FROM AUTHOR]

Recommender Systems are a predictive model for personalized suggestions utilizing past interactions and experiences. Collaborative filtering is the most popular and successful approach. The core idea behind this approach is that the users expressing similar preferences in the past are considered similar and will continue to like similar recommendations in future. The similarity among the items between past and future references, however, affects diversity and coverage of the recommendation system. In this work, we focus on a less usual direction for recommendation systems by increasing the probability of retrieving unusual and novel items in the recommendation list, which are, or can be, also relevant to the users. Most of prevailing techniques for incorporating diversity are based on re-ranking methodology, which shrinks the domain of user's exposure to serendipitous items. To overcome this issue, we propose a methodology Clus-DR (Cluster-based Diversity Recommendation) that uses individual diversity of users and then pre-trained model for diverse recommendation generation. Instead of relying on re-ranking approach, we use different clustering techniques to have different groups of users with similar diversity. Experimental results using datasets of diverse domains indicate the effectiveness of the proposed Clus-DR methodology in diversity and coverage while maintaining acceptable level of accuracy. [ABSTRACT FROM AUTHOR]

Nowadays, the increasing number of smart objects and devices which are connected to the Internet leads to the introduction of numerous IoT services. However, due to the overloading of information, many users suffer from several difficulties in obtaining useful and relevant services that meet their needs and interests. Recommender Systems (RSs) provide an efficient solution to this complex problem. Yet, the existing IoT recommender systems are based on objects owned by the users themselves because it is a serious problem, in IoT recommender systems, to find alternative sources of the missing data to fulfill the service requirements. Whereas, the IoT data generated by various objects is inefficiently used. An efficient IoT service recommendation can be reached based on Social IoT (SIoT) using the generated data by various IoT devices owned by friends and friends of friends. In this paper, we propose a hybrid technique that combines implicit collaborative filtering and ontology to recommend personalized IoT services to users. Ontology is used for modeling the SIoT, where we incorporate the social relationships among objects into the recommendation process alongside the ratings while collaborative filtering predicts ratings and generates recommendations. The evaluation results show that our proposed recommendation technique outperforms the collaborative filtering using SIoT on its own without SIoT in terms of personalization and recommendation accuracy. [ABSTRACT FROM AUTHOR]

• A simple yet flexible multi-branch neural network for collaborative filtering. • Representing a user's multifaceted interests with multiple latent vectors. • Producing better personalized recommendations via multi-branch neural matching. Personalized recommendation serves as an indispensable functionality in many online services, where the key is to model the user's preference based on past user-item interactions. Many well-known models learn a latent vector for each user and each item. Nevertheless, a single latent vector might not be able to well represent a user's multifaceted interests revealed by the wide variety of items consumed by him/her. In this paper, we present a flexible framework named Mul ti-branch S ub- i nterest m atching Net work (MulSimNet) for personalized recommendation. Each user is explicitly represented by multiple sub-interest vectors as well as a general interest vector, and an item is matched against these vectors through different branches, where each branch is realized by appropriate neural networks. The compatibility scores output by multiple sub-interest matching branches are first aggregated via the max operator, which is then fused with the output of the general interest matching branch to estimate the user's affinity with the item. A comprehensive set of experiments on benchmark datasets demonstrate the effectiveness of the proposed framework in fully capturing the user's multifaceted interests and yielding better top-N recommendations. [ABSTRACT FROM AUTHOR]

Recommender systems are popular tools used in many applications, such as e-commerce, e-learning, and social networks to help users select their desired items. Collaborative filtering is a widely used recommendation technique that employs previous ratings of users to predict their future interests. Lack of sufficient ratings often reduces the performance of collaborative filtering recommendation methods. Additional side resources, such as trust relationships and tag information can be employed to enhance the recommendation accuracy. However, trust and tag data are often heavily sparse as users mainly provide insufficient information about these side resources. Moreover, such additional resources have generally large dimensions which result in increasing the computational complexity of recommendation models in calculating similarity values. To cope with these problems, a new recommendation model is proposed that utilizes deep neural networks to model the representation of trust relationships and tag information. To this end, a sparse autoencoder is used to extract latent features from user-user trust relationships and user-tag matrices. Then, the extracted latent features are utilized to calculate similarity values between users, which are then used to form the nearest neighbors of the target user and predict unseen items. The proposed method can tackle the data sparsity problem and reduce the computational complexity of recommender systems as the extracted latent features have smaller dimensions in comparison to the original data. Experimental results on two benchmark datasets reveal the effectiveness of the proposed method and show its outperformance over state-of-the-art recommender systems. [ABSTRACT FROM AUTHOR]

The introduction of digital systems in manufacturing has increased the number of ways that assistant information reaches the human operators in assembly work spaces. Such digital systems can accustom to different degrees of support, based on the level of detail of the information and the used medium. As the production environment becomes increasingly flexible and the tasks are changing in a faster pace, a more agile way for recommendations of support or instructions is needed. This work will present and compare two ways to represent instructional support, namely, on-screen digital work instructions (DWI) and Augmented Reality (AR)- based instructions. Furthermore, in order to assist an operator with the appropriate instructional support, a recommender system (RS) is developed, which matches an operator's proficiency, a task's requirement and a level of instruction. A purely data-driven RS using Bayesian Probabilistic Matrix Factorization (BPMF) is compared with a skill-driven approach. The latter has the advantage that the recommended level of instruction is directly relatable to a skill-based operator proficiency metric. For the evaluation of the two instruction delivery approaches and for training and testing the RS, a group of 13 users were called to complete two industrial setups, alternating between the different instruction types. Both RS were able to outperform a naive baseline. [ABSTRACT FROM AUTHOR]

As recommender systems shift from rating-based to interaction-based models, graph neural network-based collaborative filtering models are gaining popularity due to their powerful representation of user-item interactions. However, these models may not produce good item ranking since they focus on explicit preference predictions. Further, these models do not consider side information since they only capture latent feature information of user-item interactions. This study proposes an approach to overcome these two issues by employing preference relation in the graph neural network model for collaborative filtering. Using preference relation ensures the model will generate a good ranking of items. The item side information is integrated into the model through a trainable matrix, which is crucial when the data is highly sparse. The main advantage of this approach is that the model can be generalized to any recommendation scenario where a graph neural network is used for collaborative filtering. Experimental results obtained using the recent RS datasets show that the proposed model outperformed the related baselines. • Pair-wise preference is used for representing user-item interactions in GNN. • Side Information is integrated to the GNN model using a learnable matrix. • Item-item similarity information is preserved during the preference generation. • The model can capture both neighborhood and latent feature information. [ABSTRACT FROM AUTHOR]

Representation learning, with its desired properties, including Alignment and Uniformity, has recently emerged as an effective approach in collaborative filtering (CF) for recommender systems. Alignment measures the distance between positive pairs, and uniformity describes the distribution of all samples on the unit hypersphere. Despite the effectiveness of existing research, the challenges of insufficient alignment and uniformity bias persist in studies on the desired properties of recommender systems: (1) Sparse interaction information leads to an insufficient alignment representation. (2) Calculating the uniformity loss based on duplicate samples introduces bias, which affects the overall representation uniformity. Building on aforementioned challenges, we propose a Similar Alignment and Unique Uniformity model called SUAU. SUAU mitigates insufficient alignment by introducing additional item-similar item pairs for model training and employing a uniqueness strategy to prevent uniformity bias. More specifically, we propose a similar items generation method that identifies the most similar items based on the interaction behavior of the original items' related users to form item-similar item pairs. The experimental results obtained on three highly sparse datasets demonstrate that the propose model achieves superior alignment and uniformity compared with state-of-the-art models, notably surpassing the best CF methods in terms of recommendation performance. We have open-sourced the source code on a publicly accessible website: https://github.com/ZzYUuuu/SUAU. [ABSTRACT FROM AUTHOR]

Co-clustering followed by applying recommendations within each user–item cluster is an effective way to enhance recommender systems. This approach reduces sparsity and ensures that users are only recommended items from their respective clusters. Recently, it has been shown that due to the nature of user–item data, items or users often belong to multiple clusters, making fuzzy co-clustering more appropriate. Unfortunately, despite its potential, few fuzzy co-clustering methods have been developed. In this paper, we introduce two novel fuzzy co-clustering methods based on non-negative matrix factorization (NMF) due to the non-negativity of the rating matrix. One method is based on linear NMF, while the other incorporates a conformal mapping called inverse stereographic projection to appropriately compute the existing similarities within NMF. Both methods offer low computational complexity and better quality compared to previous approaches. However, the adjusted method provides a nonlinear factorization that aligns more closely with the nature of the data. Implementation results on different well known datasets and different recommender systems, evaluated using multiple metrics, demonstrate the quality of these methods, with the adjusted method outperforming the other proposed method. [ABSTRACT FROM AUTHOR]

Over the past few years, deep learning networks have been applied in the estimation of the Click-Through Rate (CTR). The main task of the CTR model lies in predicting a user's positive/negative response to an item. Most of existing CTR models are constituted of two parts: a deep neural network (DNN) and a wide model (e.g., deep cross network). The wide models are generally designed to learn the feature interaction and contain much more parameters than the DNN. As a consequence, it is quite difficult for existing CTR models to work on the resource-limited devices. Blindly removing the wide model will significantly encumber the predictive performance of the algorithm. In this paper, we adopt the knowledge distillation to train a single multi-branch network and assemble all branches as a teacher network. Each branch network is a simple DNN which not only matches the ground-truth label distribution but also aligns the prediction distribution of the teacher network. One of branch networks, termed as A daptive D eep N eural N etwork (ADNN), is trained independently and further combined with a wide model to learn feature interactions. Our method does not require pre-training any high-capacity teacher models, which endows our method with higher efficiency compared with existing ones. The experimental results tested on Criteo and Avazu datasets show that the hybrid model outperforms state-of-the-art methods, and the light model ADNN also has a considerable performance accuracy over certain modern complex models, demonstrating the superiority of the methodology. [ABSTRACT FROM AUTHOR]

The ubiquity of the Internet of Things (IoT) across diverse applications underscore its pivotal role in seamlessly integrating physical devices to facilitate efficient data collection, analysis, and automation. Consequently, ensuring the Quality of Service (QoS) emerges as a critical imperative. While numerous studies have proposed methodologies focusing on resource optimization, network management, and data processing techniques, several previous approaches to QoS prediction may face constraints in the Internet of Things (IoT) environment, where the dynamic nature of IoT environments demands predictive models capable of adapting to varying conditions. Integrating user and service location data into QoS prediction models is paramount, as it enables personalized service delivery tailored to the user's specific context, thus enhancing the user experience and overall system performance. This paper presents a novel approach to QoS prediction in IoT, harnessing self-attention and collaborative filtering (CF) techniques while incorporating location-based features such as distance from the user to the service. Experimental evaluations on benchmark datasets reveal that our proposed model improves prediction accuracy by up to 7 % compared to existing methods, underscoring its efficacy in enhancing QoS provisioning in IoT environments. [ABSTRACT FROM AUTHOR]

Most approaches used in social recommender systems concentrate mostly on positive relationships within the social network, frequently ignoring the insightful information that negative relationships can offer. A more thorough understanding of user characteristics and behaviors within a social network can result from integrating positive and negative links to construct a signed graph. However, there are two major challenges in developing signed social recommender systems. Considering negative links in these systems needs to be done carefully to make sure that these relationships are appropriately represented and used. Moreover, it is difficult to predict user interactions and preferences effectively when relationships represented by positive and negative links conflict, known as social inconsistency. Therefore, working with signed graphs requires the management of social inconsistency. To address these issues, in this paper, a signed social recommender system called SiSRS is presented using a deep architecture combined with network representation learning. The SiSRS is composed of three principal components. First, the application of signed graph attention networks is explored to collate and disseminate information across the network via graph motifs, leading to the creation of node embeddings. Second, a deep autoencoder model is employed to assimilate top-k semantic signed social data within the deep learning framework. Third, a novel loss function is formulated to reduce the impact of social inconsistency. The efficacy of the SiSRS model was evaluated through extensive experiments conducted on two datasets in terms of various evaluation metrics. The results indicated a superior performance of this approach compared to existing state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Traditional recommendation techniques often prioritize target behavior in practical recommendation scenarios(e.g., follow, play and buy). However, these approaches suffer from data sparsity issues and may not fully capture user's personal preferences. To address this deficiency, multi-behavior recommendation technology has emerged, leveraging users' multi-behavioral interactions for recommendation. Nevertheless, certain multi-behavior recommendation methods learning behavioral information from each behavior separately and then aggregate them before making recommendation, which inadvertently neglects the intrinsic connections between different behaviors. In some scenarios, user behavior often occurs in a fixed order, such as view - > cart - > buy in e-commerce platforms. In this work, we propose a novel C ascading G raph C onstrastive L earning (CGCL) framework for Multi-Behavior recommendation. Specifically, we devise a graph contrastive learning block to learn distinctive user behavioral representations for each type of interaction. Leveraging the recommendation task, we aim to capture user preferences, while the contrastive learning provides supplementary supervisory signals to refine the user and item representation. By acknowledging the sequential order of behaviors, we utilize the cascading structure within our model to iteratively propagate and purify the personalized preferences of users. Extensive experimental results and ablation studies on three real-world datasets have shown that our CGCL framework outperforms various state-of-the-art recommendation methods and validated the effectiveness of our approach. • Utilizing intrinsic connections among behaviors to refine user preference. • Novel cascading graph contrastive learning approach for multi-behavior recommendation. • A simple and efficient noise-based augmentation for graph contrastive learning. • Uniform embedding distribution drives CGCL's recommendation performance. • Multi-behavioral data is undeniably effective, yet excess can be detrimental. [ABSTRACT FROM AUTHOR]

In collaborative filtering recommendations, part of studies mainly focus on quantitatively learning user preferences based on explicit rating information. However, due to the uncertainty and complexity of decision-making scenarios, user preferences usually express ambiguity and diversity. Therefore, to fully capture user preferences from both qualitative and quantitative perspectives and improve the accuracy and reliability of recommendations, we propose a novel fuzzy neural collaborative filtering for recommender systems. Firstly, we qualitatively divide user preferences into fine-grained categories based on picture fuzzy set and transform the rating matrix into four preference degree matrices. Then, we employ Bernoulli matrix factorization to independently learn preference features of each submatrix to obtain picture fuzzy numbers. To enhance the adaptability of picture fuzzy set in recommender systems, we design an improved scoring function based on picture fuzzy number to effectively distinguish differences in user preference levels. In addition, we use a simple deep neural framework to quantitatively extract global preferences of users and fuse it with fine-grained preferences of users to thoroughly predict user preferences. Experiments on four benchmark datasets show that the proposed model outperforms other comparison methods in terms of recommendation accuracy. Compared with the closest competitor method, our model has the highest improvements of 5.08% and 4.04% in the metrics HR@10 and NDCG@10, respectively. • PFS is used to divide fine-grained user preferences during BeMF training. • Improved scoring function is designed to the adaptability of PFS in RSs. • A simple neural network MLP is employed to capture global user preferences. • Two preference modules are fused to thoroughly predict user preferences. • Results reveal that our model effectively improves the performance of RSs. [ABSTRACT FROM AUTHOR]

In recommender system, the proportion of interacted items with users is extremely sparse compared to the total number of items. This data sparsity problem particularly affects collaborative filtering methods. Even in modern collaborative filtering recommendation models based on deep learning, effectively acquiring more neighbor information and exploring how to utilize this information efficiently remains a promising research area. This paper not only explores heuristic algorithms for neighbor selection but also extends the application of neighbor selection to the GAN framework and negative sampling. To capture user neighbors that contain more potential feature information, this paper proposes a neighbor strategy that integrates explicit and implicit user information. It aggregates reciprocal neighbors of users based on the fusion of their explicit and implicit information, aiming to capture user preference information through neighborhood information. Furthermore, by designing an MRNGAN model based on a twin-tower generator, this paper successfully embeds user neighborhood information as feature embedding into the generator of GAN. In addition, this paper also proposes a heuristic negative sampling mechanism in the recall phase. It uses explicit and implicit information to fuse the nearest neighbors of items with the farthest neighbors of users. Negative samples that contain user neighborhood information are embedded into the discriminator of MRNGAN. The deep relationships between nearest and farthest neighbors are explored through the adversarial training of GAN, and their impact on recommendation performance is validated. The experimental results on three real-world datasets show that the MRNGAN model significantly improves the accuracy of Top-N recommendation. [ABSTRACT FROM AUTHOR]

Nowadays, e-commerce platforms, such as Amazon, Flipkart, Netflix and YouTube, extensively use recommender systems (RS) techniques. Collaborative filtering (CF) is used widely among all RS techniques. A CF analyzes the user's preference from past data, like ratings, and then suggests actual items to the intended user. The existing techniques compute the similarity between users/items and predict the ratings. However, most of them indicate the user's preference for the items using a single technique, which may produce poor results. This paper proposes a hybrid CF technique to enhance the movie recommendation (HCFMR). The HCFMR consists of two modules. The first module finds the prediction score with the help of matrix factorization (MF) and passes the prediction score as input to the prediction algorithm, i.e., extreme gradient boosting (XGBoost). The second module generates handcrafted features, such as similar users and movies, along with the user, item and global average. Finally, these features are supplied to the XGBoost to predict the rating score of the movie and recommend the topmost movie to the user. We conduct various simulations on real-world datasets to verify the effectiveness of the proposed technique against the baseline techniques. The exploratory outcomes signify that the HCFMR technique outperforms the baselines and provides a better prediction on the benchmark datasets. • To deploy a matrix factorization model for the predictions and use them in the model. • To generate handcrafted features of the top five similar users and items. • To combine all features to make a hybrid collaborative filtering model. [ABSTRACT FROM AUTHOR]