Cold Start Issues Research Articles

TGVx: Dynamic Personalized POI Deep Recommendation Model

Personalized points-of-interest (POI) recommendation is very important for improving the service quality of location-based social network applications. It has become one of the most popular research directions in the industry and academia. However, the realization of high-quality personalized POI recommendation faces three major challenges: (i) the interest drift issue caused by the spatiotemporal dynamics of user check-in behavior, (ii) how to integrate as much heterogeneous information as possible to alleviate data sparseness and cold start issues, and (iii) how to use implicit feedback to model complex high-order nonlinear user-POI interactions. To jointly address all these challenges, we propose the TGVx recommendation framework and establish a dynamic personalized POI deep recommendation model, where T and G respectively represent time and geographic factors, V represents out-of-town visitors, and x represents time slot number. TGVx is composed of x parallel TGV models where the TG module mines high-order nonlinear user-POI interaction relationships and integrates multisource heterogeneous information, and the V module transfers the check-in records of out-of-town visitors in hometowns and generates pseudo check-in records in the target city. Technically, we design a new unsupervised deep learning network T-SemiDAE for the TG module. We built a POI-word heterogeneous network for the V module and used graph embedding technology to match the most similar POIs across cities and transfer check-in records. The experimental results on the actual datasets show that the TGVx model is always better than other advanced models in terms of accuracy and diversity for local and out-of-town recommendation scenarios. Compared with the best baseline model semi-deep auto-encoder with a conditional layer the average improvement rates of accuracy and diversity of TGVx are 17.1% to 58.6% and 2.25% to 28.86%, respectively. In theory, our research effectively uses data science and analysis methods to design a recommender system. In practice, our research is motivated by practical problems, and the research results have high practical promotion value. History: Accepted by Ram Ramesh, Area Editor for Data Science & Machine Learning. Funding: This work was supported by the National Natural Science Foundation of China [Grants 71831003, 72293563, and 72172025], the Dalian Science and Technology Talent Innovation Support Policy Project [The High-level Talent Innovation Team Project 2022RG17], and the Liaoning Provincial Department of Education [Basic Scientific Research Project LJKMZ20221582]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2023.1286 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2021.0010 ) at ( http://dx.doi.org/10.5281/zenodo.7407123 ).

Personalized Bayesian optimization for noisy problems

In many real-world applications of interest, several related optimization tasks can be encountered, where each task is associated with a specific context or personalized information. Moreover, the amount of available data for each task may be highly limited due to the expensive cost involved. Although Bayesian optimization (BO) has emerged as a promising paradigm for handling black-box optimization problems, addressing such a sequence of optimization tasks can be intractable due to the cold start issues in BO. The key challenge is to speed up the optimization by leveraging the transferable information, while taking the personalization into consideration. In this paper, optimization problems with personalized variables are formally defined at first. Subsequently, a personalized evolutionary Bayesian algorithm is proposed to consider the personalized information and the measurement noise. Specifically, a contextual Gaussian process is used to jointly learn a surrogate model in different contexts with regard to the varying personalized parameter, and an evolutionary algorithm is tailored for optimizing an acquisition function for handling the presence of personalized information. Finally, we demonstrate the effectiveness of the proposed algorithm by testing it on widely used single- and multi-objective benchmark problems with personalized variables.

HCoF: Hybrid Collaborative Filtering Using Social and Semantic Suggestions for Friend Recommendation

Today, people frequently communicate through interactions and exchange knowledge over the social web in various formats. Social connections have been substantially improved by the emergence of social media platforms. Massive volumes of data have been generated by the expansion of social networks, and many people use them daily. Therefore, one of the current problems is to make it easier to find the appropriate friends for a particular user. Despite collaborative filtering’s huge success, accuracy and sparsity remain significant obstacles, particularly in the social networking sector, which has experienced astounding growth and has a large number of users. Social connections have been substantially improved by the emergence of social media platforms. In this work, a social and semantic-based collaborative filtering methodology is proposed for personalized recommendations in the context of social networking. A new hybrid collaborative filtering (HCoF) approach amalgamates the social and semantic suggestions. Two classification strategies are employed to enhance the performance of the recommendation to a high rate. Initially, the incremental K-means algorithm is applied to all users, and then the KNN algorithm for new users. The mean precision of 0.503 obtained by HCoF recommendation with semantic and social information results in an effective collaborative filtering enhancement strategy for friend recommendations in social networks. The evaluation’s findings showed that the proposed approach enhances recommendation accuracy while also resolving the sparsity and cold start issues.

H3Rec: Higher-Order Heterogeneous and Homogeneous Interaction Modeling for Group Recommendations of Web Services

Recommendations are important web services in the era of information explosion. Particularly, group recommendations aim to suggest new items to groups such that the members of groups are likely interested in. However, existing works still suffer from sparsity and cold-start issues (e.g., cold-start groups or items) for groups with few interactions on items. Most of them model the preferences or features of entities (i.e., users, items and groups) from heterogeneous interactions (i.e., user-item, group-item and user-group interactions) between two distinct types of entities, while ignoring the homogeneous interactions (i.e., user-user, item-item and group-group interactions) between entities of one type. To this end, we propose a new model, called H3Rec, which learns the representations of entities by developing two graph embedding layers based on an interaction graph of all entities. Specifically, the two graph embedding layers make full use of the hidden information in the <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">H</b> <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">igher-order</i> <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">H</b> <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">eterogeneous and</i> <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">H</b> <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">omogeneous interactions</i> of the graph. Therefore, H3Rec can alleviate the sparsity and cold-start issues and improve the performance of group recommendations. The experimental results on two real world datasets in different domains show the superiority of H3Rec in group recommendations, especially for cold-start groups and items.

CAML: Contextual augmented meta-learning for cold-start recommendation

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.

Enhancing sequential recommendation with contrastive Generative Adversarial Network

Sequential recommendation models a user’s historical sequence to predict future items. Existing studies utilize deep learning methods and contrastive learning for data augmentation to alleviate data sparsity. However, these existing methods cannot learn accurate high-quality item representations while augmenting data. In addition, they usually ignore data noise and user cold-start issues. To solve the above issues, we investigate the possibility of Generative Adversarial Network (GAN) with contrastive learning for sequential recommendation to balance data sparsity and noise. Specifically, we propose a new framework, Enhanced Contrastive Learning with Generative Adversarial Network for Sequential Recommendation (ECGAN-Rec), which models the training process as a GAN and recommendation task as the main task of the discriminator. We design a sequence augmentation module and a contrastive GAN module to implement both data-level and model-level augmentations. In addition, the contrastive GAN learns more accurate high-quality item representations to alleviate data noise after data augmentation. Furthermore, we propose an enhanced Transformer recommender based on GAN to optimize the performance of the model. Experimental results on three open datasets validate the efficiency and effectiveness of the proposed model and the ability of the model to balance data noise and data sparsity. Specifically, the improvement of ECGAN-Rec in two evaluation metrics (HR@N and NDCG@N) compared to the state-of-the-art model performance on the Beauty, Sports and Yelp datasets are 34.95%, 36.68%, and 13.66%, respectively. Our implemented model is available via https://github.com/nishawn/ECGANRec-master.

A Critical Review on Location Based Hybrid Filtering Recommender Systems

A recommender system is basically a type of information filtering system that suggests/recommends items based on the factors that constitute what the user is most interested in. The recommendations are typically provided in relation to different decision-making processes. Tourism is a social phenomenon where people deliberately travel in search of recreation, well-being, cultural exploration or get themselves softened up. But, the amount of information available online keeps expanding at exponential rates and thus the users have expressed their feeling of frustration at how challenging it is to find the appropriate information. This problem is called information overload. This is where the recommendation system comes into play which helps in solving the information overload problem. The hybrid system addresses the disadvantages where location-based recommendation systems are used individually, of which the most notable is the cold start issue. Furthermore, in order to improve the accuracy of the prediction to recommend items, these systems search for the ideal fusion of different approaches. Thus, the hybrid recommendation method solves the challenges like ‘cold start problem’, inability to capture changes in user behavior, sparsity and selecting correct choices for users. This paper explores the hybrid recommendation systems and other filtering techniques used in various fields, their challenges, how they can also be used for tourism recommender systems based on the longitudes and latitudes.

A Multi-strategy-based Pre-training Method for Cold-start Recommendation

The cold-start issue is a fundamental challenge in Recommender Systems. The recent self-supervised learning (SSL) on Graph Neural Networks (GNNs) model, PT-GNN, pre-trains the GNN model to reconstruct the cold-start embeddings and has shown great potential for cold-start recommendation. However, due to the over-smoothing problem, PT-GNN can only capture up to 3-order relation, which cannot provide much useful auxiliary information to depict the target cold-start user or item. Besides, the embedding reconstruction task only considers the intra-correlations within the subgraph of users and items, while ignoring the inter-correlations across different subgraphs. To solve the above challenges, we propose a multi-strategy-based pre-training method for cold-start recommendation (MPT), which extends PT-GNN from the perspective of model architecture and pretext tasks to improve the cold-start recommendation performance. 1 Specifically, in terms of the model architecture, in addition to the short-range dependencies of users and items captured by the GNN encoder, we introduce a Transformer encoder to capture long-range dependencies. In terms of the pretext task, in addition to considering the intra-correlations of users and items by the embedding reconstruction task, we add an embedding contrastive learning task to capture inter-correlations of users and items. We train the GNN and Transformer encoders on these pretext tasks under the meta-learning setting to simulate the real cold-start scenario, making the model able to be easily and rapidly adapted to new cold-start users and items. Experiments on three public recommendation datasets show the superiority of the proposed MPT model against the vanilla GNN models, the pre-training GNN model on user/item embedding inference, and the recommendation task.

A hybrid recommender system using topic modeling and prefixspan algorithm in social media

AbstractRoute schema is difficult to plan for tourists, because they demand to pick points of interest (POI) in unknown areas that align with their preferences and limitations. This research proposes a novel personalized method for POI route recommendation that employs contextual data. The proposed approach enhances the existing methods by considering user preferences and multifaceted tourism contexts. Due to the sparsity of the data, the proposed method employs two-level clustering (DBSCAN based on the Manhattan distance) that reduces the time to discover POI. In specific, this approach utilizes the following: first, a topic pattern model is employed to discover the users’ attraction diffusion while improving the user–user similarity model using a novel asymmetric schema. Second, it has used explicit demographic information to alleviate the cold start issue, and third, it proposes a new strategy for assessing user preferences and also combined the context parameters in the form of a vector model with the Term Frequency Inverse Document Frequency technique to find contexts’ similarity. Furthermore, our framework discovers a list of optimal candidate trips by involving personalized POIs in sequential patterns’ mining (SPM); also, it used an adjusted forgotten function to involve the date context of each trip. Based on two datasets (Flickr and Gowalla), our methodology beats other prior approaches in F-score, RMSE, MAP, and NDCG factors in the experimental evaluation.

A New Online Education Personalized Recommendation Algorithm

For online education platforms, a personalized recommendation system is crucial, and the collaborative filtering algorithm is the primary recommendation algorithm used. This study took the recommendation of crowdfunding platforms as a sample, and enhanced the collaborative filtering algorithm based on the user score and project attribute features of the crowdfunding platform, intending to resolve the cold start issue brought on by the platform's reliance on a single data source. The study concludes with experimental proof of the paper's suggested better method. This approach can alleviate the cold start issue to some degree. The prediction accuracy has been much enhanced in comparison with the conventionally advised method. The method can also adapt to user tastes over time, learning what they like and what they don't. It also has an excellent real-time suggestion impact. The performance verification of the algorithm in this research is also conducted using data from a live crowdfunding site, lending credence to the study's claim of greater practicality.

Personalized Relationships-Based Knowledge Graph for Recommender Systems with Dual-View Items

The knowledge graph has received a lot of interest in the field of recommender systems as side information because it can address the sparsity and cold start issues associated with collaborative filtering-based recommender systems. However, when incorporating entities from a knowledge graph to represent semantic information, most current KG-based recommendation methods are unaware of the relationships between these users and items. As such, the learned semantic information representation of users and items cannot fully reflect the connectivity between users and items. In this paper, we present the PRKG-DI symmetry model, a Personalized Relationships-based Knowledge Graph for recommender systems with Dual-view Items that explores user-item relatedness by mining associated entities in the KG from user-oriented entity view and item-oriented entity view to augment item semantic information. Specifically, PRKG-DI utilizes a heterogeneous propagation strategy to gather information on higher-order user-item interactions and an attention mechanism to generate the weighted representation of entities. Moreover, PRKG-DI provides a score feature as a filter for individualized relationships to evaluate users’ potential interests. The empirical results demonstrate that our approach significantly outperforms several state-of-the-art baselines by 1.6%, 2.1%, and 0.8% on AUC, and 1.8%, 2.3%, and 0.8% on F1 when applied to three real-world scenarios for music, movie, and book recommendations, respectively.

Cross-Domain Explicit–Implicit-Mixed Collaborative Filtering Neural Network

Collaborative filtering (CF) is a classical model for recommender systems. Though neural network-based CF models have been shown to perform well in some cases, they still suffer from the sparsity and cold-start issues. To address these issues, we propose a novel neural network-based CF model, termed the cross-domain explicit–implicit-mixed CF neural network (CEICFNet). The proposed model utilizes deep neural networks to learn latent factors not only from the explicit ratings and the implicit interactions but also in a cross-domain manner. In particular, domain-shared multilayer perception (MLP) networks are designed to learn the user rating latent factors and the user interaction latent factors from the explicit ratings and the implicit interactions, respectively, which act as bridges for transferring knowledge across domains. On the other hand, domain-specific MLP networks are designed to learn the item rating latent factors and the item interaction latent factors from the explicit ratings and the implicit interactions, respectively. Then, in each domain, based on the user rating latent factors and the item rating latent factors, the rating predictive representation for each user–item pair can be learned by an MLP. Similarly, the interaction predictive representation for each user–item pair can be learned. For integrating the explicit ratings and the implicit interactions, in each domain, a fully connected layer is used to automatically assign different weights to the rating predictive representations and the interaction predictive representations, based on which the final interaction probability can be generated. Extensive experiments are conducted on five datasets and the results have confirmed the effectiveness of our model.

An Overview of Heterogeneous Information Networks based on Recommendation System

Heterogeneous Information Networks (HINs) provide a natural way to represent different relationships among entities of different kinds, as a consequence they are valuable in many domains. Analysing and extracting understanding from HINs normally is based at the idea of meta paths, which are paths in the community schema denoting family members of different semantics among entities. Moreover, real-global HINs are frequently extremely big, containing thousands and thousands of nodes and edges. Therefore, exploring and analysing HINs not only requires interdisciplinary knowhow, having the ability both to interpret and pick out suitable meta paths inside the network, but additionally to run the analysis in an efficient and scalable manner. However, there may be a loss of tools to facilitate this assignment. Most real system include a huge wide variety of interacting, multityped components, while most current researches version them as homogeneous networks, without distinguishing different forms of objects and links in the networks. In recent years, an increasing amount of works have been proposed to present helper data in recommender system to alleviate information sparsity and cold-start issues. Among them, heterogeneous information networks (HIN)- based recommender systems give a brought together way to deal with meld different helper data, which can be joined with standard suggestion calculations to successfully upgrade the presentation and interpretability of models, and accordingly have been applied in many kinds of recommendation tasks. This paper gives a far reaching and efficient overview of HIN-based recommender systems, including four perspectives: ideas, strategies, applications, and resources.

Content-Based Collaborative Filtering With Predictive Error Reduction-Based CNN Using IPU Model

Recommender systems (RS) are strong tools for addressing the internet networking overload problems by considering past user ratings on multiple items with auxiliary data and suggests the better item to the end user. Traditional collaborative filtering (CF) and content-based methods were identified the interaction or correlation between users and the items. But they have failed to identify the join user-item interactions and suffering from incomplete cold start (ICS) and complete cold start (CCS) issues. To address the deficiencies of CF-based approaches, this article offers a novel deep learning based error predictions method along with CF based user-item interactions. Initially, incentivized/penalized user-based content-based collaborative filtering (IPU-CBCF) method is introduced for learning low-dimensional vectors of users and items, separately. The simulation results shows that IPU-CBCF using PER-CNN resulted in better performance as compared to the conventional approaches for all performance metrics like F1-score, recall, and precision, respectively.

Trust-aware denoising autoencoder with spatial-temporal activity for cross-domain personalized recommendations

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.

An attention-based deep learning method for solving the cold-start and sparsity issues of recommender systems

Matrix Factorization is a successful approach for generating an effective recommender system. However, most existing matrix factorization methods suffer from the sparsity and cold-start issues of the user-item matrix as the primary information source of recommender systems. Besides, they are not much scalable to apply to large real-world applications. A main idea to overcome the cold-start and sparsity issues is to use additional information sources such as user/item profiles or user reviews on items. In this paper, a novel Attention-based Deep Learning Recommender System, so-called ADLRS, is proposed to employ the information sources in the matrix factorization method framework. The proposed method uses a language model to represent contextual information such that important features are effectively embedded. Moreover, a deep autoencoder reduces the dimensionality of item vectors embedded by the language model. Then, these vectors are used as regularization terms in the matrix factorization framework to form an objective function. Then, an iterative algorithm is designed to solve the objective function and provide a method prediction of unknown rating values. Experimental results show that the proposed method achieves superior performance compared to other state-of-the-art ones in most cases. Moreover, the improvement rate for sparse datasets and cold items proves that the proposed method effectively deals with sparsity, cold start and scalability problems.

Hotel room personalization via ontology and rule-based reasoning

Purpose The increasingly competitive hotel industry and emerging customer trends where guests are more discerning and want a personalized experience has led to the need of innovative applications. Personalization is much more important for hotels, especially now in the post-COVID lockdown era, as it challenges their business model. However, personalization is difficult to design and realize due to the variety of factors and requirements to be considered. Differences are both in the offer (hotels and their rooms) and demand (customers’ profiles and needs) in the accommodation domain. As for the implementation, critical issues are in hardware-dependent and vendor-specific Internet of Things devices which are difficult to program. Additionally, there is complexity in realizing applications that consider varying customer needs and context via existing personalization options. This paper aims to propose an ontological framework to enhance the capabilities of hotels in offering their accommodation and personalization options based on a guest’s characteristics, activities and needs. Design/methodology/approach A research approach combining both quantitative and qualitative methods was used to develop a hotel room personalization framework. The core of the framework is a hotel room ontology (HoROnt) that supports well-defined machine-readable descriptions of hotel rooms and guest profiles. Hotel guest profiles are modeled via logical rules into an inference engine exploiting reasoning functionalities used to recommend hotel room services and features. Findings Both the ontology and the inference engine module have been validated with promising results which demonstrate high accuracy. The framework leverages user characteristics, and dynamic contextual data to satisfy guests’ needs for personalized service provision. The semantic rules provide recommendations to both new and returning guests, thereby also addressing the cold start issue. Originality/value This paper extends HoROnt in two ways, to be able to add: instances of the concepts (room characteristics and services; guest profiles), i.e. to create a knowledge base, and logical rules into an inference engine, to model guests’ profiles and to be used to offer personalized hotel rooms. Thanks to the standards adopted to implement personalization, this framework can be integrated into existing reservation systems. It can also be adapted for any type of accommodation since it is broad-based and personalizes varying features and amenities in the rooms.

Deep transfer learning with multimodal embedding to tackle cold-start and sparsity issues in recommendation system.

Recommender systems (RSs) have become increasingly vital in the modern information era and connected economy. They play a key role in business operations by generating personalized suggestions and minimizing information overload. However, the performance of traditional RSs is limited by data sparseness and cold-start issues. Though deep learning-based recommender systems (DLRSs) are very popular, they underperform when considering rating matrices with sparse entries. Despite their performance improvements, DLRSs also suffer from data sparsity, cold start, serendipity, and generalizability issues. We propose a multistage model that uses multimodal data embedding and deep transfer learning for effective and personalized product recommendations, and is designed to overcome data sparsity and cold-start issues. The proposed model includes two phases. In the first—offline—phase, a deep learning technique is implemented to learn hidden features from a large image dataset (targeting new item cold start), and a multimodal data embedding is used to produce dense user feature and item feature vectors (targeting user cold start). This phase produces three different similarity matrices that are used as inputs for the second—online—phase to generate a list of top-n relevant items for a target user. We analyzed the accuracy and effectiveness of the proposed model against the existing baseline RSs using a Brazilian E-commerce dataset. The results show that our model scored 0.5882 for MAE and 0.4011 for RMSE which is lower than baseline RSs which indicates that the model achieved an improved accuracy and was able to minimize the typical cold start and data sparseness issues during the recommendation process.

Federated Social Recommendation with Graph Neural Network

Recommender systems have become prosperous nowadays, designed to predict users’ potential interests in items by learning embeddings. Recent developments of the Graph Neural Networks (GNNs) also provide recommender systems (RSs) with powerful backbones to learn embeddings from a user-item graph. However, only leveraging the user-item interactions suffers from the cold-start issue due to the difficulty in data collection. Hence, current endeavors propose fusing social information with user-item interactions to alleviate it, which is the social recommendation problem. Existing work employs GNNs to aggregate both social links and user-item interactions simultaneously. However, they all require centralized storage of the social links and item interactions of users, which leads to privacy concerns. Additionally, according to strict privacy protection under General Data Protection Regulation, centralized data storage may not be feasible in the future, urging a decentralized framework of social recommendation.As a result, we design a federated learning recommender system for the social recommendation task, which is rather challenging because of its heterogeneity, personalization, and privacy protection requirements. To this end, we devise a novel frameworkFedratedSocial recommendation withGraph neural network (FeSoG). Firstly,FeSoGadopts relational attention and aggregation to handle heterogeneity. Secondly,FeSoGinfers user embeddings using local data to retain personalization. Last but not least, the proposed model employs pseudo-labeling techniques with item sampling to protect the privacy and enhance training. Extensive experiments on three real-world datasets justify the effectiveness ofFeSoGin completing social recommendation and privacy protection. We are the first work proposing a federated learning framework for social recommendation to the best of our knowledge.

Matrix factorization with denoising autoencoders for prediction of drug-target interactions.

Drug-target interaction is crucial in the discovery of new drugs. Computational methods can be used to identify new drug-target interactions at low costs and with reasonable accuracy. Recent studies pay more attention to machine-learning methods, ranging from matrix factorization to deep learning, in the DTI prediction. Since the interaction matrix is often extremely sparse, DTI prediction performance is significantly decreased with matrix factorization-based methods. Therefore, some matrix factorization methods utilize side information to address both the sparsity issue of the interaction matrix and the cold-start issue. By combining matrix factorization and autoencoders, we propose a hybrid DTI prediction model that simultaneously learn the hidden factors of drugs and targets from their side information and interaction matrix. The proposed method is composed of two steps: the pre-processing of the interaction matrix, and the hybrid model. We leverage the similarity matrices of both drugs and targets to address the sparsity problem of the interaction matrix. The comparison of our approach against other algorithms on the same reference datasets has shown good results regarding area under receiver operating characteristic curve and the area under precision-recall curve. More specifically, experimental results achieve high accuracy on golden standard datasets (e.g., Nuclear Receptors, GPCRs, Ion Channels, and Enzymes) when performed with five repetitions of tenfold cross-validation. Display graphical of the hybrid model of Matrix Factorization with Denoising Autoencoders with the help side information of drugs and targets for Prediction of Drug-Target Interactions.