Search

Blockchain, as an emerging technology, provides a brand-new idea for the improvement of personalized recommender systems with its characteristics of decentralization, tamper-proof, anonymity and traceability.Therefore, the main problems faced by the recommender systems in recent years and the opportunities brought by blockchain technology were summarized firstly.Then, literature analysis was adopted to analyze and summarize the research on the application of blockchain technology in recommender systems from four aspects: time distribution, literature types, research questions and evaluation indicators.The results show that the blockchain is of great significance for solving the problems of data security and privacy protection, data sharing, data trustworthiness and recommendation transparency of recommender systems.Existing studies mainly focus on solving the problem of data security and privacy protection in recommender systems, while further breakthroughs are needed in cross-platform data sharing, design of incentive mechanisms and system scalability.

Recommender systems are devoted to find and automatically recommend valuable information and services for users from massive data,which can effectively solve the information overload problem,and become an important information technology in the era of big data.However,the problems of data sparsity,cold start,and interpretability are still the key technical difficulties that limit the wide application of the recommender systems.Reinforcement learning is an interactive learning technique,which can dynamically model user preferences by interacting with users and obtaining feedback to capture their interest drift in real time,and can better solve the classical key issues faced by traditional recommender systems.Nowadays,reinforcement lear-ning has become a hot research topic in the field of recommendation systems.From the perspective of survey,this paper first analyzes the improvement ideas of reinforcement learning for recommender systems based on a brief review of recommender systems and reinforcement learning.Then,the paper makes a general overview and summary of reinforcement learning based recommender systems in recent years,and further summarizes the research situation of traditional reinforcement learning based recommendation and deep reinforcement learning based recommendation respectively.Furthermore,the paper summarizes the frontiers of reinforcement learning based recommendation research topic in recent years and its application.Finally,the future development trend and application of reinforcement learning in recommender systems are analyzed.

Recommendation system is one of the most common and basic applications in E-Commerce, search engines, etc. However, the user-item rating matrix data in the recommendation system is extremely sparse, resulting in poor recommendation results. This paper proposes a dual auto encoders matrix factorization (DAE-MF) model for the sparseness of the scoring matrix. This model uses a two-way parallel network structure, which combines a convolutional auto encoder and stacked denoising autoencoder and matrix factorization technology, fully combining the advantages of the two modules, effectively improving the model’s data completion capabilities. Through experiments on MovieLens and AIV datasets, it is verified that the DAE-MF method is superior to the existing models in rating prediction tasks. [ABSTRACT FROM AUTHOR]

Collaborative filtering is a widely used technique in current recommendation systems.It leverages the similarity between different users or items to retrieve interactive information between users and items and recommends new items for target users.In recent years,graph learning has gradually become an emerging recommendation paradigm due to its excellent perfor-mance and scalability in graph representation learning.This paper systematically reviews the most recent research on recommendation field from the perspective of graph learning.First,we provide a taxonomy that groups the current recommendation scenarios into two categories according to the data type used,including recommendation systems based on interactive information that leverage user-item interaction data as the main data source and auxiliary information-enhanced recommendation systems that incorporate social information associated with users and items as well as the knowledge graph information.Then,we review the main approaches,fundamental algorithms and critical difficulties of current recommendation models from the perspectives of random walk,graph representation learning and graph neural networks.Finally,we summarize the main challenges of graph learning methods in the field of recommendation system and outline the possible future research directions.

Knowledge graph (KG) provides a data structure to generate hybrid recommendations based on content and collaborative filtering. However, the existing recommendation methods based on knowledge graph take much less account of the user attribute information than the item attribute. To solve this problem, double end knowledge graph convolutional networks (DEKGCN) for recommender systems is proposed. In this algorithm, certain amount of sample of each entity’s neighborhood in the knowledge graph is taken as its high-order acceptance domain, and the related attributes of users in the dataset are taken as its first-order receptive field. Then, when calculating the representation of a given entity and a user, the neighborhood information is combined respectively, and finally the probability of user’s preference for items is obtained. It is an end-to-end framework that integrates multiple information of both user and item sides to learn the vector representation of users and items, which effectively solves the problem of data sparsity and cold start. Experimental results on real datasets show that DEKGCN has better recommendation quality than other baselines.

To address the issue of user preference drift and capture dynamic user preferences more accurately in sequence recommendation, this paper proposed a novel model named SILSSRec. The model initially leveraged categories and frequencies of items as side information to generate personalized user embeddings based on users' historical interaction sequences. Then it created personalized temporal interval embeddings by considering the time intervals between historical and current interactions, and integrated these embeddings with item feature embeddings to form personalized temporal embeddings. The model employed attention mechanisms and gated recurrent networks to extract users' long-term and short-term preferences from the embedding representations. Furthermore, it used contrastive learning to reinforce the feature representation of preferences, and an adaptive aggregation network dynamically combined these two types of preferences to form the final preference representation of users. Experiments on eight public datasets demonstrate that SILSSRec outperforms existing baseline models on evaluation metrics, with an average increase of 3.82% in AUC, 7.2% in recall rate, and 0.3% in precision. The results validate that SILSSRec performs well in various scenarios, effectively mitigating the preference drift issue and enhancing recommendation performance. [ABSTRACT FROM AUTHOR]

Heterogeneous graph embedding models based on non-contrastive learning (NCL) do not rely on negative sampling to learn the intrinsic features and patterns, which may cause the model fail to efficiently learn the differences between vertexes. This paper proposed a heterogeneous graph embedding model based on cross-view prototype non-contrastive learning (XPNCL), which learnt better node representations for downstream tasks by finding additional positive samples with more contextual information, and reconsidered the similarity between positive samples. The model firstly designed a tree structure based on random walks in heterogeneous graph. This directed filtering tree (DFT) about positive samples contained rich neighboring and semantic information by filtering out random walk paths that satisfied local structural constraints. Secondly, to achieve the alignment of similar samples in terms of numerical and quantitative from multiple dimensions, XP-NCL defined the cross-view prototype index (ISDR) and peak operator based on the characteristics of heterogeneous graphs. Furthermore, the model trained using stop-gradient updating. Finally, experiments verify the classification and clustering performance of the node on ACM, DBLP and freebase datasets, and the results show that even without the negative samples, the XP-NCL representation can achieve superior performance in many cases compared to other homogeneous and heterogeneous graph baselines. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Blockchain, as an emerging technology, provides a brand-new idea for the improvement of personalized recommender systems with its characteristics of decentralization, tamper-proof, anonymity and traceability. Therefore, the main problems faced by the recommender systems in recent years and the opportunities brought by blockchain technology were summarized firstly. Then, literature analysis was adopted to analyze and summarize the research on the application of blockchain technology in recommender systems from four aspects: time distribution, literature types, research questions and evaluation indicators. The results show that the blockchain is of great significance for solving the problems of data security and privacy protection, data sharing, data trustworthiness and recommendation transparency of recommender systems. Existing studies mainly focus on solving the problem of data security and privacy protection in recommender systems, while further breakthroughs are needed in cross-platform data sharing, design of incentive mechanisms and system scalability [ABSTRACT FROM AUTHOR]

Recommender systems provide important functions in areas such as dealing with data overload, providing personalized consulting services, and assisting clients in investment decisions. However, the cold start problem in recommender systems has always been in urgent need of solution and optimization. Based on this, this paper classifies the traditional methods and cutting- edge methods to solve the cold start problem, and expounds the research progress and excellent methods in recent years. Firstly, three traditional solutions to the cold start problem are summarized: recommendation based on content filtering, recommendation based on collaborative filtering, and hybrid recommendation. Secondly, the current cutting-edge recommendation algorithms to solve the cold start problem are summarized, and they are classified into the data-driven strategy and the method-driven strategy. The method-driven strategy is divided into algorithms based on meta-learning, algorithms based on context information and session strategy, algorithms based on random walk, algorithms based on heterogeneous graph information and attribute graph, and algorithms based on adversarial mechanism. According to the type of cold start problem, the algorithms are divided into two categories: new users and new items. Then, according to the particularity of the recommendation field, the cold start problem of the recommendation in the multimedia information field and the online e-commerce platform field is expounded. Finally, the possible research directions to solve the cold start problem in the future are summarized. [ABSTRACT FROM AUTHOR]

In order to alleviate the inherent information differences between different user social spaces and interest spaces in recommendation systems and the problem of ignoring high-order neighbors, this paper proposed AGCRSR. Firstly, the algorithm used a mapping matrix in the embedding layer to convert the initial feature vectors into adaptive embedding. Secondly, it intro- duced an attention mechanism to aggregate different aspects of user embeddings, and used a graph convolutional network to li- nearly learn the potential representations of users and items. Finally, it used the adaptive module to aggregate user representations and predict the final recommendation results of users for the project using the inner product function. This paper conducted comparative experiments on the LastFM and Ciao datasets and compared with other baseline algorithms. The experimental results show that the recommendation performance of the AGCRSR is significantly improved compared to other algorithms. [ABSTRACT FROM AUTHOR]

The knowledge graph in the recommendation system plays a vital role in the recommendation effect of the system, and the knowledge representation in the graph becomes a key factor affecting the recommendation system, which has become one of the current research hotspots. This paper proposed a node2vec-based knowledge representation node2 vec-side based on the traditional node2vec model by adding relational representation and diversifing wandering strategy to the structural characteristics of the knowledge graph in recommendation system, which combined with the knowledge graph network structure of recommendation system to explore the potential association relationship between nodes of large-scale recommendation entities, reduced the complexity of the representation and improved interpretability. After time complexity analysis, it could be seen that the proposed knowledge representation is lower than Trans series and RGCN in terms of complexity. Link prediction experiments were conducted on the traditional knowledge graph datasets FB15K, WN18, and recommendation domain datasets MovieLens-1M, Book-Crossing, Last. FM respectively. The experimental results show that on the MovieLens-1M dataset, hits @ 10 improves 5.5%12.1% and MRR improves 0. 090. 24, respectively. On the Book-Crossing dataset, hits @ 10 improves 3.5%-20.6%, and MRR improves 0.04-0.24 on average, respectively. And on the Last. FM dataset, hits@1 improves 0.3%-8.5% and MRR improves 0.04-0. 16 on average. It is better than the existing algorithms and verifies the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Graduate education is located at the top of national education, enrollment is the beginning of the top, which has always been the high ground for universities to compete for high 一 quality students. Recommended exemption students have absolute autonomy to choose, and universities may not hinder or restrict them for any reason or in any way. They can usually get the olive branch of multiple schools at the same time, and eventually choose the best, and go to the most desired school hall, but this will inevitably cause a lot of waste of the university s indicators. In addition to the limitations of discipline strength, geographical location, economic development and other problems, the number and quality of public health and preventive medicine postgraduate enrollment in western China are even less dominant. By reviewing the development context of graduate recommendation exemption system, this paper clarifies the current situation of the public health and preventive medicine promotion and exemption enrollment in western China, and formulates a series of effective recommended enrollment measures such as expanding publicity channels, improving the ratio of reexamination, and optimizing the selection process, so as to provide references for all enrollment units to improve the quantity and quality of the promotion and exemption enrollment. [ABSTRACT FROM AUTHOR]

The existing graph neural network(GNN) -based recommendation models have a natural limitation, namely the sparse supervision signal problem and coarse-grained relation modeling. These models can't identify user-item interaction relationships at a finer granularity, potentially leading to a decrease in their actual performance. Inspired by the successful application of contrastive learning in mining supervision signals, this paper proposed a KRMVC algorithm. The KRMVC framework constructed four graph views, including a global structural view, a local intention view, a coordination view, and a collaboration view. Additionally, the algorithm designed a new GNN information aggregation scheme to extract useful information about user intents and encoded it into the representations of users and items. KRMVC performed contrastive learning across the four views at both global and local levels, used a multi-task strategy to jointly optimize the recommendation supervision task and contrastive learning task for performance improvement. Experimental results demonstrate that the model achieves improvements in AUC values by 1.1% and 0.7%, F1 scores by 1.4% and 1.0%, and recall@K scores surpass state-of-the-art baselines on the MovieLens and Yelp2018 datasets. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Chongqing University of Posts & Telecommunications (Natural Science Edition) is the property of Chongqing University of Posts & Telecommunications and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Application Research of Computers / Jisuanji Yingyong Yanjiu is the property of Application Research of Computers Edition and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Multi-modal knowledge graphs (MMKG) integrate various modal information such as vision and text, presenting knowledge structures graphically. With the advancement of artificial intelligence, MMKG have played a significant role in recommendation systems, intelligent Q&A, and knowledge search among other fields. Compared to traditional knowledge graphs, MMKG can understand and present knowledge in multiple dimensions, possessing superior representation and application capabilities. To delve deep into the study of MMKG, this review first conducts a detailed analysis and elucidation of the value and categories of MMKG. Based on different construction methods, it compares and summarizes multimodal knowledge extraction, representation learning, entity alignment, and other aspects, categorizes multi-modal knowledge integration methods. It analyzes the progress in the applications of MMKG, discusses the limitations of MMKG, and proposes future research directions in the field of MMKG. [ABSTRACT FROM AUTHOR]

Based on matrix factorization techniques, this paper implemented a modified binomial matrix decomposition algorithm in order to solve the recommender system's rating prediction problem. It supposed that the user's rating of the item was based on the binomial distribution. There were differences in the user's rating habits, popularity of the items, it would result in an offset in the user-item scoring matrix. This paper used the maximum a posteriori estimate to design model and optimized the model by a stochastic gradient descent algorithm. The experimental results show that the modified binomial matrix decomposition algorithm is superior to the traditional binomial matrix decomposition algorithm in terms of recommender accuracy and offline calculation time on the MovieLens lOOK datasets. [ABSTRACT FROM AUTHOR]

To address the problem of insufficient recommendation performance in session recommendation, this paper proposed a disentangled graph neural network model (GDST-GNN) using global graph diffusion and spatio-temporal awareness. Specifically, the model firstly constructed a global collaborative graph based on all sessions from a global perspective, and then employed graph diffusion as the message propagation paradigm for global representation learning of items to capture global information be- yond the current session. For the representation learning of the current session, this paper designed a disentangled spatio-temporal gated network modeling the complex transition pattern and temporal dependency pattern of items in the session, and then fused the learned global and local representations factor by factor. In addition, this paper employed a self-supervised task to achieve performance enhancement of the model. Finally, it generated session representations through attention networks to achieve accurate recommendation of items. Extensive experiments on four real-world datasets validate the effectiveness of the proposed model. [ABSTRACT FROM AUTHOR]

Aiming at the problem that the existing social recommendation algorithms ignore the potential association of users and the collaborative relationship between items, this paper proposed a new algorithm model GATHGN. This model framework unified modeled of user association and item relationship. Firstly, users’ mined explicit social relations, potential association relations and user-item association relations, so as to extract users’ social high-order characteristics and potential interest high-order characteristics. Then, the above two high-level features were aggregated based on the graph attention mechanism, and the user fusion features were updated layer by layer. Finally, the final recommendation results were calculated according to the updated user fusion characteristics and project characteristics. Experimental results on Yelp and Flickr data show that the HR and NDCG of GATHGN are significantly improved compared with the baseline algorithm. [ABSTRACT FROM AUTHOR]

In the recommendation system, the user’s ratings are often affected by the rating context, that is, the user’s previous ratings of some items will affect the objectivity of his rating of the current item. Sparse linear method treats user ratings affected by context as the same as other ratings when calculating item similarity. However, this partial ratings cannot objectively reflect the similarity between items. To solve the above problems, this paper proposed a recommendation algorithm combining rating context and item similarity based on sparse linear method. It divided the algorithm into three stages.The first stage used weighted ratings to calculate the item’s nearest neighbor for feature selection.In the second stage, it used the rating error weight to reduce the fitting of the ratings affected by the context of the algorithm model, and trained the item similarity matrix. In the third stage, it predicted the ratings according to the user’s ratings and the item similarity, and finally sorted the predicted ratings to complete the item recommendation. Experiments were conducted on four datasets of MovieLens, it used mean average precision (MAP), mean reciprocal rank (MRR) and normalized discounted cumulative gain (NDCG) to evaluate the effectiveness of the algorithm. The experimental results show that the fusion rating context will further improve the accuracy of item similarity and thus improve the performance of recommendation. [ABSTRACT FROM AUTHOR]

To address the challenges of cold-start prediction and data sparsity for new users in recommendation systems, this paper proposed a collaborative similarity embedding recommendation model incorporating consistent social relationships（CSECSR）. Firstly, the model propagated and aggregated graph embeddings with equal weights through a warming-up layer. Secondly, it sampled social relationship neighbors with consistency and aggregated these relationships using attention mechanisms. Finally, the model predicted the rating by computing the inner product of the final embedded values of the user and the item, and optimized the model using a loss function that combined BPR loss with adaptive margin and similarity loss. Compared with other representative recommendation models on Ciao, Epinions and Film Trust datasets, the experimental results show that the proposed recommendation model outperforms other models significantly in terms of prediction error, which confirms the effectiveness of the proposed recommendation model. [ABSTRACT FROM AUTHOR]

Copyright of Systems Engineering & Electronics is the property of Journal of Systems Engineering & Electronics Editorial Department and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Harbin University of Science & Technology is the property of Journal of Harbin University of Science & Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In recommender systems, extracting higher order information about the graph using methods such as graph convolutional network alleviates the cold start problem. In order to fuse the deep feature extraction capability of neural network collaborative filtering on this basis, this paper proposed a two-channel collaborative filtering recommendation algorithm (GCNCF-2C) based on graph convolution. Firstly, it divided the recommendation problem into an upstream task and a downstream task. Secondly, to form a feature representation of nodes, in the upstream task, a pre-trained encoder extracted node features and higher-order features of the graph using a one-dimensional convolutional layer containing residuals and multiple graph convolutional layers in two independent channels. Finally, the decoder performed end-to-end training by node features for rating prediction. Experiments on the datasets MovieLens-100k and MovieLens-1M show that compared to the baseline models, the algorithm improves RMSE metrics by 1.72% and 1.76% on average and MAE metrics by 2.7% and 1.98% on average on the two datasets. Meanwhile, it improves RMSE metrics by 5.9% on average in the cold-start experiments based on users and projects, which shows overall performance. [ABSTRACT FROM AUTHOR]

Because heterogeneous information network（HIN） contains rich network structure and semantic information, recommendation systems often use HIN for recommendation. However, the current researches of recommender systems is mainly based on indirect information provided by meta-paths for recommendation, but these researches don’t make full use of direct interactive information. To make full use of this information, this paper proposed a ternary interaction model（AMMRec） that incorporated attention mechanisms and heterogeneous information network meta-paths. This method firstly used the implicit feedback matrix to construct user similarity matrix and item similarity matrix and used the representation learning method of the HIN to obtain the corresponding feature vector embeddings in HIN. Then it used the the attention mechanism to modify the embeddings and designed attention neural network to fuse representation vectors of different meta-paths. Finally it concatenated user embeddings and meta-path embeddings and item embeddings, and generated recommendation results through fully connected neural network. The experimental results on real datasets show that AMMRec improves the recommendation accuracy by up to 9.5%. In addition, AMMRec has good interpretability for the recommendation results. [ABSTRACT FROM AUTHOR]

In the practical application of recommendation systems, item popularity bias can be amplified by feedback loops, machine learning training models, and some external factors. It results in a phenomenon where a large number of long-tail items do not get a fair chance to be recommended. To address the fairness problem caused by the feedback loop amplifying the popularity bias, this paper conducted the first fairness analysis and enhancement study by means of a stochastic model checking method. It modeled the traditional recommendation system framework based on popularity bias and feedback loops as a DTMC and verified the fairness properties. The experiment revealed that as the number of feedback loop rounds increased, the Matthew effect intensified and fairness significantly diminished. This paper presented a fairness-enhanced recommendation system framework (FERSF) guided by stochastic model checking. It added a dynamic fairness threshold detection process to the feedback loop of the traditional framework to monitor the fairness. Also, it made a fairness-enhanced adjustment of the feedback influence factor to mitigate the impact of popularity bias on the system. The experimental analysis shows that the fairness of FERSF is significantly improved compared to the traditional recommendation system. Compared with the methods based on utility functions for fairness improvement, FERSF fundamentally inhibits the amplification of popularity bias due to the dynamic nature of the combined feedback loop. Compared with other algorithm-specific fairness improvements, FERSF is highly compatible because it is modeled based on the recommendation system framework. [ABSTRACT FROM AUTHOR]

From the output functional of recommender systems, the issue of the top-N was researched in this paper. The value of N affects the recommended quality and personalization. The feasibility and general methods of personalization of the output were studied in this paper. A new method based on aggregative rank was proposed. In the method, the entire sub-group is used as output of recommendation and the number of products in the sub-group is used as candidate value of N, in order to avoid big differences between the recommended products in the recommendation results and small differences between recommended products and un-recommended products. The general aggregative rank model of product was built in this paper. Meanwhile, aggregative rank quality was assessment based on different N, only in order to get a reliable value of N. The research results not only enrich the recommended system theoretical results, but also get a new road for personalization study of recommender systems. [ABSTRACT FROM AUTHOR]

The capture of user interest hidden in the user behavior sequence is a hot research direction of recommendation algorithms in recent years. The traditional sequence prediction model uses the last product clicked by the user as the target, and establishes the association between user behavior and the target product, but does not fully dig out the sequence relationship between user sequences. This study improves on the traditional DIN model, uses continuous behavior over a period of time as the target vector, uses the transformer structure to complete the sequence-to-sequence prediction task, and further extracts and utilizes the user's deep interest in the user behavior sequence, and it is recommended in conjunction with DIN as an auxiliary feature. The experimental results on the Amazon book and the electronic data sets show that the DIN-based hybrid recommendation model proposed in this study increases the AUC index of the original DIN model by about 0.7% and 1.9%, respectively. It can be seen that the hybrid recommendation based on user behavior sequence prediction can play a certain auxiliary role in the multi-feature recommendation system. In addition, the influence of user sequence length on the model results is also explored. [ABSTRACT FROM AUTHOR]

Most sequential recommendation systems（SRSs） assume that the next item to be predicted is related to the user’s previous input. However, in real scenarios, users may click items inconsistent with their own interests and preferences（unreliable instances） by mistake during browsing. This paper proposes a sequential recommendation algorithm based on Gaussian distribution modeling to solve this problem. Firstly, the algorithm extracts the input sequence patterns by reducing the uncertainty of input items through an uncertainty aware graph ensemble network（UAN） with multiple heads of self-attention; Secondly, the extracted input sequence pattern is modeled as a Gaussian distribution, and obtain the dynamic user’s preferences and the uncertainty of preferences in the sequence information; Then, extend the traditional recommended objective function to a sampling loss function and an uncertainty regularizer, and give each training instance appropriate uncertainty; Finally, remove the unreliable examples with high loss and low uncertainty to enhance the accuracy of sequence recommendation. The algorithm is tested on three open datasets, Book Cross, Movielens-1 M and Steam. The results show that the algorithm has achieved an improvement of about 5.3% compared with the baseline with good effects, and has obtained better sequential recommendation results. The proposed algorithm can improve the recommendation accuracy by effectively reducing the uncertainty of input sequence information. [ABSTRACT FROM AUTHOR]

To address the problem of poor performance in graph neural network based recommendation systems due to the data sparse problem, this paper proposed a light relational graph attention recommendation model (LightRGAN) that fusing with profiles and text information. First, this model used user profiles and item profiles to initialize the embeddings of users and the items. Secondly, it introduced comments, descriptions and categories as auxiliary information. Meanwhile, it mined the rela-tionships between the text in the comment set and the description set of the same user through the text embedding network based on the multi-head attention mechanism. Then, it learned the embedding representations of the users and the items through a light relational graph convolutional network fused with the attention mechanism. Finally, it weighted sum the embed-ding representations at each layer and inputted them to the prediction network to calculated match score between the users and the items. The results on three public datasets show that the LightRGAN outperforms other baseline models. The evaluation metrics HR@ 20 and NDCG@20 improved at least 2.58% and 2.37% than the optimal baseline model. [ABSTRACT FROM AUTHOR]

In the field of knowledge-aware recommendation, embedding propagation methods can mine the structured information of the knowledge graph and also can obtained the semantic information between n-hop entities. However, with the increase of the spread, the embedding propagation method will lose part of the structure information, the user representation obtained in this way is coarse-grained and cannot adequately characterize the user's preferences. Aiming at these problems, this paper proposed an user preference network of fusing knowledge-aware and time-aware (FKTUPN). Firstly, FKTUPN generated the first-order embedding of the knowledge graph in the embedding-layer, and separately modeled the fine-grained user prefe-rence. Then it divided the time slot by random walk and performed temporal context encoding to get contextual information of KG. Subsequently, the contribution of the representation was distinguished and propagated through the knowledge-aware attention mechanism. Finally, it aggregated users and items of different propagation levels in the prediction layer, and obtained the recommendation result after multiplying these two representations. Experiments show that result on three datasets: Amazon-book, Last-FM and Yelp is superior to the baseline model in terms of recall@20 and NDCG@20, fine-grained user preference representation and full utilization of structured information can effectively improve the accuracy of recommendations. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The rapid development of information technology has led to information overload.Recommendation is one of 他the most effective ways to solve the information overload.In recent years,the rapid development of deep learning has also led to the advancement of recommender systems,and various deep learning based recommendation algorithms have emerged one after another.However,due to the large number of candidate items and the dynamic evolving of user interests,deep learning based recommendation algorithms suffer from computational burden of online recommendation.Therefore,it is almost impossible for these algorithms to be deployed alone in practice.With the development of deep learning based recommendation,the item recallingtechniques(also called approximated search techniques) has also made significant progress.This paper first introduces the research progress of the item recalling techniques based on the nearest neighbor search,and then discusses the research progress of the item recallingtechniquesbased on the maximum inner product search from the perspectives of indexing,locality sensitive hash,learning to hash and vector quantization. [ABSTRACT FROM AUTHOR]

Although collaborative filtering algorithms become widely applicable, they also suffer from the difficulties of scalability and extensibility. To address this problem, this paper proposed a collaborative filtering model for recommendation systems based on the classification of user complex network features（UCNFC）. Firstly, the method selects feature users based on degree values in the user set and establishes similarity thresholds to implement groups of non-feature users. Then, user-user similarity constructs the network and completes the community labeling in the network by K-core decomposition. Finally, the target users choose their neighbors within themselves group to accomplish movie rating prediction. Based on the experimental results of Movie Lens and Netflix datasets, the algorithm shows improved both temporal and spatial performance compared with the classical collaborative filtering algorithm, demonstrating more excellent scalability and extensibility. [ABSTRACT FROM AUTHOR]

Traditional collaborative filtering methods largely model in a uniform way without considering the modeling relationships at the fine-grained level of interaction intent. This paper proposes a graph neural network collaborative filtering algorithm fused with interaction intention, named INTNGCF. Firstly, the potential intention is identified by decomposing edges into multiple potential spaces; Secondly, the interaction intention fusion layer is used to determine the importance of these potential intentions. Finally, a predicted user rating of the project is generated. Compared with seven baseline models on three real datasets, the results show that the proposed algorithm has certain performance advantages compared with other advanced recommendation models. [ABSTRACT FROM AUTHOR]

Existing sequential recommendation methods based on multi-interest frameworks only learn the multi-interest representations from users' recent interaction sequences, ignoring the association information between items in the dataset. To address this problem, this paper proposed an item associations aware multi-interest sequential recommendation method (IAMIRec). Firstly, this method obtained the item association set and matrix by calculating the user' s interaction sequence in the dataset, then modeled the user's recent interaction sequence by the association martix and the multi-head self-attention mechanism, and used the multi-interest framework to model the user's multiple interest vectors. It used the obtained multiple interest vectors to the top-N recommendation for users. This paper tested and analyzed the model on three datasets, IAMIRec outperformed related methods on recall, NDCG, and hit rate metrics. The results show that IAM!Rec can achieve better recommendation performance, and also show that it can effectively enhance the multi-interest representation of users through item associations information. [ABSTRACT FROM AUTHOR]

In order to further explore ways to improve the accuracy of the clothing recommendation system and improve users' shopping experience, relevant literature on the current clothing recommendation system was analyzed. The development status and trend of clothing recommendation were sorted out from three aspects: the process of clothing recommendation, personalized recommendation and the application of clothing recommendation. The relevant technologies and their respective characteristics of clothing image target recognition and segmentation, clothing feature extraction and recommendation algorithm model involved in clothing recommendation were expounded in this paper, the research status of personalized clothing recommendation, and the application of clothing recommendation system were summarized, and it was pointed out that the future recommendation system will pay more attention to the combination of image and text, intelligence and individuation, so as to provide more targeted and accurate recommendations to meet the diverse needs of users. [ABSTRACT FROM AUTHOR]

The recommender system uses knowledge graph can solve the problems of data sparsity and cold start, but it often ignores the importance of high-order collaborative information and different collaborative information to explore users' potential preferences. Therefore, this paper proposed a recommendation model combining enhanced collaborative information and knowledge graph information ( CIKG ) . This model first used the historical interactive data of users and items to obtain first-order collaboration information and high-order collaboration information. At the same time, it used the attention mechanism to capture important information and obtain enhanced collaboration information to supplement the feature representation of users and items. Secondly, this model matched interacted items with the entities in the knowledge graph and performed the propagation in the knowledge graph. It could obtain knowledge graph information and used the information to obtain the user' s preferences and enhance the interpretability of the model. Finally, the aggregator combined the enhanced collaboration information and knowledge graph information to obtain the final representation of users and items, so as to make predictions. The experimental results on Last-fm and Book-crossing datasets show that CIKG has a great improvement over other comparative models. [ABSTRACT FROM AUTHOR]

Due to the numerous but scattered studies on the bias of recommendation systems and few relevant research reviews, according to the subject that causes the bias, this paper divides bias into the bias in user behavior, the bias in item display, and the bias in recommendation algorithm, and then subdivides it into nine specific categories, such as selection bias and conformity bias, and gives the definition and causes of each bias; This paper focuses on sorting out the research on correlation recommendation methods and causal recommendation methods to solve the bias problem, and further classifies them according to the technical means or the first, middle and last three stages of the training model, and analyzes the differences between different types of research; This paper summarizes common datasets and reviewers of bias studies and forecasts the development trend of bias-oriented recommendation methods from different categories, and hopes to provide some help for the follow-up bias research in recommender system. [ABSTRACT FROM AUTHOR]

Applying knowledge graph to recommendation system can make use of semantic relations between entities of knowledge graph to learn user and item representation. The embedding propagation method uses the graph structure of the knowledge graph to learn relevant features, but the semantic dependency between multi-hop entities decreases as the propagation range increases. In order to effectively improve the semantic expression ability of recommendation and improve the accuracy of recommendation, this paper proposes a knowledge-aware propagation recommendation algorithm based on users’ potential interests. The model adopts heterogeneous propagation method to disseminate item relevant knowledge and iteratively learn users’ potential interests, so as to enhance the representation ability of the model to users and items. Specifically, first, graph embedding layer generate initialize representation of users and items, and in the heterogeneous propagation layer, the knowledge-aware attention mechanism can distinguish the importance of entities in the same layer, so the model can generate the representation of target entities more accurately. Then the user’s potential interest propagation can effectively learn the user’s higher-order potential interest and enhance the semantic relevance of multi-hop entities. Finally, information decay factor is used in the prediction layer to distinguish the importance of different communication levels and generate the final representation of users and items. Experiments show that the AUC value of the model on the Last. FM and Book-Crossing increases by 2.25% and 4.71% compared with the most advanced baseline, and the F1 value increases by 3.05% and 1.20% respectively, and the Recall@K value is superior to the comparison baseline model. The proposed model can effectively improve the accuracy of recommendation. [ABSTRACT FROM AUTHOR]

In order to summarize previous works and provide ideas for related researchers. This paper first discussed the current basic concepts of the multi-modal knowledge graph. Then it introduced the construction of multi-modal knowledge graphs from two perspectives of graph database and knowledge graph, and summarized the ideas of two main methods. It also analyzed the key technologies and related works in the construction and application, such as multi-modal information extraction, representation learning, entity linking. In addition, it enumerated the application of multi-modal knowledge graphs in four scena-rios, including recommender systems, cross-modal retrieval, human-computer interaction and cross-modal data management. Finally, it looked forward to the development prospects of multi-modal knowledge graphs from four aspects. [ABSTRACT FROM AUTHOR]

Copyright of Geomatics & Information Science of Wuhan University is the property of Geomatics & Information Science of Wuhan University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Jilin University (Science Edition) / Jilin Daxue Xuebao (Lixue Ban) is the property of Zhongguo Xue shu qi Kan (Guang Pan Ban) Dian zi Za zhi She and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In the Open Source Community, it’s essential to find and recommend suitable participants for newly initiated issues in order to solved the issues and develop the community. This paper proposed to construct a two-layer Graph Attention Network Participant Recommendation Model (GAT-UCG) based on the cooperative relationship records and the historical participated issues records of the developers. The method used to construct the model is obtaining the information of the problem participants and the interaction information of the developers first, and building the developer problem participation graph and the developer collaboration relationship graph respectively, then redistributing the weights to the edges through the attention mechanism, finally, figuring the Top-N recommendation of the problem participants according to the Issue node embedding representation obtained by the output layer. There are 7, 352 issues from popular Github repositories for experiments. The results showed that the GAT-UCG model outperforms the baseline method in three indicators: recommendation accuracy, recall, and F-Score. [ABSTRACT FROM AUTHOR]

The imputation-based solution can alleviate the sparsity problem of recommendation system. Improved Jaccard similarity based on logistic user habit rating weighting, which is called the HabJac. By combining this metric, the K nearest neighbor ( KNN) obtains the imputation value. Furthermore, this paper proposed a new HISVD recommendation algorithm by combining the regularized singular value decomposition ( RSVD) technology to predict the unknown ratings. Firstly, the user' s habit rating was the most frequent one. Secondly, logistic weight was positively correlated with Euclidean distance between rating and habit one . The experimental results on four real data sets show that: a ) the optimal parameters of HISVD algorithm for different data sets are similar; b) HISVD surpasses the mainstream competitors for the MAE and RSME. [ABSTRACT FROM AUTHOR]

With the rapid development of the Internet, the recommendation system can handle the problem of information overload. Due to many problems in traditional recommendation systems, they can' t handle the discovery of hidden information. This paper proposed an adaptive graph convolution attention neural collaborative filtering recommendation model ( ANGCACF) . Firstly, graph convolution neural network obtained the user-item interaction figure and adaptively aggregated the user-item feature information. Secondly, the model added adaptive extended data to solve the data sparsity for the user-item feature information and used the attention mechanism to redistribute the weight to the user-item feature information and the adaptive extended data. Finally, it obtained the final recommendation result by using the algorithm framework of collaborative filtering based on matrix decomposition. Experiments on MovieLens-lM, MovieLens-lOOK and Amazon-book show that the algorithm is superior to the baseline method in five indexes : precision, recall, Mrr, hit and NDCG. [ABSTRACT FROM AUTHOR]

In recent years, many researchers take advantage of graph convolution network in multi-behavior recommendation to further alleviate the data sparsity problem. However, most of current works directly use graph convolution network, which makes time complexity of the model relatively high. These works also ignore the different weights of neighbors and the different contributions of each behavior to user' s preference . Therefore, this paper proposed a multi-behavior recommendation model based on light graph convolution and enhanced attention ( MB-LGCA) . Firstly, the model constructed a user-item bipartite graph according to the multi-behavior data, and used a light graph convolution network to aggregate the features of neighbors to obtain high-order collaborative information. At the same time, it used attention mechanism to integrate the neighbors' weights to enhance embedding representations of nodes . It used the k-order user 's embedding propagation to obtain the different importance of each behavior to user ' s preference, so that the model had better interpretability. Finally, it combined embedding representations of different layers for prediction. The experimental results on two real datasets show that the model has better performance. [ABSTRACT FROM AUTHOR]

As an effective representation of knowledge, knowledge graph network can be used to represent rich factual information between different categories and become an effective knowledge management tool. It has achieved great results in the application and research of knowledge engineering and artificial intelligence. Knowledge graph is usually expressed as a complex network structure. Its unstructured characteristics make the application of graph neural network to the analysis and research of knowledge graph become a research hotspot in academia. The purpose of this paper is to provide extensive research on knowledge graph construction technology based on graph neural network to solve two types of knowledge graph construction tasks, including knowledge extraction (entity, relationship and attribute extraction) and knowledge merging and processing (link prediction, entity alignment and knowledge reasoning, etc.). Through these tasks, the structure of knowledge graph can be further improved and new knowledge and reasoning relationships can be discovered. This paper also studies the advanced graph neural network method for knowledge graph related applications, such as recommendation system, question answering system and computer vision. Finally, the future research directions of knowledge graph application based on graph neural network are proposed. [ABSTRACT FROM AUTHOR]

With the massive quantification of networked data and the development of geo-social networks (GSNs), group activities are prevalent in people's life. The objects of recommendation systems are extended from individuals to user groups. Point-of-interest (POI) group recommendation problem is also gradually known as a hot research topic. However, the traditional methods are not suitable for group recommendation in geographic social networks, due to the multifactorial influence of user preferences in GSNs and the complexity of the group decisionmaking process. To reveal user preferences and the effect of the group decision process on group recommendation, this paper proposes a neural network-based model for dynamic POI group recommendation by leveraging multidimensional user preference. Firstly, the proposed model combines temporal and spatial factors to calculate user preferences based on user behavior activity records and builds a group-point-of-interest perception graph with group as unit. Next, this paper adds the influence of collaborative users to model group preferences, which fully considers the characteristics of GSNs, to ensure the accuracy of POI group recommendation. Finally, a neural network-based model can be constructed to simulate group decision-making, which can ensure the accuracy of POI recommendations. This paper conducts extensive experiments by comparing the existing group recommendation algorithms on the real datasets to demonstrate the performance of the method proposed in this paper. Experimental results show that the proposed method is significantly better than the existing algorithms in terms of the hit rate of POI, which proves the effectiveness of the proposed algorithm. [ABSTRACT FROM AUTHOR]