Social recommendations assume that users with social networks tend to have similar pReferences and leverage the social network of users to improve personalized recommendations. However, the scarcity of interactive and social data, along with the presence of irrelevant or fake social connections, poses challenges in accurately predicting user preferences. Recent research has leveraged diffusion models to eliminate invalid social connections from the social relation graph, but this approach incurs high resource costs for large-scale item prediction. To address these issues, we propose an efficient latent space diffusion model for social recommendation named latent diffusion method for social recommendation (LDSR), which can reduce resource costs by clustering user social relationships and performing diffusion in a low-dimensional space. During the diffusion process, we inject and eliminate Gaussian noise and residuals in multiple steps, enhancing the model’s ability to recognize noise while ensuring output diversity and determinism. Additionally, we design a reconstruction strategy to capture latent social relationships, which helps to densify the social relation graph. The nonsmooth nature of the latent space can disrupt downstream task outputs, so we introduce variation constraints to smooth the latent space, reducing the impact of latent perturbations during generation. Furthermore, we incorporate user-item collaborative information to guide the reverse process, enhancing the controllability of the generated content to provide reasonable denoising. Extensive experiments on four publicly available datasets demonstrate that LDSR outperforms the state-of-the-art models, exhibiting superior training efficiency, robustness against sparsity and noise, and enhanced interpretability.

Knowledge-aware recommendations improve performance by using knowledge graphs as auxiliary information. Recently, researchers have introduced the contrastive learning paradigm in knowledge-aware recommendations to enhance representation learning. However, most contrastive learning methods rely on manually or randomly generated knowledge views, making it challenging to generalize to different data distributions and alleviate knowledge noise effects. To solve these issues, we propose a mask diffusion-based contrastive learning method for knowledge-aware recommendation. Specifically, we apply local masked input to the diffusion model, using a mask prediction paradigm to adaptively generate views from both global and local perspectives, thereby enhancing the model's generalization capability across different data distributions. Additionally, we propose a conditional inference process, leveraging user intentions to provide reasonable denoising guidance. At the same time, we design a collaborative knowledge diffusion loss aimed at improving the consistency between generated data and user behavior patterns. In this way, we combine the diffusion model with contrastive learning for the knowledge-aware recommendation, which can improve the generalization ability of the model. Our experimental results on four datasets show the effectiveness of our model. The implementation code is available athttps://github.com/haomiaocqut/ReSys_KMDCL.

The current sequential recommendation systems mainly focus on mining information related to users to make personalized recommendations. However, there are two subjects in the user historical interaction sequence: users and items. We believe that mining sequence information only from the users' perspective is limited, ignoring effective information from the perspective of items, which is not conducive to alleviating the data sparsity problem. To explore potential links between items and use them for recommendation, we propose Intent-guided Bilateral Long and Short-Term Information Mining with Contrastive Learning for Sequential Recommendation (IBLSRec), which interpretively integrates three kinds of information mined from the sequence: user preferences, user intentions, and potential relationships between items. Specifically, we model the potential relationships between interactive items from a long-term and short-term perspective. The short-term relationship between items is regarded as noise; the long-term relationship between items is regarded as a stable common relationship and integrated with the user's personalized preferences. In addition, user intent is used to guide the modeling of user preferences to refine the representation of user preferences further. A large number of experiments on four real data sets validate the superiority of our model.

Conversational recommender systems provide users with item recommendations via interactive dialogues. Existing methods using graph neural networks have been proven to be an adequate representation of the learning framework for knowledge graphs. However, the knowledge graph involved in the dialogue context is vast and noisy, especially the noise graph nodes, which restrict the primary node’s aggregation to neighbor nodes. In addition, although the recurrent neural network can encode the local structure of word sequences in a dialogue context, it may still be challenging to remember long-term dependencies. To tackle these problems, we propose a sparse multi-hop conversational recommender model named SMCR, which accurately identifies important edges through matching items, thus reducing the computational complexity of sparse graphs. Specifically, we design a multi-hop attention network to encode dialogue context, which can quickly encode the long dialogue sequences to capture the long-term dependencies. Furthermore, we utilize a variational auto-encoder to learn topic information for capturing syntactic dependencies. Extensive experiments on the travel dialogue dataset show significant improvements in our proposed model over the state-of-the-art methods in evaluating recommendation and dialogue generation.

Knowledge graph (KG) is of growing significance in enabling explainable recommendations. Recent research works involve constructing propagation-based recommendation models. Nevertheless, most of the current propagation-based recommendation methods cannot explicitly handle the diverse relations of items, resulting in the inability to model the underlying hierarchies and diverse relations, and it is difficult to capture the high-order collaborative information of items to learn premium representation. To address these issues, we leverage hyperbolic dynamic neural networks for knowledge-aware recommendation (KHDNN). Technically speaking, we embed users and items (forming user–item bipartite graphs), along with entities and relations (constituting KGs), into hyperbolic space, followed by encoding these embeddings using an encoder. The encoded embedding is passed through a hyperbolic dynamic filter to explicitly handle relations and model different relational structures. Furthermore, we design a fresh aggregation strategy based on relations to propagate and capture higher-order collaborative signals as well as knowledge associations. Meanwhile, we extract semantic information via a bilateral memory network to fuse item collaborative signals and knowledge associations. Empirical results from four datasets show that KHDNN surpasses cutting-edge baseline methods. Additionally, we demonstrate that the KHDNN can perform knowledge-aware recommendations with complex relations.

Most rumor detection methods extract the features of rumor through two aspects of text semantics and propagation structure to achieve automatic rumor classification, while most of the existing methods do not realize that false and irrelevant interactions in the propagation structure will reduce the accuracy of rumor detection. In addition, most of the existing rumor detection methods failed to effectively extract key clues from the comments of social network users. In response to these phenomena, this article proposes a social network rumor detection method combining a dual attention mechanism and graph convolutional network (GCN) (dual-attention GCN, DA-GCN). First, build an event propagation graph; then, the GCN is used to extract the propagation structure information of each event-related microblog (tweet), and the attention mechanism is combined to suppress the false and irrelevant interactive relationships. Therefore, the anti-interference propagation structure features are extracted from the propagation graph. Second, to fully utilize the clues in users’ comments, this article makes use of the attention mechanism to fuse source microblog (tweet) with the comment–retweet information and extract interactive semantic features from it. Finally, the above two features are fused to generate a new event representation. Experimental results show that the proposed DA-GCN has an accuracy of 94.4%, 90.5%, and 90.2% on the Weibo dataset, the Twitter15 dataset, and the Twitter16 dataset, respectively, and has achieved excellent performance in the early rumor detection task, which proves that the proposed method is reasonable and effective.

Popularity bias is a massive challenge for autoencoder-based models, which decreases the level of personalization and hurts the fairness of recommendations. User reviews reflect their preferences and help mitigate bias or unfairness in the recommendation. However, most existing works typically incorporate user (item) reviews into a long document and then use the same module to process the document in parallel. Actually, the set of user reviews is completely different from the set of item reviews. User reviews are heterogeneous in that they reflect a variety of items purchased by users, while item reviews are only related to the item itself and are thus typically homogeneous. In this article, a novel asymmetric attention network fused with autoencoders is proposed, which jointly learns representations from the user and item reviews and implicit feedback to perform recommendations. Specifically, we design an asymmetric attentive module to capture rich representations from user and item reviews, respectively, which solves data sparsity and explainable problems. Furthermore, to further address popularity bias, we apply a noise-contrastive estimation objective to learn high-quality “de-popularity” embedding via the decoder structure. A series of extensive experiments are conducted on four benchmark datasets to show that leveraging user review information can eliminate popularity bias and improve performance compared to various state-of-the-art recommendation techniques.

The next point-of-interest (POI) recommendation is indispensable in enhancing the richness of users’ lives and helping service providers achieve more economic earnings. Recurrent Neural Network (RNN) based methods are remarkable in learning users’ long-term or short-term behavioral dependencies. However, existing RNN-based methods lack sufficient interaction with their contexts, and at the same time, ignore the importance of non-consecutive POIs with different degrees for understanding users’ behaviors. In order to solve these problems, we propose a novel Spatio-Temporal model based on mogrifier LSTM and attention network (named STMLA) for next POI recommendation. The STMLA model builds a parallel structure to process the users’ check-in sequences through the mogrifier LSTM and the multi-head attention network, which can achieve better contextual interaction while selectively considering nonconsecutive factors with different degrees of significance. Our STMLA algorithm explicitly integrates temporal and spatial information to capture users’ long-term and short-term preferences, incorporating spatial information to build the Location-Saltant algorithm. Through extensive experiments on several real-world datasets, we demonstrate that our model outperforms the existing state-of-the-art methods in the next POI recommendation task.

The autoencoder network has been proven to be one of the powerful techniques for recommender systems. Currently, the ways of utilizing autoencoder in recommender systems can be divided into two categories: modeling user-item interaction rely solely on autoencoder and integrating autoencoder with other models. Most existing methods based on autoencoder assume that all features of model’s input are equally the same contributing to the final prediction, which can be regarded as attention weight vectors; however, this hypothesis is not reliable, especially when exploring users’ interaction frequency with different items. Moreover, combining autoencoder with traditional methods, the usual strategy is to leverage a linear kernel of the inner product of user and item vectors to predict user preferences, which will lead to insufficient expression power and hurt the performance of recommendation when facing data sparsity and cold start problems. To tackle the above two problems, we propose a novel hybrid deep learning model for top-n recommendation, called attentive stacked sparse autoencoder (A-SAERec), which can capture attention weights vector of a user for items, and then combined with the neural matrix factorization to improve the performance of recommender model. Extensive experiments on four real-world datasets show that our A-SAERec algorithm has significant improvements over state-of-the-art algorithms.

Recommender systems are used to make recommendations about products, information, or services for users. Most existing recommender systems implicitly assume one particular type of user behavior. However, other recommender system utilizes different particular type information by combining different techniques to improve the quality of the recommendation. This paper proposed a novel personalized recommendation method that utilizes semi-supervised clustering based Gaussian mixture model, which provides a hybrid recommender method by combining demographic method and user-based collaborative filtering method. The result from various simulations using MovieLens data set shows that the proposed recommender method performs better and helps to improve the quality of recommendation rating.

Current existing representative works to semi-supervised incremental learning prefer to select unlabeled instances predicted with high confidence for model retraining. However, this strategy may degrade the classification performance rather than improve it, because relying on high confidence for data selection can lead to an erroneous estimate to the true distribution, especially when the confidence annotator is highly correlated with the confidence annotator. In this paper, a new semi-supervised incremental learning algorithm was proposed, which selected the high confidence unlabeled instances with symmetrical distribution from unlabeled data, it can reduce the bias in the estimation in some degree. In detail, expectation maximization algorithm was used to estimate the confidence of each instance, and Gaussian function was used to calculate the data distribution, then the selected unlabeled data was used for retraining model with classifier algorithm. The experimental results based on a large number of UCI data sets show that our algorithm can effectively exploit unlabeled data to enhance the learning performance.