Group recommendation based on hybrid trust metric

Sequential recommendation and group recommendation are two important branches in the field of recommender system. While considerable efforts have been devoted to these two branches in an independent way, we combine them by proposing the novel sequential group recommendation problem which enables modeling group dynamic representations and is crucial for achieving better group recommendation performance. The major challenge of the problem is how to effectively learn dynamic group representations based on the sequential user-item interactions of group members in the past time frames. To address this, we devise a Group-aware Long- and Short-term Graph Representation Learning approach, namely GLS-GRL, for sequential group recommendation. Specifically, for a target group, we construct a group-aware long-term graph to capture user-item interactions and item-item co-occurrence in the whole history, and a group-aware short-term graph to contain the same information regarding only the current time frame. Based on the graphs, GLS-GRL performs graph representation learning to obtain long-term and short-term user representations, and further adaptively fuse them to gain integrated user representations. Finally, group representations are obtained by a constrained user-interacted attention mechanism which encodes the correlations between group members. Comprehensive experiments demonstrate that GLS-GRL achieves better performance than several strong alternatives coming from sequential recommendation and group recommendation methods, validating the effectiveness of the core components in GLS-GRL.

Group Recommendation (GR) is the task of suggesting relevant items/events for a group of users in online systems, whose major challenge is to aggregate the preferences of group members to infer the decision of a group. Prior group recommendation methods applied predefined static strategies for preference aggregation. However, these static strategies are insufficient to model the complicated decision making process of a group, especially for occasional groups which are formed adhoc. Compared to conventional individual recommendation task, GR is rather dynamic and each group member may contribute differently to the final group decision. Recent works argue that group members should have non-uniform weights in forming the decision of a group, and try to utilize a standard attention mechanism to aggregate the preferences of group members, but they do not model the interaction behavior among group members, and the decision making process is largely unexplored.In this work, we study GR in a more general scenario, that is Occasional Group Recommendation (OGR), and focus on solving the preference aggregation problem and the data sparsity issue of group-item interactions. Instead of exploring new heuristic or vanilla attention-based mechanism, we propose a new social self-attention based aggregation strategy by directly modeling the interactions among group members, namely Group Self-Attention (GroupSA). In GroupSA, we treat the group decision making process as multiple voting processes, and develop a stacked social self-attention network to simulate how a group consensus is reached. To overcome the data sparsity issue, we resort to the relatively abundant user-item and user-user interaction data, and enhance the representation of users by two types of aggregation methods. In the training process, we further propose a joint training method to learn the user/item embeddings in the group-item recommendation task and the user-item recommendation task simultaneously. Finally, we conduct extensive experiments on two real-world datasets. The experimental results demonstrate the superiority of our proposed GroupSA method compared to several state-of-the-art methods in terms of HR and NDCG.

We propose a hybrid trust model for enhancing security in distributed applications by combining the and trust relationships. The motivation is to compensate for the drawbacks associated with individual hard trust and soft trust models. We develop a formal hybrid trust model for specifying the hybrid trust relationships and associated operations. The new model provides a flexible way for specifying the security relevant trust requirements which compensate the drawbacks of the individual trust models while preserving their advantages. We provide application guidelines on how the hybrid trust relationships can be applied to achieve mutual improvements between the individual models

KTPGN: Novel event-based group recommendation method considering implicit social trust and knowledge propagation

Recommending events to groups is different from to single-user in event-based social networks (EBSN), which involves various complex factors. Generally, group recommendation methods are either based on recommendation fusion or model fusion. However, most existing methods neglect the fact that user preferences change over time. Moreover, they believe that the weights of different factors that affect group decision-making are fixed in different periods. Recently, there are a few works using the attention mechanism for group recommendation. Although they take into account the dynamic variability of user preferences and the dynamic adjustment of user features weights, they haven’t discussed more features of groups and events affecting group decision-making. To this end, we propose a novel Feature-level Hierarchical Attention Network (FHAN) for group event recommendation for EBSN. Specifically, group decision-making factors are divided into group-feature factors and event-feature factors, which are integrated into a two-layer attention network. The first attention layer is constructed to learn the influence weights of words of group topics and event topics, which generates better thematic features. The second attention layer is built to learn the weights of group-feature factors and event-feature factors affecting group decision-making, which results in better comprehensive representation of groups and events. All influence weights of different features in the model can be dynamically adjusted over time. Finally, we evaluate the suggested model on three real-world datasets. Extensive experimental results show that FHAN outperforms the state-of-the-art approaches.

A POI group recommendation method in location-based social networks based on user influence

Group recommendation has attracted significant research efforts for its importance in benefiting a group of users. This paper investigates the Group Recommendation problem from a novel aspect, which tries to maximize the satisfaction of each group member while minimizing the unfairness between them. In this work, we present several semantics of the individual utility and propose two concepts of social welfare and fairness for modeling the overall utilities and the balance between group members. We formulate the problem as a multiple objective optimization problem and show that it is NP-Hard in different semantics. Given the multiple-objective nature of fairness-aware group recommendation problem, we provide an optimization framework for fairness-aware group recommendation from the perspective of Pareto Efficiency. We conduct extensive experiments on real-world datasets and evaluate our algorithm in terms of standard accuracy metrics. The results indicate that our algorithm achieves superior performances and considering fairness in group recommendation can enhance the recommendation accuracy.

We present PolyLens, a new collaborative filtering recommender system designed to recommend items for groups of users, rather than for individuals. A group recommender is more appropriate and useful for domains in which several people participate in a single activity, as is often the case with movies and restaurants. We present an analysis of the primary design issues for group recommenders, including questions about the nature of groups, the rights of group members, social value functions for groups, and interfaces for displaying group recommendations. We then report on our PolyLens prototype and the lessons we learned from usage logs and surveys from a nine-month trial that included 819 users. We found that users not only valued group recommendations, but were willing to yield some privacy to get the benefits of group recommendations. Users valued an extension to the group recommender system that enabled them to invite non-members to participate, via email.

Due to the emergence of a large number of new nodes in peer-to-peer network, the trust matrix is sparse and data is insufficient. Therefore, global trusts of peers are inaccurate which are computed by trust matrix iteration and the success rates of transactions become low. PSRTrust, a mixed trust model combining global trusts and local trusts based on preference similarities, is proposed to restore sparse trust matrix by Similarity Random Walk. It optimizes the unreasonable assumption, improves the power-law distribution, and identifies collusion attack by preference similarity. Mathematic analyses and simulation results show that the proposed model is more robust under general conditions that collusion attacks in an attempt to deliberately subvert the system, and the success rates of transactions are higher compared to the current trust model.

In vehicular ad hoc networks (VANETs), the presence of untrustworthy nodes poses a significant threat, impacting the network’s reliability. This has led to the emergence of node-centric misbehavior detection as a crucial aspect of VANET security, focusing on the behavior of vehicles rather than the content of their interactions. While the trust model is a popular approach, the computational complexity of trust computations and management in VANETs is attributed to the intricate relationships among vehicles and the dynamic autonomous movement of nodes. To tackle these challenges, we developed a hybrid trust model scheme for node-centric misbehavior detection. Our method represents complex vehicular relationships using a hyper-graph within a dynamic behavior-homogeneous cluster. The model incorporates direct and indirect trust in a multi-layered hybrid trust framework, enabling accurate computation of the aggregate trust level for each cluster member vehicle. Experimental results demonstrate the effectiveness of our scheme, particularly in high-density vehicle cooperation scenarios, highlighting its promising ability to detect misbehaving nodes.

Random group recommendation model based on fuzzy clustering

With the enrichment of social activities, people are more inclined to group activities in life, study and work. Therefore, how to make a recommendation for group user has become a research hotspot. As the result of the existed recommendation systems are designed for individuals, this paper proposes a group recommendation method based on user's interaction behavior. First, a global trust estimation model based on the interaction between users is proposed, which can be used to estimate the global trust of each user. Second, each user's score of the project can be obtained through collaborative filtering. In the third step, the results obtained in the above two parts are combined to obtain the scoring prediction model proposed in this paper. Finally, according to the characteristics of the group itself, a model GROUP-MAE which is more suitable for evaluating the accuracy of group recommendation systems is proposed in this paper. By comparing experiments on EPINIONS_EXTENDED dataset, it is shown that the proposed model is superior to the traditional cooperative fusion strategy in precision and accuracy.

Point of Interest (POI) recommender systems help provide their users with a location or place that they might be interested in visiting. When combined with Location-Based Social Networks (LBSNs), POI recommender systems can be restructured to recommend POI for groups of users and not just individuals. The research focused on Group Recommender Systems (GRSs) and specifically, POI GRSs are scarce when compared to recommender systems for individuals. There are two main techniques that are used for POI GRSs, the Group Profiling methods and the Users Score Aggregation methods. Both methods have their drawbacks, as the Group Profiling methods do not recommend well for new groups and the Users Score Aggregation methods generally do not perform as well as the Group Profiling methods for established groups. In this paper, we propose new result aggregation methods that use both the Group Profiling method and the Users Score Aggregation method's results to provide the best POI group recommendations without the aforementioned drawbacks.

Apart from the fact that trust between doctors and patients is a source of sustainable relations, affecting the behaviour of both parties, it also has its therapeutic value.Therefore, this paper aims to construct a model of trust in the doctor-patient relationship based on qualitative research (analysis of the contents of Internet message boards).The study has revealed that trust towards doctors is a result of overlapping and interpenetration of two levels of trust: macro-and meso-.Macro-trust can be seen as a context in which all the dimensions of institutional trust are 'embedded'.Whereas meso-trust (institutional) is described in terms of three dimensions: benevolence, competence and integrity.

Collaborative filtering based methods have a low performance in the context of social recommendation due to the data sparsity issue and not considering the social network information that can be exploited to improve the performance. Trust-based methods attempt to reduce the data sparsity by utilizing the social network information. In this paper, we propose a hybrid personal trust model which adaptively combines the rating-based trust model and explicit trust metric to resolve the drawback caused by insufficient past rating records. The proposed method takes user-item matrix and user interaction information as inputs, and calculates the degree of direct trust to produce an initiative trust matrix．Then，following the predefined trust propagation rules，the algorithm infers the degree of indirect trust between users and transforms the initiative trust matrix to a denser trust matrix．The denser trust matrix and rating matrix are collectively used to find the k nearest neighbors for the target user. We present our empirical experiments in a real sparse data set MovieLens, the experimental results shows our algorithm can increase the data density and achieve lower MAE, that is to say, the proposed approach can efficiently improve recommendation quality.