Heterogeneous Hyperbolic Hypergraph Neural Network for Friend Recommendation in Location-based Social Networks

Trust-aware location recommendation in location-based social networks: A graph-based approach

Heterogeneous graph-based joint representation learning for users and POIs in location-based social network

New developments in mobile computing, wireless networking and social networking leverage opportunities to create location-based social networks (LBSNs) such as Foursquare, Gowalla, Facebook Place. Pointof-interest (POI) recommendation is a significant task in LBSNs, since it provides recommendations that respect personal preferences and capture social and environmental parameters. Moreover, employing Wireless Sensor Networks (WSNs) in POIs enables to improve recommendation accuracy. Most of existing POI recommendation methods deals with the only single domain, whereas the use of information available from other application domains potentially provides better POI recommendation results. This is particularly relevant for the newly launched LBSNs, where little check in information is available. In this paper, to the best of our knowledge, we are the first to propose a cross domain POI recommendation method in LBSNs and apply transfer learning for this purpose. To this end, we first incorporate both user and POI knowledge in auxiliary domains by discovering the principle coordinates. Next, we transfer them to the target domain and then make predictions using the transferred knowledge. The experimental results demonstrate high performance of our method and the merits of cross-domain POI recommendation in LBSNs.

Point-of-interest (POI) recommendations in location-based social networks (LBSNs) allow online users to discover various POIs for social activities occurring in the near future close to their current locations. Research has verified that people’s preferences regarding POIs are significantly affected by various internal and external contextual factors, which are therefore worth extensive study for POI recommendation. However, although psychological effects have also been demonstrated to be significantly correlated with an individual’s preferences, such effects have been largely ignored in previous studies on POI recommendation. For this paper, inspired by the famous memory theory in psychology, we were interested in whether memory-based preferences could be derived from users’ check-in data. Furthermore, we investigated how to incorporate these memory-based preferences into an effective POI recommendation scheme. Consequently, we refer to Ebbinghaus’s theory on memory, which describes the attenuation of an individual’s memory in the form of a forgetting curve over time. We first created a memory-based POI preference attenuation model and then adopted it to evaluate individuals’ check-ins. Next, we employed the memory-based values of check-ins to calculate the POI preference similarity between users in an LBSN. Finally, based on this memory-based preference similarity, we developed a novel POI recommendation method. We experimentally evaluated the proposed method on a real LBSN data set crawled from Foursquare. The results demonstrate that our method, which incorporates the proposed memory-based preference similarity for POI recommendation, significantly outperforms other methods. In addition, we found the best value of the parameter H in the memory-based preference model that optimizes the recommendation performance. This value of H implies that an individual’s memory usually has an effect on their daily travel choices for approximately 300 days.

The task of points-of-interest (POI) recommendations has become an essential feature in location-based social networks (LBSNs) with the significant growth of shared data on LBSNs. However it remains a challenging problem, because the decision process of a user choosing to visit a POI depends on numerous factors. The high level of sparsity of the data in LBSNs makes the POI recommendation problem even more challenging, especially for large geographical areas and worldwide datasets. Moreover, in this context the mobility behavior of the users is very heterogeneous, ranging from urban to worldwide mobility. In this paper, we explore the impact of spatial clustering on the recommendation quality. The proposed approach combines spatial clustering with users' influences. It is based on a Poisson factorization model built on an implicit social network, inferred from the geographical mobility patterns. We conduct a comprehensive performance evaluation of our approach on the YFCC dataset (a very large-scale real-world dataset). The experiments show that our approach achieves a significantly superior recommendation quality compared to other state-of-the-art recommendation techniques.

Virtual friend recommendations in virtual worlds

A survey of graph neural network based recommendation in social networks

The rising prosperity of Location-based Social Networks (LBSNs) witnessed an explosion in the availability of geo-tagged social media data, which enables tremendous location-aware online services, especially next point of interest (POI) recommendation. However, previous next POI recommendation studies usually adopt fix-length time windows for user check-in sequence modeling, leading to a limited capacity in capturing fine-grained user temporal preferences that easily change over time. Besides, existing methods often directly leverage multi-modal contexts as auxiliary to alleviate the data sparsity issue, which fails to fully exploit the sequential patterns of contextual information for inferring user interest drift. To address the above challenges, we propose a novel framework named iTourSPOT which extends traditional collaborative filtering methods with a context-aware POI embedding architecture. For enhancing temporal interests modeling capacity, we associate the context feature extraction with varying-length sessions and incorporate check-in frequencies of POIs as prior knowledge to instruct the session representation learning of our model. Moreover, a collaborative sequence transduction model is designed for joint context sequence modeling and session-based POI recommendation. Experimental results on a real-world geo-tagged photo dataset clearly demonstrate the effectiveness of the proposed framework when compared with state-of-the-art baseline methods, especially in both sparse and cold-start scenarios.

With the proliferation of social network services (e.g., Facebook, Twitter, and Instagram), identifying cohesive subgraphs in networks is a key and fundamental problem to understand users' activity. Recently this problem has received extensive research attention. Identifying cohesive subgraph structures in networks has many applications such as recommendation, detecting fraudsters, event organization, and graph analysis. In this thesis, we study three important problems of finding cohesive subgraphs. Our studies include finding a query-based cohesive subgraph search in social networks and location-based social networks, and detecting all cohesive subgraphs in attributed bipartite networks. In our study of Density Modularity based Community Search (DMCS), we propose the modularity-based community search problem in networks. Modularity is a widely used measure of the structure of networks which checks the strength of partition of a network into communities. Most of the existing community search models only focus on the internal cohesiveness of a community. However, a high-quality community often has high modularity, which implies dense connections inside communities and sparse connections to the nodes outside the community. Hence, we conduct a pioneer study on searching for a community with high modularity. We point out that while modularity has been used in community detection (unrelated to query nodes), its application in community search (related to query nodes) brings in different challenges named free-rider effect and resolution limit problems. Both problems indicate that intrinsically optimizing the modularity can fail to find a small-sized community as a result for the community search problem. We mitigate these problems by designing a new graph modularity function named Density Modularity. To the best of our knowledge, this is the first work on the community search problem by using the graph modularity. The community search based on the density modularity, termed as DMCS, is to find a community in a social network that contains all the query nodes and has high density modularity. We prove that the DMCS problem is NP-hard, and thus there is no exact algorithm that is scalable to large graphs. To efficiently address DMCS, we present new algorithms that run in log-linear time to the graph size. We conduct extensive experimental studies in real-world and synthetic networks, which offer insights into the efficiency and effectiveness of our approximate algorithms. In our study of GeoSocial Community Search(GCS) problem, we propose the community search problem in location-based social networks. Specifically, we aim at finding a social community and a location cluster that are densely connected in a location-based social network. GCS can be useful for finding marketing targets and user/location recommendations. To the best of our knowledge, this is the first work to find a social community and a location cluster that are densely connected from location-based social networks. We prove that the GCS problem is NP-hard and is not in APX, unless P = NP. To solve GCS problem, we propose three effective and efficient algorithms: Core-based Basic Algorithm, Top-down Greedy Removing Algorithm, and Bottom-up Greedy Expansion Algorithms. Finally, we report extensive experimental studies that offer insights into the efficiency and effectiveness of the proposed solutions. In our study of Attributed Bipartite Co-clustering (ABC), we propose the co-clustering problem in attributed bipartite networks. Co-clustering in bipartite networks is a fundamental and important problem. Specifically, we unify two main concepts: (i) bipartite modularity optimization, and (ii) attribute cohesiveness. To the best of our knowledge, this is the first work to find a set of co-clusters while considering the attribute cohesiveness. We prove that ABC problem is NP-hard and propose three effective and efficient algorithms: (1) Top-Down Algorithm; (2) Bottom-Up Algorithm; and (3) Group-based Matching Algorithm. Finally, extensive experimental results on real-world attributed bipartite networks demonstrate the efficiency and effectiveness of our algorithms.

One of the most attractive problems of social network analysis is the link prediction. Social networks' user growth is mostly supported with data driven friend recommendations which are provided by link predictors. Previously, we had studied new features to improve prediction accuracy in Location Based Social Networks (LBSNs) where users share temporal location information with check-in interactions. In this paper, we focused on the efficiency of link predictors as the speed of prediction is as critical as its accuracy in LBSNs. Extraction time costs and prediction accuracy of individual LBSN features are mined to pick a feature subset that is achieving faster link prediction while not losing from accuracy.

With the development of Internet of Things technology, the will to make cities smarter is growing. In the context of smart cities, where people are surrounded by a tremendous number of points of interests (POIs), POI recommendation is of great significance. The massive amount of user check-in data collected by location-based social networks (LBSNs) can help users explore new places via POI recommendations. With rich auxiliary data becomeing available in LBSNs, purely exploiting users' check-in information for POI recommendation is not sufficient. Although several attempts have been done to employ geographical influence to enhance POI recommendation, they simply use a certain distribution function to measure the geographical influence between POIs, which may lead to biased results. To this end, this article proposes a geography-aware inductive matrix completion (GAIMC) approach for personalized POI recommendation. Specifically, the GAIMC model consists of two parts, including geographic feature extraction via a Gaussian mixture model (GMM) and inductive matrix completion for recommendation. The GAIMC model first captures the geographical influence among users and POIs by the GMM, which can mine clustering information with hierarchical structures based on the users' check-in data and POIs' location information. Then, a matrix-completion method called inductive matrix completion, which can incorporate geographical features with the user-POI association metric, is utilized to recommend POIs. The experimental results on two real-world LBSN data sets demonstrate that our proposed model can achieve the best recommendation performance in comparison with the state-of-the-art counterparts.

These proceedings contain the papers selected for presentation at the seventh edition of the ACM SIGSPATIAL Workshop on Location-Based Social Networks (LBSN 2014) which is being held in conjunction with the ACM SIGSPATIAL GIS conference in 2014. Social networks have been prevalent on the Internet and become a hot research topic attracting many professionals from a variety of fields. The advances in location-acquisition and mobile communication technologies empower people to use location data with existing online social networks. The dimension of location helps bridge the gap between the physical world and online social networking services. Furthermore, people in an existing social network can expand their social structure with the new interdependency derived from their locations. As location is one of the most important components of user context, extensive knowledge about an individual's interests, behaviors, and relationships with others can be learned from locations. These kinds of location-embedded and location-driven social structures are known as location-based social networks. The objective of this workshop is to provide professionals, researchers, and technologists with a single forum where they can discuss and share the state-of-the-art of LBSN development and applications, present their ideas and contributions, and set future directions in emerging innovative research for location based social networks. This year program is composed of two tracks of predominantly "social" and "spatial" views on LBSNs.

In the last two decades, the issue of Influence Maximization (IM) in traditional online social networks has been extensively studied since it was proposed. It is to find a seed set which has maximum influence spread under a specific network transmission model. However, in real life, the information can be spread not only through online social networks, but also between neighbors who are close to each other in the physical world. Location-Based Social Network (LBSN) is a new type of social network which is emerging increasingly nowadays. In a LBSN, users can not only make friends, but also share the events they participate in at different locations by checking in. In this paper, we aim to study the IM in LBSNs, where we consider both the influence of online and offline interactions. A two-layer network model and an information propagation model are proposed. Also, we formalize the IM problem in LBSNs and present an algorithm obtaining an approximation factor of \((1 - 1/e - \epsilon \)) in near-linear expected time. The experimental results show that the algorithm is efficient meanwhile offering strong theoretical guarantees.

With the development of location-based social networks (LBSNs), users can check in Point-of-Interest (POIs) at any time. However, users do not check in all places they have visited, so the POI trajectory sequence generated through LBSNs is incomplete. An incomplete POI trajectory will have a negative impact on subsequent tasks such as POI recommendation and next POI prediction. Therefore, we complete the missing POI in the user trajectory sequence. Since the POI trajectory sequence is incomplete, it is a challenge to use the pre-order and post-order trajectory sequences with missing POIs. Therefore, we propose a masked POI trajectory model (MPTM) that uses the bidirectionality of BERT to complete the missing POIs in user’s behavior sequence. By masking the missing POIs, MPTM fully explores the relationship between the missing POIs and the known POIs to predict the missing POIs. In order to strengthen the relationship between POIs in the user trajectory sequence, we build a graph for each user’s incomplete POIs sequence to explore the user’s hidden behavior habits. Besides, we design experiments to explore the relationship between the continuity of the number of missing POIs and the predictive ability of the model. The experimental results demonstrate that our MPTM outperforms the state-of-the-art models for completion on missing POIs of user’s behavior sequence.

Leveraging graph neural networks for point-of-interest recommendations