Heterogeneous Graph Neural Networks (HGNNs) have achieved significant success in various real-world applications. However, existing methods often employ meta-paths, the random walk or shallow homogeneous convolutional graph neural networks such as GCN and GAT to aggregate neigh-borhood information. The meta-path requires prior knowledge of domain experts, while the random walk is not conducive to the link prediction task compared with the direct neighbor method. And the shallow models may have limited capability to capture the information of high-order neighbors. To address these challenges, we propose Multi-View Heterogeneous Graph Attention Network (MHAN), a novel approach that captures the neighbor information from both the local and the global view to effectively learn node representations. In MHAN, we first map the attributes of nodes of different types into the same feature space. Secondly, we utilize the attention mechanism to aggregate information from the first-order and high-order neighbors, respectively. Finally, we generate the final node representations by aggregating the obtained node embeddings from different views. Extensive experiments on three real-world datasets of heterogeneous graphs show that our model outperforms state-of-the-art baselines on the link prediction task.

Online Agricultural Question and Answering (Q&A) Communities aim to help farmers obtain useful suggestions in the form of interactive Q&A dialogues. However, with the increasing number of users, a large number of questions cannot be answered in time. Moreover, questions in agricultural Q&A communities often contain limited textual information which also becomes a major challenge for expert recommendation as traditional methods are not applicable. In this article, we leverage heterogeneous graph neural networks to address the issue. To do this, we take a popular agricultural Q&A community in China as an example and we propose DAHGNN, a Dual-Attention Heterogeneous Graph Neural Network model for the expert recommendation problem. Our model considers the influence of both the node type and different neighbors of the same type. Through extensive experimental evaluations, we show that our model can obtain a better performance compared to many state-of-art methods. Our work helps to improve user participation in the community so as to efficiently assist farmers to solve problems.

Social live video streaming has been a major Internet phenomenon with the rise of platforms like Facebook-Live, Youtube-Live, and Twich. New features tailored by these platforms provide unprecedented opportunities for viewers to engage and for broadcasters to gain tangible rewards through user donation. In this article, we conduct an in-depth analysis of user donations, a distinctive feature and prominent phenomenon in social live streaming, which however we have very limited understandings with. Based on a publicly available (anonymized) dataset with detailed information on over 4 million donation relationships that cover over 2 million users and worth in total over 200 million US dollars, we quantitatively reveal donation disparity and dynamics of donation relationships. Among other results, we find that repeated donation relationships largely exist and the strength significantly increases with the repetition level. Finally, we adopt machine-learned classifiers to accurately predict future donations. Our findings shed lights on the user retention problem and the design of social live video streaming services.

Online social games such as World of Tanks (WoT) have attracted hundreds of millions of users not only to play the games but also to share game replays and to watch games played by others in online game replay sharing communities. To date, user retention, including player retention for games and uploader retention for replay sharing communities, remains a key challenge in game design and in maintaining the community prosperity. In this article, we revisit the user retention problem and we deal with the player and the uploader retention problem simultaneously, through leveraging both the in- and out-game features uniquely provided by online game replay sharing communities. To this end, we have collected a large-scale dataset from a replay sharing community named WoTreplays, which contains records of over 380 000 game replays and nearly 2 million users. For the player retention problem, we examine traditional factors including the activity level, the gaming performance, and the social relationships of players, and we include additional features that are intuitively informative, such as how many users have watched a player, which are provided uniquely by the game replay sharing communities. Similarly, for the uploader retention problem, we study the activity level and the popularity of uploaders in game replay sharing communities and additionally include their gaming performance and social relationships in the games as a supplement. Based on our findings, we build machine-learned classifiers that can accurately predict users who will play intensively and users who will share a large number of game replays in the future. Among other results, we find that: 1) for both game playing and replay sharing, the activity level and the popularity of the users are highly skewed; 2) the influence between the gaming activity and the gaming performance is mutual; 3) while the majority of users do not form long-term relationships, a small fraction of users connect very closely; and 4) the gaming performance and the user popularity are very informative for predicting highly active users.

Network Representation Learning embeds each node in a network into a low-dimensional real-value vector which can be used for downstream tasks such as link prediction and recommendation. Many existing approaches use unsupervised or (semi-)supervised methods to explore the network topology and learn representations from it. In contrast, we propose, reinforcement learning network representations (RLNet), which explores the idea of using reinforcement learning to learn to explore the network and to obtain network representations. Based on reward signals, RLNet learns an actor which uses a policy to determine the network navigation actions. RLNet uses node representations to parameterize its policy, and the representations are learned together with the policy. Through experiments based on multiple datasets, we show that RLNet can obtain promising results in link prediction tasks.

Online gaming franchises such as World of Tanks, Defense of the Ancients, and StarCraft have attracted hundreds of millions of users who, apart from playing the game, also socialize with each other through gaming and viewing gamecasts. As a form of User Generated Content (UGC), gamecasts play an important role in user entertainment and gamer education. They deserve the attention of both industrial partners and the academic communities, corresponding to the large amount of revenue involved and the interesting research problems associated with UGC sites and social networks. Although previous work has put much effort into analyzing general UGC sites such as YouTube, relatively little is known about the gamecast sharing sites. In this work, we provide the first comprehensive study of gamecast sharing sites, including commercial streaming-based sites such as Amazon’s Twitch.tv and community-maintained replay-based sites such as WoTreplays. We collect and share a novel dataset on WoTreplays that includes more than 380,000 game replays, shared by more than 60,000 creators with more than 1.9 million gamers. Together with an earlier published dataset on Twitch.tv, we investigate basic characteristics of gamecast sharing sites, and we analyze the activities of their creators and spectators. Among our results, we find that (i) WoTreplays and Twitch.tv are both fast-consumed repositories, with millions of gamecasts being uploaded, viewed, and soon forgotten; (ii) both the gamecasts and the creators exhibit highly skewed popularity, with a significant heavy tail phenomenon; and (iii) the upload and download preferences of creators and spectators are different: while the creators emphasize their individual skills, the spectators appreciate team-wise tactics. Our findings provide important knowledge for infrastructure and service improvement, for example, in the design of proper resource allocation mechanisms that consider future gamecasting and in the tuning of incentive policies that further help player retention.

Summary User interactions are indispensable for any online network to thrive, especially for BitTorrent‐like and Web real‐time communication‐based distributed online networks that rely on users' collective contributions instead of the help of central servers. User interactions provide fine‐grained information for many applications, such as security enhancement and cooperation promotion. To date, several schemes forestimating user interaction strengthin centralized online networks have been proposed. In contrast, we present design, deployment, and analysis of UISE for user interaction strength estimation in distributed online networks. Among the strong points of UISE is that it captures both direct and indirect user interactions, and that it scales with only partial information dissemination. We apply UISE to devise the first distributed scheme for online time estimation and we implement it into Tribler, a distributed online network for media and social applications like file sharing, streaming, and voting. We demonstrate the accuracy and the scalability of UISE with different information dissemination protocols and user behaviors using simulations, emulations, and a real‐world deployment. Copyright © 2015 John Wiley & Sons, Ltd.

Multiplayer Online Games (MOGs) like Defense of the Ancients and StarCraft II have attracted hundreds of millions of users who communicate, interact, and socialize with each other through gaming. In MOGs, rich social relationships emerge and can be used to improve gaming services such as match recommendation and game population retention, which are important for the user experience and the commercial value of the companies who run these MOGs. In this work, we focus on understanding social relationships in MOGs. We propose a graph model that is able to capture social relationships of a variety of types and strengths. We apply our model to real-world data collected from three MOGs that contain in total over ten years of behavioral history for millions of players and matches. We compare social relationships in MOGs across different game genres and with regular online social networks like Facebook. Taking match recommendation as an example application of our model, we propose SAMRA, a Socially Aware Match Recommendation Algorithm that takes social relationships into account. We show that our model not only improves the precision of traditional link prediction approaches, but also potentially helps players enjoy games to a higher extent.

BitTorrent (BT) plays an important role in Internet content distribution. Because public BTs suffer from the free-rider problem, Darknets are becoming increasingly popular, which use Sharing Ratio Enforcement to increase their efficiency. We crawled and traced 17 Darknets from September 2009 to February 2011, and obtained datasets about over 5 million torrents. We conducted a broad range of measurements, including traffic, sites, torrents, and users activities. We found that some of the features of Darknets are noticeably different from public BTs. The results of our study reflect both macroscopic and microscopic aspects of the overall ecosystem of BitTorrent Darknets.

Many peer-to-peer communities, including private BitTorrent communities that serve hundreds of thousands of users, utilize credit-based or sharing ratio enforcement schemes to incentivize their members to contribute. In this paper, we analyze the performance of such communities from both the system-level and the user-level perspectives. We show that both credit-based and sharing ratio enforcement policies can lead to system-wide "crunches" or "crashes," where the system seizes completely due to too little or too much credit, respectively. We present a theoretical model that identifies the conditions that lead to these system pathologies and we design an adaptive credit system that automatically adjusts credit policies to maintain sustainability. Given private communities that are sustainable, it has been demonstrated that they are greatly oversupplied in terms of excessively high seeder-to-leecher ratios. We further analyze the user-level performance by studying the effects of oversupply. We show that although achieving an increase in the average downloading speed, the phenomenon of oversupply has three undesired effects: long seeding times, low upload capacity utilizations, and an unfair playing field for late entrants into swarms. To alleviate these problems, we propose four different strategies, which have been inspired by ideas in social sciences and economics. We evaluate these strategies through simulations and demonstrate their positive effects.

Enhancing reciprocity has been one of the primary motivations for the design of incentive policies in BitTorrent-like P2P systems. Reciprocity implies that peers need to contribute their bandwidth to other peers if they want to receive bandwidth in return. However, the over-provisioning that characterizes today's BitTorrent communities and the development of many next-generation P2P systems with real-time constraints (e.g., for live and on-demand streaming) suggest that more effort can be devoted to reducing the inequity (i.e., the difference of service received) among peers, rather than only enhancing reciprocity. Inspired by this observation, in this work we analyze in detail several incentive mechanisms that are used in BitTorrent systems, and explore several strategies that influence the balance between reciprocity and equity. Our study shows that (i) reducing inequity leads to a better overall system performance, and (ii) the behavior of seeders (i.e., peers that hold a complete copy of the file and upload it for free) influences whether reciprocity is enhanced or inequity reduced.

Providing incentives for user contribution has been one of the primary design goals of Peer-to-Peer systems. The newly-emerged BitTorrent private communities adopt Sharing Ratio Enforcement (SRE) on top of BitTorrent's incentive mechanism, Tit-For-Tat, in order to strictly enforce a minimum contribution a member has to provide, in relation to the amount of service it has received. In this paper, we provide a theoretical model to analyze 1) how SRE provides seeding incentives, and 2) how SRE influences the download performance in the system. Specifically, we study the influence of the SRE threshold (i.e., the minimum sharing ratio requirement) and the bandwidth heterogeneity of the peers in the system. In our analysis, we assume users to be rational, i.e., peers seed only the minimum amount required by SRE, and we show that the download performance as predicted by our model represents a lower bound for the actual performance that can be reached in a BitTorrent private community. Hence, following our model, community administrators can predict the minimum performance level their systems will be able to reach.

Many private BitTorrent communities employ Sharing Ratio Enforcement (SRE) schemes to incentivize users to contribute their upload resources. It has been demonstrated that communities that use SRE are greatly oversupplied, i.e., they have much higher seeder-to-leecher ratios than communities in which SRE is not employed. The first order effect of oversupply under SRE is a positive increase in the average downloading speed. However, users are forced to seed for extremely long times to maintain adequate sharing ratios to be able to start new downloads. In this paper, we propose a fluid model to study the effects of oversupply under SRE, which predicts the average downloading speed, the average seeding time, and the average upload capacity utilization for users in communities that employ SRE. We notice that the phenomenon of oversupply has two undesired negative effects: a) Peers are forced to seed for long times, even though their seeding efforts are often not very productive (in terms of low upload capacity utilization); and b) SRE discriminates against peers with low bandwidth capacities and forces them to seed for longer durations than peers with high capacities. To alleviate these problems, we propose four different strategies for SRE, which have been inspired by ideas in social sciences and economics. We evaluate these strategies through simulations. Our results indicate that these new strategies release users from needlessly long seeding durations, while also being fair towards peers with low capacities and maintaining high system-wide downloading speeds.

The "tracker" of a P2P system is used to lookup which peers hold (or partially hold) a given object. There are various designs for the tracker function, from a single-server tracker, to multiple-server tracker system, to DHT-based serverless systems. In this paper, we classify the different designs, discuss the different considerations for these designs, and simple models to evaluate the reliability of these designs.