Towards efficient communications in federated learning: A contemporary survey

In the traditional distributed machine learning scenario, the user’s private data is transmitted between clients and a central server, which results in significant potential privacy risks. In order to balance the issues of data privacy and joint training of models, federated learning (FL) is proposed as a particular distributed machine learning procedure with privacy protection mechanisms, which can achieve multi-party collaborative computing without revealing the original data. However, in practice, FL faces a variety of challenging communication problems. This review seeks to elucidate the relationship between these communication issues by methodically assessing the development of FL communication research from three perspectives: communication efficiency, communication environment, and communication resource allocation. Firstly, we sort out the current challenges existing in the communications of FL. Second, we have collated FL communications-related papers and described the overall development trend of the field based on their logical relationship. Ultimately, we discuss the future directions of research for communications in FL.     

基于本地化差分隐私的政务数据共享隐私保护算法研究

[目的/意义]为了合理化决策,通常一个政府部门会根据业务需求向其他部门共享某类数据,为本部门管理或服务决策提供辅助参考依据。数据共享在其中至关重要,但若在没有适当预防措施的情况下就共享政务数据,将容易造成隐私信息的泄露。[方法/过程]针对政府部门间共享统计数据的场景,提出一种基于本地化差分隐私的政务数据共享方法。该方法在算法Generalized randomized response(GRR)的基础上引入数据分箱思想,通过等宽分箱将数据记录分入更小的数据域范围内,以克服当前隐私保护算法在数据域较大且数据量较少时统计误差大的问题。[结果/结论]将所提算法与GRR算法在仿真数据集和真实数据集上均进行了对比分析,实验结果表明该算法可有效降低统计误差,并能在不同分布和数据域大小下保持其效用性。     

Efficient federated item similarity model for privacy-preserving recommendation

Previous federated recommender systems are based on traditional matrix factorization, which can improve personalized service but are vulnerable to gradient inference attacks. Most of them adopt model averaging to fit the data heterogeneity of federated recommender systems, requiring more training costs. To address privacy and efficiency, we propose an efficient federated item similarity model for the heterogeneous recommendation, called FedIS, which can train a global item-based collaborative filtering model to eliminate user feature dependencies. Specifically, we extend the neural item similarity model to the federated model, where each client only locally optimizes the shared item feature matrix. We then propose a fast-convergent federated aggregation method inspired by meta-learning to address heterogeneous user updates and accelerate the convergence of global training. Furthermore, we propose a two-stage perturbation method to protect both local training and transmission while reducing communication costs. Finally, extensive experiments on four real-world datasets validate that FedIS can provide more competitive performance on federated recommendations. Our proposed method also shows significant training efficiency with less performance degradation.     

Passengers information in public transport and privacy: Can anonymous tickets prevent tracking?

Modern public transportation companies often record large amounts of information. Privacy can be safeguarded by discarding nominal tickets, or introducing anonymization techniques. But is anonymity at all possible when everything is recorded? In this paper we discuss travel information management in the public transport scenario and we present a revealing case study (relative to the city of Cesena, Italy), showing that even anonymous 10-ride bus tickets may betray a user's privacy expectations. We also propose a number of recommendations for the design and management of public transport information systems, aimed at preserving the users’ privacy, while retaining the useful analysis features enabled by the e-ticketing technology.     

A scalable privacy-preserving recommendation scheme via bisecting k-means clustering

Privacy-preserving collaborative filtering is an emerging web-adaptation tool to cope with information overload problem without jeopardizing individuals’ privacy. However, collaborative filtering with privacy schemes commonly suffer from scalability and sparseness as the content in the domain proliferates. Moreover, applying privacy measures causes a distortion in collected data, which in turn defects accuracy of such systems. In this work, we propose a novel privacy-preserving collaborative filtering scheme based on bisecting k-means clustering in which we apply two preprocessing methods. The first preprocessing scheme deals with scalability problem by constructing a binary decision tree through a bisecting k-means clustering approach while the second produces clones of users by inserting pseudo-self-predictions into original user profiles to boost accuracy of scalability-enhanced structure. Sparse nature of collections are handled by transforming ratings into item features-based profiles. After analyzing our scheme with respect to privacy and supplementary costs, we perform experiments on benchmark data sets to evaluate it in terms of accuracy and online performance. Our empirical outcomes verify that combined effects of the proposed preprocessing schemes relieve scalability and augment accuracy significantly.     

User k-anonymity for privacy preserving data mining of query logs

The anonymization of query logs is an important process that needs to be performed prior to the publication of such sensitive data. This ensures the anonymity of the users in the logs, a problem that has been already found in released logs from well known companies. This paper presents the anonymization of query logs using microaggregation. Our proposal ensures the k-anonymity of the users in the query log, while preserving its utility. We provide the evaluation of our proposal in real query logs, showing the privacy and utility achieved, as well as providing estimations for the use of such data in data mining processes based on clustering.     

Zero-shot stance detection via multi-perspective contrastive learning with unlabeled data

Stance detection is to distinguish whether the text’s author supports, opposes, or maintains a neutral stance towards a given target. In most real-world scenarios, stance detection needs to work in a zero-shot manner, i.e., predicting stances for unseen targets without labeled data. One critical challenge of zero-shot stance detection is the absence of contextual information on the targets. Current works mostly concentrate on introducing external knowledge to supplement information about targets, but the noisy schema-linking process hinders their performance in practice. To combat this issue, we argue that previous studies have ignored the extensive target-related information inhabited in the unlabeled data during the training phase, and propose a simple yet efficient Multi-Perspective Contrastive Learning Framework for zero-shot stance detection. Our framework is capable of leveraging information not only from labeled data but also from extensive unlabeled data. To this end, we design target-oriented contrastive learning and label-oriented contrastive learning to capture more comprehensive target representation and more distinguishable stance features. We conduct extensive experiments on three widely adopted datasets (from 4870 to 33,090 instances), namely SemEval-2016, WT-WT, and VAST. Our framework achieves 53.6%, 77.1%, and 72.4% macro-average F1 scores on these three datasets, showing 2.71% and 0.25% improvements over state-of-the-art baselines on the SemEval-2016 and WT-WT datasets and comparable results on the more challenging VAST dataset.     

Federated reinforcement learning approach for detecting uncertain deceptive target using autonomous dual UAV system

This paper develops a cooperative federated reinforcement learning (RL) strategy that enables two unmanned aerial vehicles (UAVs) to cooperate in learning and predicting the movements of an intelligent deceptive target in a given search area. The proposed strategy allows the UAVs to autonomously cooperate, through information exchange of the gained experience to maximize the target detection performance and accelerate the learning speed while maintaining privacy. Specifically, we consider a monitoring model that includes a search area, a charging station, two cooperative UAVs, an intelligent deceptive uncertain moving target, and a fake (false) target. Each UAV is equipped with a limited-capacity rechargeable battery and a communication unit for exchanging the gained experience. The problem of maximizing the detection probability of the uncertain deceptive target using cooperative UAVs is mathematically modeled as a search-benefit maximization problem, which is then reformulated as a Markov decision process (MDP) due to the uncertainty nature of the problem. Because there is no prior information on the targets’ movement, a cooperative RL, is utilized to tackle the problem. The proposed cooperative RL-based algorithm is a distributed collaborative mechanism that enables the two UAVs, i.e., agents, to individually interact with the operating environment and maximize their cumulative rewards by converging to a shared policy while achieving privacy. Simulation results indicate that a cooperative RL-based dual UAV system can noticeably improve the target detection probability, reduce the detection performance, and accelerate the learning speed.     

Meta-prompt based learning for low-resource false information detection

The wide spread of false information has detrimental effects on society, and false information detection has received wide attention. When new domains appear, the relevant labeled data is scarce, which brings severe challenges to the detection. Previous work mainly leverages additional data or domain adaptation technology to assist detection. The former would lead to a severe data burden; the latter underutilizes the pre-trained language model because there is a gap between the downstream task and the pre-training task, which is also inefficient for model storage because it needs to store a set of parameters for each domain. To this end, we propose a meta-prompt based learning (MAP) framework for low-resource false information detection. We excavate the potential of pre-trained language models by transforming the detection tasks into pre-training tasks by constructing template. To solve the problem of the randomly initialized template hindering excavation performance, we learn optimal initialized parameters by borrowing the benefit of meta learning in fast parameter training. The combination of meta learning and prompt learning for the detection is non-trivial: Constructing meta tasks to get initialized parameters suitable for different domains and setting up the prompt model’s verbalizer for classification in the noisy low-resource scenario are challenging. For the former, we propose a multi-domain meta task construction method to learn domain-invariant meta knowledge. For the latter, we propose a prototype verbalizer to summarize category information and design a noise-resistant prototyping strategy to reduce the influence of noise data. Extensive experiments on real-world data demonstrate the superiority of the MAP in new domains of false information detection.     

Student privacy in learning analytics: An information ethics perspective

ABSTRACT

Higher education institutions have started using big data analytics tools. By gathering information about students as they navigate information systems, learning analytics employs techniques to understand student behaviors and to improve instructional, curricular, and support resources and learning environments. However, learning analytics presents important moral and policy issues surrounding student privacy. We argue that there are five crucial questions about student privacy that we must address in order to ensure that whatever the laudable goals and gains of learning analytics, they are commensurate with respecting students' privacy and associated rights, including (but not limited to) autonomy interests. We address information access concerns, the intrusive nature of information-gathering practices, whether or not learning analytics is justified given the potential distribution of consequences and benefits, and issues related to student autonomy. Finally, we question whether learning analytics advances the aims of higher education or runs counter to those goals.     

What the fake? Probing misinformation detection standing on the shoulder of novelty and emotion

One of the most time-critical challenges for the Natural Language Processing (NLP) community is to combat the spread of fake news and misinformation. Existing approaches for misinformation detection use neural network models, statistical methods, linguistic traits, fact-checking strategies, etc. However, the menace of fake news seems to grow more vigorous with the advent of humongous and unusually creative language models. Relevant literature reveals that one major characteristic of the virality of fake news is the presence of an element of surprise in the story, which attracts immediate attention and invokes strong emotional stimulus in the reader. In this work, we leverage this idea and propose textual novelty detection and emotion prediction as the two tasks relating to automatic misinformation detection. We re-purpose textual entailment for novelty detection and use the models trained on large-scale datasets of entailment and emotion to classify fake information. Our results correlate with the idea as we achieve state-of-the-art (SOTA) performance (7.92%, 1.54%, 17.31% and 8.13% improvement in terms of accuracy) on four large-scale misinformation datasets. We hope that our current probe will motivate the community to explore further research on misinformation detection along this line. The source code is available at the GitHub.²     

Digital Images for the Information Professional,by Melissa M. Terras. Burlington,VT: Ashgate, 2008. 245 pp. $114.95 cloth. ISBN 978-0-7546-4860-4 (cloth)

Privacy has largely been equated with every individual's right to privacy. Accordingly, current efforts to protect privacy on the Internet have sought anonymity by breaking, where possible, links with personally identifiable information (PII)—all uses of aggregated data stripped of PII are considered legitimate. This article argues that we need to use a broader concept, general or group identifying information (GII), because even aggregated data stripped of PII violate privacy at the community level. The search engine companies, or anyone else with access to their log files, can use these data to generate a moment-by-moment view of what is on the collective mind. Such a view can be used in a variety of ways, some with deep economic and even political impact. In order to frame this discussion, it is necessary to examine some of the realities of the search engine-mediated associative interface to the World Wide Web. While this interface has enormous benefits for the networked world, it also fundamentally changes a number of issues underlying various current debates about Internet governance.     

Who Searches the Searchers? Community Privacy in the Age of Monolithic Search Engines

Privacy has largely been equated with every individual's right to privacy. Accordingly, current efforts to protect privacy on the Internet have sought anonymity by breaking, where possible, links with personally identifiable information (PII)—all uses of aggregated data stripped of PII are considered legitimate. This article argues that we need to use a broader concept, general or group identifying information (GII), because even aggregated data stripped of PII violate privacy at the community level. The search engine companies, or anyone else with access to their log files, can use these data to generate a moment-by-moment view of what is on the collective mind. Such a view can be used in a variety of ways, some with deep economic and even political impact. In order to frame this discussion, it is necessary to examine some of the realities of the search engine-mediated associative interface to the World Wide Web. While this interface has enormous benefits for the networked world, it also fundamentally changes a number of issues underlying various current debates about Internet governance.     

Whose posts to read: Finding social sensors for effective information acquisition

In the era of big data, it is extremely challenging to decide what information to receive and filter out in order to effectively acquire high-quality information, particularly in social media where large-scale User Generated Contents (UGC) is widely and quickly disseminated. Considering that each individual user in social network can take actions to drive the process of information diffusion, it is naturally appealing to aggregate spreading information effectively at the individual level by regarding each user as a social sensor. Along this line, in this paper, we propose a framework for effective information acquisition in social media. To be more specific, we introduce a novel measurement, the preference-based Detection Ability to evaluate the ability of social sensors to detect diffusing events, and the problem of effective information acquisition is then reduced to achieving social sensing maximization through discovering valid social sensors. In pursuit of social sensing maximization, we propose two algorithms to resolve the longstanding problems in traditional greedy methods from the perspectives of efficiency and performance. On the one hand, we propose an efficient algorithm termed LeCELF, which resolves the redundant re-evaluations in the traditional Cost-Effective Lazy Forward (CELF) algorithm. On the other hand, we observe the participation paradox phenomenon in the social sensing network, and proceed to propose a randomized selection-based algorithm called FRIENDOM to choose social sensors to improve the effectiveness of information acquisition. Experiments on a disease spreading network and real-world microblog datasets have validated that LeCELF greatly reduces the running time, whereas FRIENDOM achieves a better detection performance. The proposed framework and corresponding algorithms can be applicable in many other settings in resolving information overload problems.     

Anonymity

Anonymity is a form of nonidentifiability which I define as noncoordinatability of traits in a given respect. This definition broadens the concept, freeing it from its primary association with naming. I analyze different ways anonymity can be realized. I also discuss some ethical issues, such as privacy, accountability and other values which anonymity may serve or undermine. My theory can also conceptualize anonymity in information systems where, for example, privacy and accountability are at issue.     

Informed Patch Enhanced HyperGCN for skeleton-based action recognition

Human skeleton, as a compact representation of action, has attracted numerous research attentions in recent years. However, skeletal data is too sparse to fully characterize fine-grained human motions, especially for hand/finger motions with subtle local movements. Besides, without containing any information of interacted objects, skeleton is hard to identify human–object interaction actions accurately. Hence, many action recognition approaches that purely rely on skeletal data have met a bottleneck in identifying such kind of actions. In this paper, we propose an Informed Patch Enhanced HyperGraph Convolutional Network that jointly employs human pose skeleton and informed visual patches for multi-modal feature learning. Specifically, we extract five informed visual patches around head, left hand, right hand, left foot and right foot joints as the complementary visual graph vertices. These patches often exhibit many action-related semantic information, like facial expressions, hand gestures, and interacted objects with hands or feet, which can compensate the deficiency of skeletal data. This hybrid scheme can boost the performance while keeping the computation and memory load low since only five extra vertices are appended to the original graph. Evaluation on two widely used large-scale datasets for skeleton-based action recognition demonstrates the effectiveness of the proposed method compared to the state-of-the-art methods. Significant accuracy improvements are reported using X-Sub protocol on NTU RGB+D 120 dataset.     

The obstacle problem of integro-partial differential equations with applications to stochastic optimal control/stopping problem

This paper is devoted to existence and uniqueness of minimal mild super solutions to the obstacle problem governed by integro-partial differential equations. We first study the well-posedness and local Lipschitz regularity of L^p solutions (p?≥?2) to reflected forward-backward stochastic differential equations (FBSDEs) with jump and lower barrier. Then we show that the solutions to reflected FBSDEs provide a probabilistic representation for the mild super solution via a nonlinear Feynman–Kac formula. Finally, we apply the results to study stochastic optimal control/stopping problems.     

Stewardship of personal data on social networking sites

The extant Information Management literature highlights the asymmetric distribution of power between users and online platforms, while the issues related to the stewardship of personal data on such platforms remain problematic and largely unresolved. To address that lacuna, we propose a conceptual design that can help to overcome many of the challenges related to storage, analysis, and integrity associated with the stewardship of personal data on online platforms. We adopt a systemic perspective and propose a shift from the current user-platform relationship to one in which users control the level of access to their data, organisations are relieved from the burden of maintaining personal data, and the data are not decoupled from information about their provenance and context of origin. We apply our conceptual design to the context of social networking sites, where we specifically address issues related to privacy, and identity and pave the path to a broader set of possible applications. We discuss the significance and timeliness of our proposed conceptual design for the stewardship of personal data, and the importance of our findings for future research, as well as for the design of online platforms.