Recognizing whether sensors are on the same body

In an open mobile health (mHealth) sensing system, users will be able to seamlessly pair sensors with their cellphone and expect the system to just work. This ubiquity of sensors, however, creates the potential for users to accidentally wear sensors that are not paired with their own cellphone. Our method probabilistically detects this situation by finding correlations between embedded accelerometers in the cellphone and sensor. We evaluate our method over a dataset of seven individuals with sensors in various positions on their body and experimentally show that our method is capable of achieving an accuracy of 85%.     

Expertise ranking using activity and contextual link measures

The Internet has transformed from a Web of content to a people-centric Web. People actively use social networking platforms to stay in contact with friends and colleagues. The availability of rich Web-based applications allows people to collaborate and interact online. These connected online societies provide an immense potential for future business models such as crowdsourcing. Based on the idea of crowdsourcing, we developed a framework that enables people to offer their skills and expertise as human-provided services (HPS) which can be discovered and requested on demand. Automated techniques for expertise mining become thus essential in such applications. We introduce a link intensity based ranking model for recommending relevant users in human collaborations. Here we argue that an expertise ranking model must consider the users' availability, activity level, and expected informedness. We present DSARank for estimating the relative importance of persons based on reputation mechanisms in collaboration networks. We test the applicability of our ranking model by using datasets obtained from real human interaction networks including mobile phone and email communications. The results show that DSARank is better suited for recommending users in collaboration networks than traditional degree-based methods.     

Classifying watermelon ripeness by analysing acoustic signals using mobile devices

This work addresses the problem of distinguishing between ripe and unripe watermelons using mobile devices. Through analysing ripeness-related features extracted by thumping watermelons, collecting acoustic signals by microphones on mobile devices, our method can automatically identify the ripeness of watermelons. This is possible in real time, making use of machine learning techniques to provide good accuracy. We firstly collect a training dataset comprising acoustic signals generated by thumping both ripe and unripe watermelons. Audio signal analysis on this helps identify features related to watermelon ripeness. These features are then used to construct a classification model for future signals. Based on this, we developed a crowdsourcing application for Android which allows users to identify watermelon ripeness in real time while submitting their results to us allowing continuous improvement of the classification model. Experimental results show that our method is currently able to correctly classify ripe and unripe watermelons with an overall accuracy exceeding 89 %.     

Personal mobile services

Ubiquitous information access through mobile devices has become a typical practice in everyday life. The mobile service paradigm shifts the role of mobile devices from consumers to providers, opening up new opportunities for a multitude of collaborative services and applications ranging from sharing personal information to collaborative participatory sensing. Although many basic principles of the standard Web service approach continue to apply, the inherent constraints of mobile devices and broadband wireless access render the deployment of the standard architecture in mobile environments inefficient. This paper introduced personal services, a user-centric paradigm that enables service-oriented interactions among mobile devices that are controlled via user-specified authorization policies. Personal services exploit the user’s contact list (ranging from phonebook to social lists) in order to publish and discover Web services while placing users in full control of their own personal data and privacy. Experimental validation demonstrates the ability of personal services to foster a new generation of collaborative mobile services. Performance evaluation results show that the publication and discovery through contact lists are efficient and that service announcements and discovery requests can reach a huge number of users in a few seconds. Results also support a conclusion that resources-constrained devices can collaborate to carry out functionalities beyond the ability of their resources limitations.     

Efficient mining of group patterns from user movement data

In this paper, we present a new approach to derive groupings of mobile users based on their movement data. We assume that the user movement data are collected by logging location data emitted from mobile devices tracking users. We formally define group pattern as a group of users that are within a distance threshold from one another for at least a minimum duration. To mine group patterns, we first propose two algorithms, namely AGP and VG-growth. In our first set of experiments, it is shown when both the number of users and logging duration are large, AGP and VG-growth are inefficient for the mining group patterns of size two. We therefore propose a framework that summarizes user movement data before group pattern mining. In the second series of experiments, we show that the methods using location summarization reduce the mining overheads for group patterns of size two significantly. We conclude that the cuboid based summarization methods give better performance when the summarized database size is small compared to the original movement database. In addition, we also evaluate the impact of parameters on the mining overhead.     

Ubiquitous conference management system for mobile recommendation services based on mobilizing social networks: A case study of u-conference

In this paper, we propose a social network-based mechanism to be aware of user contexts and to provide contextually relevant mobile services to users. A social network among users is regarded as the channel for exchanging and propagating their contexts. To efficiently discover the contexts of a certain users the contexts of his neighbors can be fused to provide mobile recommendation service to mobile subscribers. However, since the social network of a user has been fragmented into more than one, it is difficult to put the number of contexts from the fragmented social networks together (Jung, 2009b). Thereby, we mobilize all possible on- and off-line social networks to build an ego-centric social network. We have implemented the proposed system, which is called u-conference system, by collecting the social network dataset from online (e.g., Facebook, Twitter, CyWorld, and co-authoring patterns in major Korean journals) and offline (e.g., co-participation patterns in a number of Korean domestic conferences). Once we have implemented the system, we have provided mobile services to the conference participants by sending text messages about time schedule of relevant presentations.     

Private discovery of common social contacts

Digital services that are offered, and consumed, on the basis of social relationships form the backbone of social clouds—an emerging new concept that finds its roots in online social networks. The latter have already taken an essential role in people’s daily life, helping users to build and reflect their social relationships to other participants. A key step in establishing new links entails the reconciliation of shared contacts and friends. However, for many individuals, personal relationships belong to the private sphere, and, as such, should be concealed from potentially prying eyes of strangers. Consequently, the transition toward social clouds cannot set aside mechanisms to control the disclosure of social links. This paper motivates and introduces the concept of Private Discovery of Common Social Contacts, which allows two users to assess their social proximity through interaction and learn the set of contacts (e.g., friends) that are common to both users, while hiding contacts that they do not share. We realize private contact discovery using a new cryptographic primitive, called contact discovery scheme (CDS), whose functionality and privacy is formalized in this work. To this end, we define a novel privacy feature, called contact-hiding, that captures our strong privacy goals. We also propose the concept of contact certification and show that it is essential to thwart impersonation attacks on social relationships. We build provably private and realistically efficient CDS protocols for private discovery of mutual contacts. Our constructions do not rely on a trusted third party (TTP)—all contacts are managed independently by the users. The practicality of our proposals is confirmed both analytically and experimentally on different computing platforms. We show that they can be efficiently deployed on smartphones, thus allowing ad hoc and ubiquitous contact discovery outside of existing social networks. Our CDS constructions allow users to select their (certified) contacts to be included in individual protocol executions. That is, users may perform context-dependent contact discovery using any subset (circle) of their contacts.     

Discovering Typed Communities in Mobile Social Networks

Mobile social networks,which consist of mobile users who communicate with each other using cell phones,are reflections of people’s interactions in social lives.Discovering typed communities(e.g.,family communities or corporate communities) in mobile social networks is a very promising problem.For example,it can help mobile operators to determine the target users for precision marketing.In this paper we propose discovering typed communities in mobile social networks by utilizing the labels of relationships between users.We use the user logs stored by mobile operators,including communication and user movement records,to collectively label all the relationships in a network,by employing an undirected probabilistic graphical model,i.e.,conditional random fields.Then we use two methods to discover typed communities based on the results of relationship labeling:one is simply retaining or cutting relationships according to their labels,and the other is using sophisticated weighted community detection algorithms.The experimental results show that our proposed framework performs well in terms of the accuracy of typed community detection in mobile social networks.     

Hierarchy sustains partial cooperation and induces a Braess-like paradox in slotted aloha-based networks

This paper studies a hierarchical distributed choice of retransmission probabilities in slotted aloha. In particular, we consider a wireless system composed of one central receiver and several selfish mobile users communicating via the slotted aloha protocol. The set of mobile users is split into two classes: leaders and followers. We then study the induced non-cooperative hierarchical game based on the Stackelberg equilibrium concept. Using a 4D Markovian model, we compute the steady state of the system and derive the average throughput and the expected delay as well. We start by discussing the protocol design and propose a controlled slotted aloha using a virtual controller. The virtual controller can sustain partial cooperation among concurrent mobile users when accessing the channel by making the channel lossy. This leads us to identify a Braess-like paradox in which reducing capacity to the system may improve the performance of all mobile users. We then investigate the impact of hierarchy among mobile users in such a random access protocol and discuss how to distribute leader/follower roles. We show that the global performance of the system is improved compared to standard slotted aloha system. However, slight performances slow-down may be observed for the followers group when the total number of mobile users is relatively small.     

基于群体行为分析的人群异常聚集预测方法

随着智能通信设备的普及和通信基站定位精度的提升,利用通信基站记录的用户行为数据监测和预测人群密度成为可能。由于人群异常聚集事件具有突发性,利用时间序列分析方法和概率模型进行预测的效果较差。针对该问题,提出一种基于群体行为分析的预测方法。通过分析聚集人群的上网行为和基站间的人群移动行为特征,得到两者之间的相关性,结合基站的人群密度时间序列信息,利用扩张因果卷积神经网络和逻辑回归模型得出预测结果。运营商提供的手机用户上网记录数据集上的实验结果表明,该预测方法的精确率为0.93,召回率为0.97,显著优于ARIMA算法、LSTM算法和XGBoost算法,证明了引入用户群体的上网行为和移动特征能够有效提升人群异常聚集预测的准确性。     

Effects of the aesthetic design of icons on app downloads: evidence from an android market

With the rapid development of the mobile app market, understanding the determinants of mobile app success has become vital to researchers and mobile app developers. Extant research on mobile applications primarily focused on the numerical and textual attributes of apps. Minimal attention has been provided to how the visual attributes of apps affect the download behavior of users. Among the features of app “appearance”, this study focuses on the effects of app icon on demand. With aesthetic product and interface design theories, we analyze icons from three aspects, namely, color, complexity, and symmetry, through image processing. Using a dataset collected from one of the largest Chinese Android websites, we find that icon appearance influences the download behavior of users. Particularly, apps with icons featuring higher colorfulness, proper complexity, and slight asymmetry lead to more downloads. These findings can help developers design their apps.     

Discovery and Integration of Mobile Communications in Everyday Life

We report on the results of a study in which 19 new mobile telephone users were closely tracked for the first six weeks after service acquisition. Results show that novices tend to rapidly modify their perceptions of social appropriateness around mobile phone use, that actual nature of use frequently differs from initial predictions, and that comprehension of service-based technologies can be problematic. We also describe instances and features of mobile telephony practice. When in use, mobile phones occupy multiple social spaces simultaneously, spaces with norms that sometimes conflict: the physical space of the mobile phone user and the virtual space of the conversation.     

Service-oriented mobile malware detection system based on mining strategies

The large number of mobile internet users has highlighted the importance of privacy protection. Traditional malware detection systems that run within mobile devices have numerous disadvantages, such as overconsumption of processing resources, delayed updating, and difficulty in intersection. This study proposed a novel detection system based on cloud computing and packet analysis. The system detects the malicious behavior of the mobile malwares through their packets with the use of data mining methods. This approach completely avoids the defects of traditional methods. The system is service-oriented and can be deployed by mobile operators to send alarms to users who have malwares on their devices. To improve system performance, a new clustering strategy called contraction clustering was created. This strategy uses prior knowledge to reduce dataset size. Moreover, a multi-module detection scheme was introduced to enhance system accuracy. The results of this scheme are produced by integrating the detection results of several algorithms, including Naive Bayes and Decision Tree.     

An Empirical Study of the Impacts of Perceived Security and Knowledge on Continuous Intention to Use Mobile Fintech Payment Services

Financial technology (Fintech) services using emerging technology such as the Internet of Things (IoT) is becoming more prevalent. The recent proliferation of the mobile payment sector led by innovative mobile Fintech payment services such as Apple Pay and Samsung Pay is the most important and fastest growing Fintech services from consumers’ perspective. Although businesses have been making efforts to spread the use of the services, security is crucial in the diffusion of the services. Despite the importance, the role of perceived security in continuous intention to use mobile, Fintech services has not yet been investigated in depth. Thus, this study investigates the relationships between perceived security, knowledge regarding the services, confirmation, perceived usefulness, and satisfaction. We propose a research model using an extended post-acceptance model (EPAM) as a theoretical framework in the context of Fintech services. We then validate the model using the data collected from the service users. The analyzed results show that knowledge and perceived security in mobile Fintech services have a significant influence on users’ confirmation and perceived usefulness. However, perceived security does not directly influence users’ satisfaction and continual intention to use. We further find significant relationships among confirmation, perceived usefulness, satisfaction, and continual intention to use of the services. We discuss theoretical and practical contributions of the study.     

Greedy-based user selection for federated graph neural networks with limited communication resources

Recently, graph neural networks (GNNs) have attracted much attention in the field of machine learning due to their remarkable success in learning from graph-structured data. However, implementing GNNs in practice faces a critical bottleneck from the high complexity of communication and computation, which arises from the frequent exchange of graphic data during model training, especially in limited communication scenarios. To address this issue, we propose a novel framework of federated graph neural networks, where multiple mobile users collaboratively train the global model of graph neural networks in a federated way. The utilization of federated learning into the training of graph neural networks can help reduce the communication overhead of the system and protect the data privacy of local users. In addition, the federated training can help reduce the system computational complexity significantly. We further introduce a greedy-based user selection for the federated graph neural networks, where the wireless bandwidth is dynamically allocated among users to encourage more users to attend the federated training of neural networks. We perform the convergence analysis on the federated training of neural networks, in order to obtain some more insights on the impact of critical parameters on the system design. Finally, we perform the simulations on the coriolis ocean for reAnalysis (CORA) dataset and show the advantages of the proposed method in this paper.     

You never call: Demoting unused contacts on mobile phones using DMTR

Throughout their lives, people gather contacts on their mobile phones. Some of these are unused contacts—contacts that have not been used for a long time and are less likely to be used in future calls. These contacts compete for the users’ attention and the mobile phone’s limited screen capacity. To address this problem, we developed a prototype contact list interface called DMTR, which automatically demotes unused contacts by presenting them in a smaller font at the bottom of the contact list. In phase I of this research, we asked 18 participants to assess for how long they had not used each of their mobile phone contacts. Results show that 47% of all their contacts had not been used for over 6?months or had never been used at all. In phase II, we demoted these unused contacts using DMTR and asked our participants to locate contacts that they had recently used, with and without the prototype. Results indicate that the use of DMTR reduced both the number of key strokes and the retrieval time significantly. The majority of participants indicated that it was easier for them to access their contacts using DMTR and that they would like to use it in their next mobile phone. The results provide strong evidence for the demotion principle suggested by the user-subjective approach.     

SNEFL: Social network explicit fuzzy like dataset and its application for Incel detection

In this paper, with respect to reviewing and comparing existing social networks’ datasets, we introduce SNEFL dataset: the first social network dataset that includes the level of users’ likes (fuzzy like) data in addition to the likes between users. With users’ privacy in mind, the data has been collected from a social network. It includes several additional features including age, gender, marital status, height, weight, educational level and religiosity of the users. We have described its structure, analysed its features and evaluated its advantages in comparison with other social network datasets. On top of that, using unique feature of SNEFL dataset (fuzzy like) for the first time a rule-based algorithm has been developed to detect involuntary celibates (Incels) in social networks. Despite Incels activities in online social networks, until now no study on computer science has been performed to identify them. This study is the first step to address this challenge that society is facing today. Experimental results show that the accuracy of the proposed algorithm in identifying Incels among all social network users is 23.21% and among users who have fuzzy like data is 68.75%. In addition to the Incel detection, SNEFL dataset can be used by researchers in different fields to produce more accurate results. Some study areas that SNEFL dataset can be used in are network analysis, frequent pattern mining, classification and clustering.

     

基于位置信息的流式移动应用推送系统的研究

目前与位置相关的移动应用越来越多,传统应用分发模式中,用户需要手动对应用进行搜索、下载、安装以及卸载,不利于提升用户使用应用服务的体验。设计并实现了基于位置信息的流式移动应用推送系统,该系统中服务器利用移动终端位置信息将相关应用解析、安装,并推送到移动终端显示,移动终端根据用户的选择从服务器流式加载应用。该系统使得用户在切换位置时,不用下载、安装即可使用和当前位置相关的应用服务。实验表明,和传统应用分发模式相比,在3G网络环境下可以减少64.37%的应用获取时延,在4G网络环境下可以减少74.49%的应用获取时延。     

Stealing PINs via mobile sensors: actual risk versus user perception

In this paper, we present the actual risks of stealing user PINs by using mobile sensors versus the perceived risks by users. First, we propose PINlogger.js which is a JavaScript-based side channel attack revealing user PINs on an Android mobile phone. In this attack, once the user visits a website controlled by an attacker, the JavaScript code embedded in the web page starts listening to the motion and orientation sensor streams without needing any permission from the user. By analysing these streams, it infers the user’s PIN using an artificial neural network. Based on a test set of fifty 4-digit PINs, PINlogger.js is able to correctly identify PINs in the first attempt with a success rate of 74% which increases to 86 and 94% in the second and third attempts, respectively. The high success rates of stealing user PINs on mobile devices via JavaScript indicate a serious threat to user security. With the technical understanding of the information leakage caused by mobile phone sensors, we then study users’ perception of the risks associated with these sensors. We design user studies to measure the general familiarity with different sensors and their functionality, and to investigate how concerned users are about their PIN being discovered by an app that has access to all these sensors. Our studies show that there is significant disparity between the actual and perceived levels of threat with regard to the compromise of the user PIN. We confirm our results by interviewing our participants using two different approaches, within-subject and between-subject, and compare the results. We discuss how this observation, along with other factors, renders many academic and industry solutions ineffective in preventing such side channel attacks.     

From the private into the public: privacy-respecting mobile interaction techniques for sharing data on surfaces

Interactive horizontal surfaces provide large semi-public or public displays for colocated collaboration. In many cases, users want to show, discuss, and copy personal information or media, which are typically stored on their mobile phones, on such a surface. This paper presents three novel direct interaction techniques (Select&Place2Share, Select&Touch2Share, and Shield&Share) that allow users to select in private which information they want to share on the surface. All techniques are based on physical contact between mobile phone and surface. Users touch the surface with their phone or place it on the surface to determine the location for information or media to be shared. We compared these three techniques with the most frequently reported approach that immediately shows all media files on the table after placing the phone on a shared surface. The results of our user study show that such privacy-preserving techniques are considered as crucial in this context and highlight in particular the advantages of Select&Place2Share and Select&Touch2Share in terms of user preferences, task load, and task completion time.