Search

The emergence of mobile crowdsensing (MCS) has revolutionized data collection method. As an important means of guaranteeing data quality, user recruitment is critical to sensing task completion. Aiming at the problem of user privacy disclosure in user recruitment, particularly when sensing platforms lack prior knowledge of user quality, we propose a Privacy-Preserving User Recruitment scheme (PPUR) which can maximize sensing quality in a lightweight and efficient manner. We design multiple secure protocols for both user quality calculation and user recruitment based on additive secret sharing (ASS). Specifically, we propose Secure user Quality Calculation (SQC) protocol to assess user quality instead of requiring user interaction in the case of unknown ground truth. Combinatorial multi-armed bandit (CMAB) based Secure User Recruitment (SUR) protocol, effectively tackles the challenge of recruiting multiple users without prior knowledge and user interactivity while adhering to budget and time limitations. Theoretical analysis confirms lightweight overhead of the PPUR scheme and its multi-class data security. Experimental results show that SQC has superior performance in both computational cost and communication overhead. The regret indicator's findings demonstrate that SUR can effectively utilize budget and time to achieve optimal user recruitment decision. [ABSTRACT FROM AUTHOR]

User interaction data comprises events that capture individual actions that a user performs on their computer. Such events provide detailed records about how users carry out their tasks in a process, even when this involves different applications. Although the comprehensiveness of such data provides a promising basis for process mining, user interaction events cannot be used directly for this purpose, because they do not meet two essential requirements. In particular, they neither indicate their relation to a process-level activity nor their relation to a specific process execution. Therefore, user interaction data needs to be transformed so that it meets these requirements before process mining techniques can be applied. This transformation problem comprises identifying tasks and their types and determining the relation between tasks and process executions. While some existing approaches tackle parts of this problem, none address it comprehensively. Therefore, we propose an unsupervised approach for recognizing task-level events from user interaction data that addresses it in full. It segments user interaction data to identify tasks, categorizes these according to their type, and relates tasks to each other via object instances it extracts from the user interaction events. In this manner, our approach creates task-level events that meet the requirements of process mining settings. Our evaluation demonstrates the approach's efficacy and shows that its combined consideration of control-flow, data, and semantic information allows it to outperform baseline approaches in both online and offline settings. [ABSTRACT FROM AUTHOR]

The construction of high-precision machine learning models relies on large-scale data collection from IoT devices. Since the training data involves sensitive user information, it is essential to design a privacy-preserving machine learning (PPML) paradigm to obtain reliable models without leaking data. Due to the diversity of IoT devices, the collected data from multiple sources often has different attributes, namely vertically partitioned data. PPML over vertically settings (abbreviated as VPPML) is more challenging than horizontal cases since VPPML requires both secure feature integration and sample aggregation. Most of existing VPPML approaches require multi-round interactions among multiple users and servers, therefore exhibiting high computation and communication overheads. To close this gap, we propose an efficient framework VPPLR to realize the security training for logistic regression. We employ a well-designed structure to reformulate the gradient update rules, ensuring that any user can go offline after uploading the rearranged local data to the cloud server in secret sharing. In the training phase, we introduce vectorization approach to complete the global parameter update in the plain-domain through common channel, thereby going beyond approaches depending on homomorphic encryption in efficiency. We illustrate the performance of our VPPLR in developing classifiers using private IoT data in vertically partitioned scenarios by comparing it with the advanced PPLR methods. [ABSTRACT FROM AUTHOR]

Musculoskeletal disorders (MSDs) account for almost 70 million physician office visits per year in the United States and are the most common workplace injuries. These are conditions involving the nerves, tendons, muscles, and supporting structures of the body. Previous studies have concluded that computer users are at high risk of developing work-related musculoskeletal disorders (WRMSDs). As computer users, medical dosimetrists are at risk of developing WRMSDs, yet there is a lack of information regarding the incidence of WRMSDs among medical dosimetrists. The purpose of this study was to determine the incidence of WRMSDs and variables of workstation ergonomics that contribute to the increased risk of WRMSDs in medical dosimetrists. A Qualtrics survey was created to support the 3 research questions guiding this study. The survey was distributed to 2,646 full members of the American Association of Medical Dosimetrists (AAMD), which included only certified medical dosimetrists (CMDs), via email. The distribution of email surveys sent through the AAMD email distribution list resulted in 988 emails opened, for a contact rate of 37% (988/2646). One hundred sixty-four responses were recorded yielding a completion rate of 17% (164/988). Fifty-five percent (90/163) of participants responded that they have experienced WRMSDs. Forty-four percent (289/652) of responses indicated WRMSDs have a slight or moderate interference on work. Sixty-two percent (94/152) of participants felt that their workstations were not ergonomically designed; even greater 68% (104/153) did not feel their workstations were designed for their individually needs. Of those respondents 64% (98/152) would like to see further adaptations made to their workspaces. [ABSTRACT FROM AUTHOR]

A Network Intrusion Detection System (NIDS) is a widely used security monitoring technology for detecting attacks against network services, beaconing activity of infected end user nodes, malware propagation, and other types of malicious network traffic. Unfortunately, NIDS technologies are known to generate a large number of alerts, with a significant proportion of them having low importance. During the last two decades, many machine learning and data mining based approaches have been proposed for highlighting high-importance alerts that require human attention. However, NIDS alert classification systems based on active learning have received marginal attention in the specialized research literature. This neglects the potential benefits of active learning which involves a human expert in the machine learning model life cycle. The current paper fills this research gap and studies the use of active learning techniques for NIDS alert classification. [ABSTRACT FROM AUTHOR]

The aim of this randomized controlled study was to explore if an intervention with biofeedback training in combination with ergonomic discussions, could improve working technique and work postures, and reduce pain intensity and perceived exertion in young adult computer users with ongoing neck and upper extremity symptoms. 39 participants were divided into an intervention group and a control group. The intervention consisted of 4 sessions during a three-month period. Working technique, working postures, rated perceived exertion, pain intensity, and duration of computer use were measured at baseline and follow ups after 6 and 12 months. The intervention did not significantly improve working technique and working postures, nor reduce pain intensity and perceived exertion in the intervention group compared to the control group. However, there was a statistically significant reduction in reported pain intensity in the neck/shoulder for the whole group. Also, there was a trend that time spent with computer work without breaks was more reduced in the intervention group than in the control group. • Biofeedback training in combination with ergonomic discussions did not improve work technique and postures. • A trend towards reduced time spent with computer use without breaks. • A reduction in reported pain intensity in the neck/shoulder for both groups. [ABSTRACT FROM AUTHOR]

Presents a study which investigated library anxiety among Israeli Bachelor of Education (B.Ed.) students. Independent variables and their relationship to library anxiety; Level of library anxiety among Israeli B.Ed. students; Discussion.

Background: Sixty-four to 90% of computer users experience symptoms of computer vision syndrome (CVS). People with CVS symptoms experience neck pain (NP), and people with NP can have visual symptoms.Objectives: (1) To examine differences in CVS symptoms in computer users with and without persistent NP. (2) To determine relationships between visual ergonomics, demographics, and CVS in those with and without persistent NP DESIGN: Comparative cross-sectional observational study METHOD: An online survey consisting of multiple questionnaires including the Computer Vision Syndrome Questionnaire (CVS-Q), Visual Symptoms Survey (VSS), and Convergence Insufficiency Symptoms Survey (CISS) was completed by 167 participants.Results: The persistent NP group had significantly higher CVS-Q, VSS, and CISS scores (p <0.01) compared to controls. No relationships were found between visual ergonomics, demographics, and visual symptoms in both groups.Conclusions: People with persitent NP are more likely to present with CVS than controls. Clinicians should consider assessing visual symptoms in people presenting with persisting NP. [ABSTRACT FROM AUTHOR]

Smartphones have become a common tool of daily communication, and a number of text input methods are extensively used. Qwerty and T9 are two universally used smartphone input methods that differ in user experience (UX) when entering phrases with different complexity. Phrase complexity is an indication of the number and type of characters used in a phrase. This study evaluated the impact of both experience and phrase complexity on the UX of Qwerty and T9. Thirty-two participants with different experience levels (high and low) used both Qwerty or T9 to input 40 phrases (both simple and complex phrases). Input time, speed, error rate, galvanic skin response, heart rate, and perceived usability were analyzed as dependent variables. The results indicated that Qwerty was more effective than T9 for entering complex phrases, while no obvious difference was found for simple phrases. Experience had a positive impact on the input method, and Qwerty was more helpful for participants with low experience level. Performance, physiological reaction, and perceived usability data showed similar results. These findings aid the optimal design of text input systems for smartphones. [ABSTRACT FROM AUTHOR]

This paper presents an analysis of human-human-computer triads (HHC triads) in institutional settings, more specifically of the way in which advisor, customer and computer cooperatively generate the maximum loan amount in mortgage orientation consultations (MOCs). HHC triads involve a participation framework in which three actors collaborate in order to fulfill a specific consultation goal. We argue that the computer, despite its restricted interactional repertoire, is a full participant in this framework, considered in terms of the turn-taking system. First, we will focus on the global organization of HHC triads, on how these triads emerge in the course of institutional interactions and how they are closed. Second, we analyze the interactional dynamics of HHC triads in the MOC. We demonstrate how turn-taking rules apply, how turn-taking is managed by the participants and how the participants, including the computer, collaborate to perform a task. In the final section we will describe the local sequential structure of HHC triads and identify the actions of the participants, including the computer. The use of HHC-triads not only facilitates easy transitions for the advisors between activities in the consultation, it also cleverly combines the advisor's computer use and customer-centeredness. As the computer becomes an actual "participant" in this institutional goal-oriented participation framework, both advisors and customers show that they make use of the interactional affordances the framework creates. • The analysis is based on a unique corpus of mortgage orientation consultations • We identified a triadic participation framework of human-human-computer interaction • In human-human-computer triads the computer is essential for fulfilling a specific conversational goal • We demonstrate the global sequence organization of Human-Human-Computer triads • We demonstrate the local sequential structures within Human-Human-Computer triads [ABSTRACT FROM AUTHOR]

the need for a reliable and complementary identifier mechanism in a digital forensic analysis is the focus of this study. Mouse dynamics have been applied in information security studies, particularly, continuous authentication and authorization. However, the method applied in security is void of specific behavioral signature of a user, which inhibits its applicability in digital forensic science. This study investigated the likelihood of the observation of a unique signature from mouse dynamics of a computer user. An initial mouse path model was developed using non-finite automata. Thereafter, a set-theory based adaptive two-stage hash function and a multi-stage rule-based semantic algorithm were developed to observe the feasibility of a unique signature for forensic usage. An experimental process which comprises three existing mouse dynamics datasets were used to evaluate the applicability of the developed mechanism. The result showed a low likelihood of extracting unique behavioral signature which can be used in a user attribution process. Whilst digital forensic readiness mechanism could be a potential approach that can be used to achieve a reliable behavioral biometrics modality, the lack of unique signature presents a limitation. In addition, the result supports the logic that the current state of behavioral biometric modality, particularly mouse dynamics, is not suitable for forensic usage. Hence, the study concluded that whilst mouse dynamics-based behavioral biometrics may be a complementary modality in security studies, more will be required to adopt it as a forensic modality in litigation. Furthermore, the result from this study finds relevance in other human attributional studies such as user identification in recommender systems, e-commerce, and online profiling systems, where the degree of accuracy is not relatively high. [ABSTRACT FROM AUTHOR]

Security warning is a form of message that is designed to imitate authorized security alerting function to notify, warn and advice user about the consequence of an action by allowing random applications to run on the computer system. On the other hand, regular computer users tend to ignore the security warning that conveys excessive technical messages which are difficult for them to understand and lead to lack of motivation for decision making. Consequently, an online survey has been utilized by a group of 250 participants to gain general insights based on the user's understanding of security warning. This paper describes one main finding utilizing the Chi-Square test. From the finding, it can be suggested that 2 hypotheses were statistically significant in terms of user's understanding and their assessments towards computer security warning. On the other hand, variable of gender indicates with the highest significant values with three statistically significant 0.004, 0.017 and 0.000 respectively The overall results from the test given indicate that for each variable were connected between computer user's skills and their understanding in order to improve security warning in the future. [ABSTRACT FROM AUTHOR]

Highlights • Relational social recommendation provides highly detailed explanations to users. • Users find these explanations to be intriguing and often surprising. • Linking people based on entity associations reveals novel connections of interest. • Recommendation relevance and serendipity may be improved given user feedbacks. Abstract This paper outlines RSR, a relational social recommendation approach applied to a social graph comprised of relational entity profiles. RSR uses information extraction and learning methods to obtain relational facts about persons of interest from the Web, and generates an associative entity-relation social network from their extracted personal profiles. As a case study, we consider the task of peer recommendation at scientific conferences. Given a social graph of scholars, RSR employs graph similarity measures to rank conference participants by their relatedness to a user. Unlike other recommender systems that perform social rankings, RSR provides the user with detailed supporting explanations in the form of relational connecting paths. In a set of user studies, we collected feedbacks from participants onsite of scientific conferences, pertaining to RSR quality of recommendations and explanations. The feedbacks indicate that users appreciate and benefit from RSR explainability features. The feedbacks further indicate on recommendation serendipity using RSR, having it recommend persons of interest who are not apriori known to the user, oftentimes exposing surprising inter-personal associations. Finally, we outline and assess potential gains in recommendation relevance and serendipity using path-based relational learning within RSR. [ABSTRACT FROM AUTHOR]

When investigating an incident it is important to document user activity, and to document which storage device was connected to which computer. We present a new approach to documenting user activity in computer systems using the NTFS file system by using the $ObjId Index to document user activity, and to correlate this index with the corresponding records in the MFT table. This may be the only possible approach when investigating external NTFS storage devices, and is hence a valuable addition to the storage forensics toolbox. [ABSTRACT FROM AUTHOR]

Abstract With the spectacular development of pervasive computing and the widespread use of mobile technologies, the user usually resorts to a wide variety of user platforms. Moreover, the applications deployed on such platforms usually lack specific support for people suffering from disabilities. This is especially valid in the Tunisian context where the current status of digital accessibility is far from being perfect since the developer's community still ignores accessibility-related programming techniques. Consequently, further efforts are still required to resolve the issues related to the growing variety of available devices and of accessibility difficulties met by different categories of users. Therefore, in this paper we are introducing a model driven approach that aims to assist local and worldwide communities of developers throughout the development process of accessible user interfaces (UIs) targeting several emerging platforms. Based on a set of models from different abstraction levels followed by transformations, this approach provides a generic framework that integrates accessibility concerns starting from the design time. It allows the generation of accessible interfaces with a focus on the graphical and adaptation aspects on various platforms regardless of the developer expertise in UI accessibility. 1 Highlights • Improving the development of accessible applications for developers worldwide. • Metamodels of accessible user interface, interaction platforms, execution platforms. • Transformations from abstract accessible UI models to executable UIs. • Development of model editors to assist users create required models. • Evaluation of the framework to show the usability and efficiency of the approach. [ABSTRACT FROM AUTHOR]

Abstract Currently there are tools that support the customisation of users' gestures. In general, the inclusion of new gestures implies writing new lines of code that strongly depend on the target platform where the system is run. In order to avoid this platform dependency, gestUI was proposed as a model-driven method that permits (i) the definition of custom touch-based gestures, and (ii) the inclusion of the gesture-based interaction in existing user interfaces on desktop computing platforms. The objective of this work is to compare gestUI (a MDD method to deal with gestures) versus a code-centric method to include gesture-based interaction in user interfaces. In order to perform the comparison, we analyse usability through effectiveness, efficiency and satisfaction. Satisfaction can be measured using the subjects' perceived ease of use, perceived usefulness and intention to use. The experiment was carried out by 21 subjects, who are computer science M.Sc. and Ph.D. students. We use a crossover design, where each subject applied both methods to perform the experiment. Subjects performed tasks related to custom gesture definition and modification of the source code of the user interface to include gesture-based interaction. The data was collected using questionnaires and analysed using non-parametric statistical tests. The results show that gestUI is more efficient and effective. Moreover, results conclude that gestUI is perceived as easier to use than the code-centric method. According to these results, gestUI is a promising method to define custom gestures and to include gesture-based interaction in existing user interfaces of desktop-computing software systems. Highlights • gestUI, a model-driven method includes gesture-based interaction in user interface. • Also, customised gestures are defined and included in user interfaces using gestUI. • A comparative empirical evaluation is used to evaluate the usability of gestUI. • We compare gestUI with a code-centric method in the empirical evaluation. • Usability is evaluated in terms of efficiency, effectiveness and intention of use. [ABSTRACT FROM AUTHOR]

Abstract Traditional recommender systems rely on user profiling based on either user ratings or reviews through bi-sentimental analysis. However, in real-world scenarios, there are two common phenomena: (1) users only provide ratings for items but without detailed review comments. As a result, the historical transaction data available for recommender systems are usually unbalanced and sparse; (2) in many cases, users' opinions can be better grasped in their reviews than ratings. For the reason that there is always a bias between ratings and reviews, it is really important that users' ratings and reviews should be mutually reinforced to grasp the users' true opinions. To this end, in this paper, we develop an opinion mining model based on convolutional neural networks for enhancing recommendation. Specifically, we exploit two-step training neural networks, which utilize both reviews and ratings to grasp users' true opinions in unbalanced data. Moreover, we propose a S entiment C lassification scoring (SC) method, which employs dual attention vectors to predict the users' sentiment scores of their reviews rather than using bi-sentiment analysis. Next, a combination function is designed to use the results of SC and user–item rating matrix to catch the opinion bias. It can filter the reviews and users, and build an enhanced user–item matrix. Finally, a M ultilayer perceptron based M atrix F actorization (MMF) method is proposed to make recommendations with the enhanced user–item matrix. Extensive experiments on several real-world datasets (Yelp, Amazon, Taobao and Jingdong) demonstrate that (1) our approach can achieve a superior performance over state-of-the-art baselines; (2) our approach is able to tackle unbalanced data and achieve stable performances. [ABSTRACT FROM AUTHOR]

Abstract Analysis of network users' behaviors is an important part for improving network security. This paper analyzes users' group behavior via their interactions within the Office Application (OA) system. Specifically, we construct a heterogeneous information network model based on the interactive messages among users in the OA system. The model contains two types of nodes: user and topic nodes, and relationships between users and topics that are encoded in matrices. We then elicit several meta paths in the model, which integrates both user interaction and semantic information, and propose a community division method for group user behavior analysis. To verify the proposed method, we conduct experiments on five-year data collected from a real OA system. The results evaluated by an improved indicator named Normalized Mutual Information show the effectiveness of our work. Highlights • A new analysis method of network users' behaviors has been proposed. • This method is based on the heterogeneous information network model. • Both user interaction and semantic information are integrated in the model for group user behavior analysis. • Five-year data collected from a real OA system are studied in the experiments. [ABSTRACT FROM AUTHOR]

Abstract Sine Cosine Algorithm is a recently developed algorithm based on the characteristics of sine and cosine trigonometric functions, to solve global optimization problems. This paper introduces a novel improved version of sine cosine algorithm, which enhances the exploitation ability of solutions and reduces the overflow of diversity present in the search equations of classical SCA. The proposed algorithm is named as ISCA. The key feature in the proposed algorithm is the hybridization of exploitation skills of crossover with personal best state of individual solutions and integration of self-learning and global search mechanisms. To evaluate these skills in ISCA, a classical set of well-known benchmark problems, standard IEEE CEC 2014 benchmark test and a recent set of benchmark problems, IEEE CEC 2017 have been taken. Several performance metrics (such as convergence, statistical test, performance index), employed on ISCA, ensure the robustness and efficiency of the algorithm. In the paper, the proposed algorithm ISCA is also used to solve five well-known engineering optimization problems. At the end of the paper, the proposed algorithm is also used for multilevel thresholding in image segmentation. The numerical experiments and analysis demonstrate that the proposed algorithm (ISCA) can be highly effective in solving real-life optimization problems. Highlights • A new method called ISCA is proposed for global optimization problems. • The ISCA improves the SCA using crossover and personal best memory of agents. • The classical, CEC 2014 and CEC 2017 benchmarks are used to examine ISCA. • The ISCA is used for engineering problems and image thresholding problem. • Comparisons illustrate the improvement on the performance of ISCA. [ABSTRACT FROM AUTHOR]

Abstract Social networks are the most preferred mean for the people to communicate. Therefore, it is quite usual that experts use them to carry out Group Decision Making processes. One disadvantage that recent Group Decision Making methods have is that they do not allow the experts to use free text to express themselves. On the contrary, they force them to follow a specific user–computer communication structure. This is against social network nature where experts are free to express themselves using their preferred text structure. This paper presents a novel model for experts to carry out Group Decision Making processes using free text and alternatives pairwise comparisons. The main advantage of this method is that it is designed to work using social networks. Sentiment analysis procedures are used to analyze free texts and extract the preferences that the experts provide about the alternatives. Also, our method introduces two ways of applying consensus measures over the Group Decision Making process. They can be used to determine if the experts agree among them or if there are different postures. This way, it is possible to promote the debate in those cases where consensus is low. [ABSTRACT FROM AUTHOR]

Highlights • Social media renders possible a new form of digital-based governance. • A social media-based governance integrates citizens in government decision-making. • Social media unveils a dialectical relationship between online and offline actions. Abstract This study sets out to analyze the trajectory of the popular demonstrations that occurred in Brazil in June 2013, commonly known as the "20 Cents Movement," in order to understand their relationship to the use of social media. Thus, using Actor-Network Theory (ANT) associated with a single case study, the relevant actors in the formation of a socio-technical network resulting in the popular demonstrations that occurred in Brazil were identified. The study reveals that these demonstrations were comprised of different actors aligned around a single obligatory passage point, namely the social media. It was established that social media played an important role in this popular demonstration, prompting a change in the way citizens interact and unite, making formal organizations less necessary, and broadening the political participation of society. Nevertheless, the misalignment between the global and local network of actors led this uprising to fade out over time. Lastly, implications for practice and theory accrued from the investigation of this social media-enabled movement are set forth. [ABSTRACT FROM AUTHOR]

Highlights • Tested uncertainty avoidance and social media usage as antecedents to engagement. • Severity, susceptibility and emotion serially mediated crisis psychological process. • Uncertainty avoidance's effect on engagement differ by arousal state (high vs. low). • High and low arousal negative emotions have varying engagement activation powers. • Social media usage affected information seeking only through cognitive appraisal. Abstract This study examines the psychological mechanisms underlying the process that enables publics' individual differences (e.g., the levels of uncertainty avoidance and social media usage) to give rise to varying post-crisis social media engagement intentions (e.g., information seeking, support seeking, and negative word-of-mouth). The study confirms that this process is serially mediated by perceived threat severity, perceived susceptibility, and negative emotions. These psychological mediators, however, function differently between uncertainty avoidance and social media usage. For uncertainty avoidance, a high arousal negative emotion (e.g., anger, fear) is a more essential step to influence engagement intentions than a low arousal emotion (e.g., shame, guilt). For social media usage, however, the type of aroused negative emotions does not matter. Rather, the relationship social media usage has with information seeking is different from that with support seeking and negative word-of-mouth intentions. Social media usage induces information seeking intentions only through publics' cognitive appraisals of the situation without the activation of negative emotions, while it induces supporting seeking and negative word-of-mouth intentions through both cognitive appraisals and negative emotion arousals. [ABSTRACT FROM AUTHOR]

Abstract Malware is the umbrella term that denotes attacking any system by malicious software. During the last few years, the popularity of Android smartphones led to the sneak of several malware applications into different Android markets without any difficulty. As a consequence of this, malware applications have been grown exponentially in the Android markets. Unfortunately, most of these markets suffer from an inability to detect malware, which results in increasing the probability of infecting users' phones with these malware applications. The present paper focuses on developing an efficient computational framework based on Deep Belief Networks for malware detection. The proposed framework merges high level static analysis, dynamic analysis and system calls in feature extraction in order to achieve the highest accuracy. The evaluation compares the most familiar machine learning approaches that were applied in malware detection with the proposed framework. The obtained results demonstrate that Deep Belief Networks technique can realize 99.1% accuracy with the presented dataset. Over and above that, we develop our complete static analysis jar which adopts different efficient methods in an attempt to facilitate and speed up the static analysis by handling all the Android applications in only one step rather than considering one application at a time. [ABSTRACT FROM AUTHOR]

Ensuring scalability of recommender systems without sacrificing the quality of the recommendations produced, presents significant challenges, especially in the large-scale, real-world setting of mobile ad targeting. In this paper, we propose MobRec, a novel two-stage user similarity based approach to recommendation which combines information provided by slowly-changing features of the mobile context and implicit user feedback indicative of user preferences. MobRec uses the contextual features to cluster, during an off-line stage, users that share similar patterns of mobile behavior. In the online stage, MobRec focuses on the cluster consisting of users that are most similar to the target user in terms of their contextual features as well as implicit feedback. MobRec also employs a novel strategy for robust estimation of user preferences from noisy clicks. Results of experiments using a large-scale real-world mobile advertising dataset demonstrate that MobRec outperforms the state-of-the-art neighborhood-based as well as latent factor-based recommender systems, in terms of both scalability and the quality of the recommendations. [ABSTRACT FROM AUTHOR]

Sentiment summarization is the process of automatically creating a compressed version of the opinionated information expressed in a text. This paper presents a machine learning-based approach to summarize user's opinion expressed in reviews using: (1) Sentiment knowledge to calculate a sentence sentiment score as one of the features for sentence-level classification. It integrates multiple strategies to tackle the following problems: sentiment shifter, the types of sentences and word coverage limit. (2) Word embedding model, a deep-learning-inspired method to understand meaning and semantic relationships among words and to extract a vector representation for each word. (3) Statistical and linguistic knowledge to determine salient sentences. The proposed method combines several types of features into a unified feature set to design a more accurate classification system ( “True”: the extractive reference summary; “False”: otherwise ). Thus, to achieve better performance scores, we carried out a performance study of four well-known feature selection techniques and seven of the most famous classifiers to select the most relevant set of features and find an efficient machine learning classifier, respectively. The proposed method is applied to three different datasets and the results show the integration of support vector machine-based classification method and Information Gain (IG) as a feature selection technique can significantly improve the performance and make the method comparable to other existing methods. Furthermore, our method that learns from this unified feature set can obtain better performance than one that learns from a feature subset. [ABSTRACT FROM AUTHOR]

As one kind of typical network big data, social-information networks (such as Weibo and Twitter) include both the complex network structure among users and the rich microblog/tweets information published by users. Understanding the interplay of rich content and social relationships is potentially valuable to the fundamental network mining task, i.e. the link prediction. Although some of the link prediction methods have been proposed by combining topological and non-topological information simultaneously, the in-depth analysis of the rich content still being in a minority, and the rich content in the social-information networks is still underused in solving link prediction. In this paper, we approach the link prediction problem in social-information network by combining network structure and topic information which is extracted from users’ rich content. We first define a kind of user-to-user topic inclusion degree (TID) based on the dissemination mechanism of the published content in the social-information networks, and then construct a TID-based sparse network. On the basis, we build a fusion probabilistic matrix factorization model which solves the link prediction problem by fusing the information of the original following/followed network and the TID-based network in a unified probabilistic matrix factorization framework. We conduct link prediction experiments on two types of real social-information network datasets, i.e. Twitter and Weibo. The experimental results demonstrate that the proposed method is more effective in solving the link prediction problem in social-information networks. [ABSTRACT FROM AUTHOR]

Mobile crowdsensing is a recent model in which a group of mobile users uses their smart devices such as smartphones or wearable devices to cooperatively perform a large-scale sensing task. In this paper, a novel model will be introduced for recognizing/classifying human activities that were collected from sensor units on the chest, legs, and arms. The proposed model employed the k -Nearest Neighbor ( k -NN) classifier which is one of the most common classifiers. k -NN has only one parameter, k , to determine the number of selected nearest neighbors to the test or unknown samples for predicting the class labels of the unknown samples. Searching for the value of k which has a great impact on the classification performance is difficult especially with high dimensional data. This paper employs the Particle Swarm Optimization (PSO) algorithm to search for the optimal value of the k parameter in the k -NN Classifier. This paper shows first experimentally how the PSO in the proposed algorithm searches for the optimal value of k parameter to reduce the misclassification rate of the k -NN classifier. Then, in the second experiment, ten standard datasets are utilized to benchmark the performance of the proposed algorithm. For verification, the results of the PSO- k NN algorithm are compared with two well-known algorithms: Genetic Algorithm (GA) and Ant Bee Colony Optimization (ABCO). In the third experiment, the proposed PSO- k NN algorithm was employed for recognizing human activities. The experimental results proved that the PSO- k NN algorithm is able to find the optimal or near optimal value(s) of the k parameter which enhances the accuracy of k -NN classifier. The results also demonstrated lower error rates compared when GA and ABCO algorithms. [ABSTRACT FROM AUTHOR]

We present a tag-aware dynamic music recommendation framework that achieves personalized and accurate music recommendations to users. The proposed framework leverages the available semantic labels (in terms of tags) of music tracks to complement a highly sparse user-item interaction matrix, which effectively addresses the data sparsity issue faced by most music recommendation systems. Music tracks are more accurately represented by aggregating the latent factors derived from both the tag space and the user interaction information. The proposed framework further employs a Gaussian state-space model to capture the evolving nature of users’ preferences over time, which helps achieve time-sensitive recommendation of music. A variational approximation is developed to achieve fast inference and learning of model parameters. Experiments conducted using actual music data and comparison with state-of-the-art competitive recommendation algorithms help demonstrate the effectiveness of the proposed framework. [ABSTRACT FROM AUTHOR]

Due to the widespread use of mobile devices, services based on the users current indoor location are growing in significance. Such services are developed in the Machine Learning and Experst Systems realm, and ranges from guidance for blind people to mobile tourism and indoor shopping. One of the most used techniques for indoor positioning is WiFi fingerprinting, being its use of widespread WiFi signals one of the main reasons for its popularity, mostly on high populated urban areas. Most issues of this approach rely on the data acquisition phase; to manually sample WiFi RSSI signals in order to create a WiFi radio map is a high time consuming task, also subject to re-calibrations, because any change in the environment might affect the signal propagation, and therefore degrade the performance of the positioning system. The work presented in this paper aims at substituting the manual data acquisition phase by directly calculating the WiFi radio map by means of a radiosity signal propagation model. The time needed to acquire the WiFi radio map by means of the radiosity model dramatically reduces from hours to minutes when compared with manual acquisition. The proposed method is able to produce competitive results, in terms of accuracy, when compared with manual sampling, which can help domain experts develop services based on location faster. [ABSTRACT FROM AUTHOR]

With the influence and social ripple effect of social media sites, diverse studies are in progress to analyze the contents generated by users. Numerous contents generated in real time contain information about social issues and events such as natural disasters. In particular, users show not only information about the events that occurred but also their sentiments. In this paper, we propose Polaris, a system for analyzing and predicting users’ sentimental trajectories for events analyzed in real time out of the massive social media contents, and show the results of preliminary validation work that we have done. We show both trajectory analysis and sentiment analysis so that users can obtain the insight at a glance. Also, we increased the accuracy in sentiment analysis and prediction by making use of the latest deep-learning technique. [ABSTRACT FROM AUTHOR]

Abstract User-centered design (UCD) as a concept has begun to enter composition studies, particularly through scholar-teachers of technical communication. We question further incorporation—specifically, the collision course of industry-driven language such as "efficiency" and "expediency" and the potential positioning of students as "users" in the composition classroom. We argue that this positioning places us in unproductive opposition to multimodal composition. Rather than a wholesale incorporation of UCD into the composition classroom, we outline a "theory + play" approach that combines scholarship in rhetoric, speculative design, and multimodal composition. This approach, we argue, better aligns with the political and social investments of Johnson's (1998) theory of user-centered technology, in which our students are critical makers and engaged citizens in the public sphere. [ABSTRACT FROM AUTHOR]

Abstract While usability testing can help instructors improve the design of online writing instruction (OWI), its emphasis on student-users sometimes overlooks the networked rhetorical ecologies in which those student-users operate, tilting online composition pedagogy toward neoliberal models of higher education that cater to the student-as-consumer. In response, I propose augmenting usability testing in online writing education with the theories of digital rhetoric. In the context of OWI, usability theory and digital rhetoric share a similar emphasis on the student as a user or audience, but they also have at least two key differences. First, unlike digital rhetoric, usability testing typically elides the political and ideological implications of student-users' experiences. Second, the usability theories of Jakob Nielsen, for example, tend to view online interfaces as static objects manipulated by users. Digital rhetoric, on the other hand, sees interfaces as dynamic, real-time interactions. Although both paradigms have their advantages, I argue that integrating usability testing with the theories of digital rhetoric can add complexity to researchers' understanding of OWI by revealing not only how students use online writing environments but also how they use them rhetorically. [ABSTRACT FROM AUTHOR]

Emerging technologies such as virtual reality and robots are evolving to increasingly integrate the user into the interface. During this temporary merge, users experience a digital or a robotic body of an avatar as their own. Embodied cognition sustains that the body and its interactions with the environment play an important role in cognition. I argue that the adoption of mediated embodiment technologies to explore cognitive development might substantially contribute to demonstrating the postulates of embodied cognition. [ABSTRACT FROM AUTHOR]

Abstract With the development of cloud technology, sharing date between users securely becomes more and more important. Moreover, the cloud server should provide a flexible control on sharing date. This work presents a notion named fuzzy conditional broadcast proxy re-encryption (FC-BPRE), in which a semi-honest proxy can transform a delegator’s ciphertext to a new ciphertext for a set of delegatees when the ciphertext satisfies a certain condition set by the delegator. However, the proxy cannot learn any useful information about the underlying plaintext. The proxy can convert a ciphertext, encrypted under a condition set W , to a ciphertext encrypted under a condition set W ′ using a broadcast re-encryption key, if and only if S and S ′ are close to each other. We first formalize the notion of FC-PBRE scheme and then proposed a construction. [ABSTRACT FROM AUTHOR]

In recent years, blogging research has grown rapidly in social networks and the number of posts has continued to grow. An effective search method for these growing posts is to use a blog search engine to help bloggers quickly and accurately find the information they need. The blog search engine faces an important problem of the short query entered by the user like the general search engine. This problem makes it difficult for search engines to correctly define the nature of user queries. Relevant literature shows that many researchers have tried to solve this problem by using different semantic analysis models. However, these models are not suitable for big data environments such as search engines because they need significant computing time. In this paper, we propose a semantic analysis model with a dynamic judgment mechanism. According to the experimental results, our model can achieve a cost-effective solution in computing time and execution performance. That is, we can use a relatively small amount of computing time to achieve a near-optimal solution performance. [ABSTRACT FROM AUTHOR]

A cloud’s elasticity is enhanced by inter-operating with other clouds and borrowing resources from them. Hence, a request from any such inter-operating cloud’s user is fulfilled with resources of some individual cloud in the network. However, the request is denied on unavailability of sufficient resources in any individual cloud. In this work, this denial of a user’s request is prevented by obtaining resources from multiple clouds and hence, the probability of satisfying the request is improved. This proposal intends to replace the existing resource provisioning technique in an open framework, Cloud Inter-operation Toolkit ( CIT ), to achieve an increased transaction success rate which is evident from the experimental results obtained from a real-time heterogeneous cloud environment set up using Eucalyptus, OpenNebula and OpenStack. In spite of satisfying a request with resources from multiple clouds, these resources are aggregated and provided in a way that is easily accessible by the user. [ABSTRACT FROM AUTHOR]

The wide adoption of social networking and microblogging platforms by a large number of users across the globe has provided a rich source of unstructured information for understanding users’ behaviors, interests and opinions at both micro and macro levels. An active area in this space is the detection of important real-world events from user-generated social content. The works in this area identify instances of events that impact a large number of users. However, a more nuanced form of an event, known as life event, is also of high importance, which in contrast to real-world events, does not impact a large number of users and is limited to at most a few people. For this reason, life events, such as marriage, travel, and career change, among others, are more difficult to detect for several reasons: i) they are specific to a given user and do not have a wider reaching reflection; ii) they are often not reported directly and need to be inferred from the content posted by individual users; and iii) many users do not report their life events on social platforms, making the problem highly class-imbalanced. In this paper, we propose a semantic approach based on word embedding techniques to model life events. We then use word mover’s distance to measure the similarity of a given tweet to different types of life events, which are used as input features for a multi-class classifier. Furthermore, we show that when a sequence of tweets that have appeared before and after a given tweet of interest (temporal stacking) are considered, the performance of the life event detection task improves significantly. [ABSTRACT FROM AUTHOR]

Collaborative filtering has become one of the most widely used methods for providing recommendations in various online environments. Its recommendation accuracy highly relies on the selection of appropriate neighbors for the target user/item. However, existing neighbor selection schemes have some inevitable inadequacies, such as neglecting users’ capability of providing trustworthy recommendations, and ignoring users’ preference changes. Such inadequacies may lead to drop of the recommendation accuracy, especially when recommender systems are facing the data sparseness issue caused by the dramatic increase of users and items. To improve the recommendation accuracy, we propose a novel two-layer neighbor selection scheme that takes users’ capability and trustworthiness into account. In particular, the proposed scheme consists of two modules: (1) capability module that selects the first layer neighbors based on their capability of providing recommendations and (2) a trust module that further identifies the second layer neighbors based on their dynamic trustworthiness on recommendations. The performance of the proposed scheme is validated through experiments on real user datasets. Compared to three existing neighbor selection schemes, the proposed scheme consistently achieves the highest recommendation accuracy across data sets with different degrees of sparseness. [ABSTRACT FROM AUTHOR]

Assistant robots are designed to perform specific tasks for the user, but their performance is rarely optimal, hence they are required to adapt to user preferences or new task requirements. In the previous work, the potential of an interactive learning framework based on user intervention and reinforcement learning (RL) was assessed. The framework allowed the user to correct an unfitted segment of the robot trajectory by using hand movements to guide the robot along a corrective path. So far, only the usability of the framework was evaluated through experiments with users. In the current work, the framework is described in detail and its ability to learn from a set of sample trajectories using an RL algorithm is analyzed. To evaluate the learning performance, three versions of the framework are proposed that differ in the method used to obtain the sample trajectories, which are: human-guided learning, autonomous learning, and combined human-guided with autonomous learning. The results show that the combination of the human-guided and autonomous learning achieved the best performance, and although it needed a higher number of sample trajectories than the human-guided learning, it required less user involvement. Autonomous learning alone obtained the lowest reward value and needed the highest number of sample trajectories. [ABSTRACT FROM AUTHOR]

Online social networks have become the hotbeds of many rumors as information can propagate much faster than ever. In order to detect the few but potentially harmful rumors to prevent the public issues they may cause, we propose an unsupervised learning model combining Recurrent Neural Networks and Autoencoders to distinguish rumors as anomalies from other credible microblogs based on users’ behaviors. Some features based on comments posted by other users are newly proposed and are then analyzed over their posting time so as to exploit the crowd wisdom to improve the detection performance. The experimental results show that our model achieves a high accuracy of 92.49% and F1 measure of 89.16%. [ABSTRACT FROM AUTHOR]

Identifying online user reputation is significant for online social systems. In this paper, taking into account the preference physics of online user collective behaviors, we present an improved group-based rating method for ranking online user reputation based on the user preference (PGR). All the ratings given by each specific user are mapped to the same rating criteria. By grouping users according to their mapped ratings, the online user reputation is calculated based on the corresponding group sizes. Results for MovieLens and Netflix data sets show that the AUC values of the PGR method can reach 0.9842 (0.9493) and 0.9995 (0.9987) for malicious (random) spammers, respectively, outperforming the results generated by the traditional group-based method, which indicates that the online preference plays an important role for measuring user reputation. [ABSTRACT FROM AUTHOR]

From web search and data mining, users’ click and purchase behaviors contain valuable information, thus numerous approaches have been proposed to identify embedded useful knowledge from them. In these real-life situations, each user may perform the same action/event multiple times, and multiple accessed events product different profit. Many utility-oriented data mining approaches thus have been extensively studied. Previous studies have the limitation that the overall utility of traditional pattern is limited since they rarely consider the inherent correlation. For example, from the purchase behavior, the low-utility patterns sometimes with a very high-utility pattern will be considered as a valuable pattern even if this behavior may be not highly correlated. A more intelligent framework that provides non-redundant and correlated behavior based on utility measure is thus desired. In this paper, we first present a novel method to extract non-redundant correlated purchase behaviors considering the utility and correlation factors. The high qualified patterns can be derived with high profit and strong correlation, which can lead to higher recall and reveal better precision. In the proposed projection-based approach, an efficient projection mechanism and a sorted downward closure property are developed to reduce the database size. Several pruning strategies are further developed to efficiently and effectively discover the desired patterns. An extensive experimental study showed that the novel non-redundant correlated high-utility pattern has more effectiveness than the previous knowledge representation. Moreover, the proposed algorithm is efficient in terms of execution time and memory usage. [ABSTRACT FROM AUTHOR]

The increased popularity of event-based social networks (EBSNs) connects online social communities with offline event activities, and makes it interesting and necessary to understand users’ event participation behaviors on this new type of online social platform, especially when groups are explicitly defined as event organizers and have declared memberships from individual users. Accurate event participation prediction can help guide more effective event participation, event organization, and community development. In this work, using data collected from the popular Meetup.com EBSN spanning three major cities with diverse cultures, we first conduct detailed analysis on group- and user-specified event behaviors. Based on this analysis, we propose a group-based event participation prediction framework that uses personalized random walk with restart on a hybrid EGU (event-group category-user) network to capture intrinsic social relationships, and fuses that with newly designed content and contextual features to predict event participation by group members. Detailed evaluations using the Meetup datasets demonstrate that our prediction frameworks achieves high prediction performance with the proposed features. [ABSTRACT FROM AUTHOR]

This work attempts to recognize the gender of an unknown user with data derived only from keystroke dynamics. Keystroke dynamics, which can be described as the way a user is typing, usually amount to tens of thousands of features, each of them enclosing some information. The question that arises is which of these characteristics are most suitable for gender classification. To answer this question, a new dataset was created by recording users during the daily usage of their computer, the information gain of each keystroke dynamics feature was calculated, and five well-known classification models were used to test the feature sets. The results show that the gender of an unknown user can be identified with an accuracy of over 95% with only a few hundred features. This percentage, which is the highest found in the literature, is quite promising for the development of reliable systems that can alert an unsuspecting user to being a victim of deception. Moreover, having the ability to identify the gender of a user who types a certain piece of text is of significant importance in digital forensics. This holds true, as it could be the source of circumstantial evidence for “putting fingers on the keyboard” and for arbitrating cases where the true origin of a message needs to be identified. [ABSTRACT FROM AUTHOR]

Here, we present IDNet, a user authentication framework from smartphone-acquired motion signals. Its goal is to recognize a target user from their way of walking, using the accelerometer and gyroscope (inertial) signals provided by a commercial smartphone worn in the front pocket of the user’s trousers. IDNet features several innovations including: (i) a robust and smartphone-orientation-independent walking cycle extraction block, (ii) a novel feature extractor based on convolutional neural networks, (iii) a one-class support vector machine to classify walking cycles, and the coherent integration of these into (iv) a multi-stage authentication technique. IDNet is the first system that exploits a deep learning approach as universal feature extractors for gait recognition, and that combines classification results from subsequent walking cycles into a multi-stage decision making framework. Experimental results show the superiority of our approach against state-of-the-art techniques, leading to misclassification rates (either false negatives or positives) smaller than 0.15% with fewer than five walking cycles. Design choices are discussed and motivated throughout, assessing their impact on the user authentication performance. [ABSTRACT FROM AUTHOR]

Recently, it has become difficult for ordinary users to find their interests when facing massive information accompanied by the popularity and development of social networks. The recommendation system is considered to be the most promising way to solve the problem by developing a personalized interest model and pushing potentially interesting content to each user. However, traditional recommendation methods (including collaborative filtering, which is currently the most mature and widely used method) are facing challenges of data sparsity, diversity and more issues that are causing unsatisfactory performance. In this paper, we propose the WLRRS, a new recommendation system based on weighted linear regression models. Compared with traditional methods, the WLRRS has the best predictive accuracy (RMSE) and the best classification accuracy (F-measure) with less fluctuation. WLRRS also provides better time performance compared to the collaborative filtering method, which meets the requirements of the real production environment. [ABSTRACT FROM AUTHOR]

This paper studies the optimization problem of energy efficiency in multi-cell networks with multiple users such as 5G network. However, the spectrum allocation and inter-cell interference (ICI) in multi-cell networks are the important challenges. The Soft Frequency Reuse (SFR) scheme which manages the spectrum and reduces ICI is used to build SFR-cellular networks. We take the global energy efficiency of the SFR-cellular network as the objective function of the optimal problem to obtain the maximum energy efficiency. Unfortunately, the objective function is a non-concave function, which is significantly difficult to be solved directly. Therefore, we utilize the fractional program, successive convex approximation, Lagrange dual, and Karush–Kuhn–Tucker (KKT) conditions to transform the objective function into a concave function. In such case, we could solve the problem with the convex optimization method. Finally, based on SFR, we propose a global energy efficiency optimization algorithm to search for the optimal energy efficiency of networks. Simulation results show that the proposed algorithm holds the better performance. [ABSTRACT FROM AUTHOR]

Sharing cloud resources between groups of users is a challenge. Cloud providers do not commonly support users in sharing their spare dedicated resources with others. In developing countries, it is often too expensive for people to acquire a virtual machine of their own. Users may, therefore, wish to manage costs and increase computational resource usage by sharing their instances with others. This paper presents a container based cloud resource bartering (CRB) model for sharing user's computational resources through a social network. In our approach, we have integrated a Facebook account with the computational cloud to enable tenants to share their unused cloud resources with other users. The performance of the proposed prototype is evaluated under different workloads. Based on our experimental results we conclude that the proposed model is well suited for the creation of a low-cost social cloud in developing countries. [ABSTRACT FROM AUTHOR]

While AI applications are very popular in many domains, they should be harmonious with domain exerts and end users. Furthermore, in order to develop such harmonious AI applications, we need agile AI platforms for not only developers but domain experts. Currently, we have developed Practical Intelligent applications (PRINTEPS), which is a user-centric platform to develop integrated intelligent applications. This paper evaluates the service quality of a multi-robot cafe at a campus festival and reports on the cafe as a practical application of PRINTEPS. It is not clear if robot services are perceived as attractive and/or valuable, and how the concept of robot services could lead to customer satisfaction. Therefore, the evaluation of the quality of such services is necessary to identify the key factors that could contribute to improving customer satisfaction. The objective of this study is to contribute to the realization of practical and economically feasible robot services by obtaining effective knowledge for improving customer satisfaction. This will be achieved through the evaluation of the service quality of the multi-robot cafe. [ABSTRACT FROM AUTHOR]