Showing papers by "Stevens Institute of Technology published in 2015"

A convolutional neural network cascade for face detection

Abstract: In real-world face detection, large visual variations, such as those due to pose, expression, and lighting, demand an advanced discriminative model to accurately differentiate faces from the backgrounds. Consequently, effective models for the problem tend to be computationally prohibitive. To address these two conflicting challenges, we propose a cascade architecture built on convolutional neural networks (CNNs) with very powerful discriminative capability, while maintaining high performance. The proposed CNN cascade operates at multiple resolutions, quickly rejects the background regions in the fast low resolution stages, and carefully evaluates a small number of challenging candidates in the last high resolution stage. To improve localization effectiveness, and reduce the number of candidates at later stages, we introduce a CNN-based calibration stage after each of the detection stages in the cascade. The output of each calibration stage is used to adjust the detection window position for input to the subsequent stage. The proposed method runs at 14 FPS on a single CPU core for VGA-resolution images and 100 FPS using a GPU, and achieves state-of-the-art detection performance on two public face detection benchmarks.

Charge-transfer-based gas sensing using atomic-layer MoS2.

Abstract: Two-dimensional (2D) molybdenum disulphide (MoS2) atomic layers have a strong potential to be used as 2D electronic sensor components. However, intrinsic synthesis challenges have made this task difficult. In addition, the detection mechanisms for gas molecules are not fully understood. Here, we report a high-performance gas sensor constructed using atomic-layered MoS2 synthesised by chemical vapour deposition (CVD). A highly sensitive and selective gas sensor based on the CVD-synthesised MoS2 was developed. In situ photoluminescence characterisation revealed the charge transfer mechanism between the gas molecules and MoS2, which was validated by theoretical calculations. First-principles density functional theory calculations indicated that NO2 and NH3 molecules have negative adsorption energies (i.e., the adsorption processes are exothermic). Thus, NO2 and NH3 molecules are likely to adsorb onto the surface of the MoS2. The in situ PL characterisation of the changes in the peaks corresponding to charged trions and neutral excitons via gas adsorption processes was used to elucidate the mechanisms of charge transfer between the MoS2 and the gas molecules.

Tracking Vital Signs During Sleep Leveraging Off-the-shelf WiFi

Abstract: Tracking human vital signs of breathing and heart rates during sleep is important as it can help to assess the general physical health of a person and provide useful clues for diagnosing possible diseases. Traditional approaches (e.g., Polysomnography (PSG)) are limited to clinic usage. Recent radio frequency (RF) based approaches require specialized devices or dedicated wireless sensors and are only able to track breathing rate. In this work, we propose to track the vital signs of both breathing rate and heart rate during sleep by using off-the-shelf WiFi without any wearable or dedicated devices. Our system re-uses existing WiFi network and exploits the fine-grained channel information to capture the minute movements caused by breathing and heart beats. Our system thus has the potential to be widely deployed and perform continuous long-term monitoring. The developed algorithm makes use of the channel information in both time and frequency domain to estimate breathing and heart rates, and it works well when either individual or two persons are in bed. Our extensive experiments demonstrate that our system can accurately capture vital signs during sleep under realistic settings, and achieve comparable or even better performance comparing to traditional and existing approaches, which is a strong indication of providing non-invasive, continuous fine-grained vital signs monitoring without any additional cost.

In situ structural analysis of the human nuclear pore complex

Abstract: Nuclear pore complexes are fundamental components of all eukaryotic cells that mediate nucleocytoplasmic exchange. Determining their 110-megadalton structure imposes a formidable challenge and requires in situ structural biology approaches. Of approximately 30 nucleoporins (Nups), 15 are structured and form the Y and inner-ring complexes. These two major scaffolding modules assemble in multiple copies into an eight-fold rotationally symmetric structure that fuses the inner and outer nuclear membranes to form a central channel of ~60 nm in diameter. The scaffold is decorated with transport-channel Nups that often contain phenylalanine-repeat sequences and mediate the interaction with cargo complexes. Although the architectural arrangement of parts of the Y complex has been elucidated, it is unclear how exactly it oligomerizes in situ. Here we combine cryo-electron tomography with mass spectrometry, biochemical analysis, perturbation experiments and structural modelling to generate, to our knowledge, the most comprehensive architectural model of the human nuclear pore complex to date. Our data suggest previously unknown protein interfaces across Y complexes and to inner-ring complex members. We show that the transport-channel Nup358 (also known as Ranbp2) has a previously unanticipated role in Y-complex oligomerization. Our findings blur the established boundaries between scaffold and transport-channel Nups. We conclude that, similar to coated vesicles, several copies of the same structural building block--although compositionally identical--engage in different local sets of interactions and conformations.

Internet Research in Psychology

Abstract: Today the Internet plays a role in the lives of nearly 40% of the world's population, and it is becoming increasingly entwined in daily life. This growing presence is transforming psychological science in terms of the topics studied and the methods used. We provide an overview of the literature, considering three broad domains of research: translational (implementing traditional methods online; e.g., surveys), phenomenological (topics spawned or mediated by the Internet; e.g., cyberbullying), and novel (new ways to study existing topics; e.g., rumors). We discuss issues (e.g., sampling, ethics) that arise when doing research online and point to emerging opportunities (e.g., smartphone sensing). Psychological research on the Internet comes with new challenges, but the opportunities far outweigh the costs. By integrating the Internet, psychological research has the ability to reach large, diverse samples and collect data on actual behaviors, which will ultimately increase the impact of psychological researc...

Identification of molybdenum oxide nanostructures on zeolites for natural gas conversion

Abstract: Direct methane conversion into aromatic hydrocarbons over catalysts with molybdenum (Mo) nanostructures supported on shape-selective zeolites is a promising technology for natural gas liquefaction. We determined the identity and anchoring sites of the initial Mo structures in such catalysts as isolated oxide species with a single Mo atom on aluminum sites in the zeolite framework and on silicon sites on the zeolite external surface. During the reaction, the initial isolated Mo oxide species agglomerate and convert into carbided Mo nanoparticles. This process is reversible, and the initial isolated Mo oxide species can be restored by a treatment with gas-phase oxygen. Furthermore, the distribution of the Mo nanostructures can be controlled and catalytic performance can be fully restored, even enhanced, by adjusting the oxygen treatment.

Similarity learning on an explicit polynomial kernel feature map for person re-identification

Abstract: In this paper, we address the person re-identification problem, discovering the correct matches for a probe person image from a set of gallery person images. We follow the learning-to-rank methodology and learn a similarity function to maximize the difference between the similarity scores of matched and unmatched images for a same person. We introduce at least three contributions to person re-identification. First, we present an explicit polynomial kernel feature map, which is capable of characterizing the similarity information of all pairs of patches between two images, called soft-patch-matching, instead of greedily keeping only the best matched patch, and thus more robust. Second, we introduce a mixture of linear similarity functions that is able to discover different soft-patch-matching patterns. Last, we introduce a negative semi-definite regularization over a subset of the weights in the similarity function, which is motivated by the connection between explicit polynomial kernel feature map and the Mahalanobis distance, as well as the sparsity constraint over the parameters to avoid over-fitting. Experimental results over three public benchmarks demonstrate the superiority of our approach.

Viscoelasticity of biofilms and their recalcitrance to mechanical and chemical challenges

Abstract: We summarize different studies describing mechanisms through which bacteria in a biofilm mode of growth resist mechanical and chemical challenges. Acknowledging previous microscopic work describing voids and channels in biofilms that govern a biofilms response to such challenges, we advocate a more quantitative approach that builds on the relation between structure and composition of materials with their viscoelastic properties. Biofilms possess features of both viscoelastic solids and liquids, like skin or blood, and stress relaxation of biofilms has been found to be a corollary of their structure and composition, including the EPS matrix and bacterial interactions. Review of the literature on viscoelastic properties of biofilms in ancient and modern environments as well as of infectious biofilms reveals that the viscoelastic properties of a biofilm relate with antimicrobial penetration in a biofilm. In addition, also the removal of biofilm from surfaces appears governed by the viscoelasticity of a biofilm. Herewith, it is established that the viscoelasticity of biofilms, as a corollary of structure and composition, performs a role in their protection against mechanical and chemical challenges. Pathways are discussed to make biofilms more susceptible to antimicrobials by intervening with their viscoelasticity, as a quantifiable expression of their structure and composition.

Pattern-Coupled Sparse Bayesian Learning for Recovery of Block-Sparse Signals

Abstract: We consider the problem of recovering block-sparse signals whose cluster patterns are unknown a priori. Block-sparse signals with nonzero coefficients occurring in clusters arise naturally in many practical scenarios. However, the knowledge of the block partition is usually unavailable in practice. In this paper, we develop a new sparse Bayesian learning method for recovery of block-sparse signals with unknown cluster patterns. A pattern-coupled hierarchical Gaussian prior is introduced to characterize the pattern dependencies among neighboring coefficients, where a set of hyperparameters are employed to control the sparsity of signal coefficients. The proposed hierarchical model is similar to that for the conventional sparse Bayesian learning. However, unlike the conventional sparse Bayesian learning framework in which each individual hyperparameter is associated independently with each coefficient, in this paper, the prior for each coefficient not only involves its own hyperparameter, but also its immediate neighbor hyperparameters. In doing this way, the sparsity patterns of neighboring coefficients are related to each other and the hierarchical model has the potential to encourage structured-sparse solutions. The hyperparameters are learned by maximizing their posterior probability. We exploit an expectation-maximization (EM) formulation to develop an iterative algorithm that treats the signal as hidden variables and iteratively maximizes a lower bound on the posterior probability. In the M-step, a simple suboptimal solution is employed to replace a gradient-based search to maximize the lower bound. Numerical results are provided to illustrate the effectiveness of the proposed algorithm.

The Cool Scent of Power: Effects of Ambient Scent on Consumer Preferences and Choice Behavior

Abstract: The present research examines how ambient scents affect consumers' spatial perceptions in retail environments, which in turn influence customers' feelings of power and, thus, product preference and purchasing behavior. Specifically, the authors demonstrate that in a warm- (vs. cool-) scented and thus perceptually more (vs. less) socially dense environment, people experience a greater (vs. lesser) need for power, which manifests in increased preference for and purchase of premium products and brands. This research extends knowledge on store atmospherics and customer experience management through the effects of ambient scent on spatial perceptions and builds on recent research on power in choice contexts.

The Emergence of Online Community Leadership

Abstract: Compared to traditional organizations, online community leadership processes and how leaders emerge are not well studied. Previous studies of online leadership have often identified leaders as those who administer forums or have high network centrality scores. Although communication in online communities occurs almost exclusively through written words, little research has addressed how the comparative use of language shapes community dynamics. Using participant surveys to identify leading online community members, this study analyzes a year of communication network history and message content to assess whether language use differentiates leaders from other core community participants. We contribute a novel use of textual analysis to develop a model of language use to evaluate the utterances of all participants in the community. We find that beyond communication network position-in terms of formal role, centrality, membership in the core, and boundary spanning-those viewed as leaders by other participants, post a large number of positive, concise posts with simple language familiar to other participants. This research provides a model to study online language use and points to the emergent and shared nature of online community leadership.

Snooping Keystrokes with mm-level Audio Ranging on a Single Phone

Abstract: This paper explores the limits of audio ranging on mobile devices in the context of a keystroke snooping scenario. Acoustic keystroke snooping is challenging because it requires distinguishing and labeling sounds generated by tens of keys in very close proximity. Existing work on acoustic keystroke recognition relies on training with labeled data, linguistic context, or multiple phones placed around a keyboard --- requirements that limit usefulness in an adversarial context. In this work, we show that mobile audio hardware advances can be exploited to discriminate mm-level position differences and that this makes it feasible to locate the origin of keystrokes from only a single phone behind the keyboard. The technique clusters keystrokes using time-difference of arrival measurements as well as acoustic features to identify multiple strokes of the same key. It then computes the origin of these sounds precise enough to identify and label each key. By locating keystrokes this technique avoids the need for labeled training data or linguistic context. Experiments with three types of keyboards and off-the-shelf smartphones demonstrate scenarios where our system can recover $94\%$ of keystrokes, which to our knowledge, is the first single-device technique that enables acoustic snooping of passwords.

Securing physical-layer communications for cognitive radio networks

Abstract: This article investigates the physical-layer security of CR networks, which are vulnerable to various newly arising attacks targeting the weaknesses of CR communications and networking. We first review a range of physical-layer attacks in CR networks, including primary user emulation, sensing falsification, intelligence compromise, jamming, and eavesdropping attacks. Then we focus on the physical-layer security of CR networks against eavesdropping and examine the secrecy performance of cognitive communications in terms of secrecy outage probability. We further consider the use of relays for improving CR security against eavesdropping and propose an opportunistic relaying scheme, where a relay node that makes CR communications most resistant to eavesdropping is chosen to participate in assisting the transmission from a cognitive source to its destination. It is illustrated that the physical- layer secrecy of CR communications relying on opportunistic relaying can be significantly improved by increasing the number of relays, showing the security benefit of exploiting relay nodes. Finally, we present some open challenges in the field of relay-assisted physical-layer security for CR networks.

Mediated semiquantum key distribution

Abstract: In this work, we design a quantum key distribution protocol, allowing two limited semiquantum or ``classical'' users to establish a shared secret key with the help of a fully quantum server. A semiquantum user can prepare and measure qubits only in the computational basis and so must rely on this quantum server to produce qubits in alternative bases and also to perform alternative measurements. However, we assume that the server is untrusted and we prove the unconditional security of our protocol even in the worst case: when this quantum server is an all-powerful adversary. We also compute a lower bound of the key rate of our protocol, in the asymptotic scenario, as a function of the observed error rate in the channel, allowing us to compute the maximally tolerated error of our protocol. Our results show that a semiquantum protocol may hold similar security to a fully quantum one.

An augmented Lagrangian method for distributed optimization

Abstract: We propose a novel distributed method for convex optimization problems with a certain separability structure. The method is based on the augmented Lagrangian framework. We analyze its convergence and provide an application to two network models, as well as to a two-stage stochastic optimization problem. The proposed method compares favorably to two augmented Lagrangian decomposition methods known in the literature, as well as to decomposition methods based on the ordinary Lagrangian function.

On the limits of predictability in real-world radio spectrum state dynamics: from entropy theory to 5G spectrum sharing

Abstract: A range of applications in cognitive radio networks, from adaptive spectrum sensing to predictive spectrum mobility and dynamic spectrum access, depend on our ability to foresee the state evolution of radio spectrum, raising a fundamental question: To what degree is radio spectrum state (RSS) predictable? In this article we explore the fundamental limits of predictability in RSS dynamics by studying the RSS evolution patterns in spectrum bands of several popular services, including TV bands, ISM bands, cellular bands, and so on. From an information theory perspective, we introduce a methodology of using statistical entropy measures and Fano inequality to quantify the degree of predictability underlying real-world spectrum measurements. Despite the apparent randomness, we find a remarkable predictability, as large as 90 percent, in real-world RSS dynamics over a number of spectrum bands for all popular services. Furthermore, we discuss the potential applications of prediction-based spectrum sharing in 5G wireless communications.

Business Process Analytics

Abstract: Business Process Management systems (BPMS) are a rich source of events that document the execution of processes and activities within these systems. Business Process Management analytics is the family of methods and tools that can be applied to these event streams in order to support decision making in organizations. The analysis of process events can focus on the behavior of completed processes, evaluate currently running process instances, or focus on predicting the behavior of process instances in the future. This chapter provides an overview of the different methods and technologies that can be employed in each of these three areas of process analytics. We discuss the underlying format and types of process events as the common source of analytics information, present techniques for the aggregation and composition of these events, and outline methods that support backward- and forward-looking process analytics.

Impact of 3D Hierarchical Nanostructures on the Antibacterial Efficacy of a Bacteria-Triggered Self-Defensive Antibiotic Coating.

Abstract: Titanium is often applied in implant surgery, but frequently implicated in infections associated with bacterial adhesion and growth on the implant surface. Here, we show that hierarchical nanostructuring of titanium and the subsequent coating of resulting topographical features with a self-defensive, antibacterial layer-by-layer (LbL) film enables a synergistic action of hierarchical nanotopography and localized, bacteria-triggered antibiotic release to dramatically enhance the antibacterial efficiency of surfaces. Although sole nanostructuring of titanium substrates did not significantly affect adhesion and growth of Staphylococcus aureus, the coating of 3D-nanopillared substrates with an ultrathin tannic acid/gentamicin (TA/G) LbL film resulted in a 10-fold reduction of the number of surface-attached bacteria. This effect is attributed to the enlarged surface area of the nanostructured coating available for localized bacteria-triggered release of antibiotics, as well as to the lower bacterial adhesion forces resulting in subsided activation of bacterial antibiotic-defense mechanisms when bacteria land on nanopillar tips. The result shows that a combination of 3D nanostructuring with a bacteria-triggered antibiotic-releasing coating presents a unique way to dramatically enhance antibacterial efficacy of biomaterial implants.

User Verification Leveraging Gait Recognition for Smartphone Enabled Mobile Healthcare Systems

Abstract: The rapid deployment of sensing technology in smartphones and the explosion of their usage in people’s daily lives provide users with the ability to collectively sense the world. This leads to a growing trend of mobile healthcare systems utilizing sensing data collected from smartphones with/without additional external sensors to analyze and understand people’s physical and mental states. However, such healthcare systems are vulnerable to user spoofing, in which an adversary distributes his registered device to other users such that data collected from these users can be claimed as his own to obtain more healthcare benefits and undermine the successful operation of mobile healthcare systems. Existing mitigation approaches either only rely on a secret PIN number (which can not deal with colluded attacks) or require an explicit user action for verification. In this paper, we propose a user verification system leveraging unique gait patterns derived from acceleration readings to detect possible user spoofing in mobile healthcare systems. Our framework exploits the readily available accelerometers embedded within smartphones for user verification. Specifically, our user spoofing mitigation framework (which consists of three components, namely Step Cycle Identification, Step Cycle Interpolation, and Similarity Comparison) is used to extract gait patterns from run-time accelerometer measurements to perform robust user verification under various walking speeds. We show that our framework can be implemented in two ways: user-centric and server-centric, and it is robust to not only random but also mimic attacks. Our extensive experiments using over 3,000 smartphone-based traces with mobile phones placed on different body positions confirm the effectiveness of the proposed framework with users walking at various speeds. This strongly indicates the feasibility of using smartphone based low grade accelerometer to conduct gait recognition and facilitate effective user verification without active user cooperation.

Restraining Overconfident CEOs through Improved Governance: Evidence from the Sarbanes-Oxley Act

Abstract: The literature posits that some CEO overconfidence benefits shareholders, though high levels may not. We argue adequate controls and independent viewpoints provided by an independent board mitigates the costs of CEO overconfidence. We use the concurrent passage of the Sarbanes-Oxley Act and changes to the NYSE/NASDAQ listing rules (collectively, SOX) as natural experiments to examine whether board independence improves decision-making by overconfident-CEOs. The results are strongly supportive: Post-SOX, overconfident CEOs reduce investment and risk exposure, increase dividends, improve post-acquisition performance, and have better operating performance and market value. Importantly, these changes are absent for overconfident-CEO firms that were compliant prior to passage.

Esterification and transesterification of waste cooking oil over Amberlyst 15 and modified Amberlyst 15 catalysts

Abstract: Biodiesel production via the simultaneous esterification and transesterification reactions of waste cooking oil (WCO) with methanol was investigated. A batch reflux reactor was used in conjunction with acidic ion-exchange catalysts, i.e., Amberlyst 15 and modified Amberlyst 15. The activity of the Amberlyst 15 catalysts could be altered via heat treatment at temperatures that are ≥493 K, as a consequence of the reduction of the active surface ( SO3H). The hydrogen ion-exchange capacities of the catalysts ranged from 2.5 meq H+/g to 5.1 meq H+/g. The reaction parameters included the molar ratio of alcohol to oil (6:1–15:1), reaction temperature (298–338 K), catalyst concentration (1–9 wt.%), reaction time (1–72 h) and the free fatty acid content (1.04–8.04%). It was observed that the activities of the Amberlyst 15 and modified Amberlyst 15 catalysts increased linearly with increasing hydrogen ion-exchange capacities of the catalysts. Amberlyst 15 exhibited the highest hydrogen ion-exchange capacity and consequently generated the highest biodiesel yield of (78 ± 3.39%). The esterification reaction of excess fatty acids was also investigated. When the esterification and transesterification reactions were carried out simultaneously, the water generated from the esterification of free fatty acids led to the hydrolysis of triglycerides and reduced the biodiesel yield emphasizing the necessity for the removal of the water during concomitant esterification of excess fatty free acids present and the transesterification of the triglycerides for biodiesel production.

Fine-grained sleep monitoring: Hearing your breathing with smartphones

Abstract: Sleep monitoring has drawn increasingly attention as the quality and quantity of the sleep are important to maintain a person's health and well-being. For example, inadequate and irregular sleep are usually associated with serious health problems such as fatigue, depression and cardiovascular disease. Traditional sleep monitoring systems, such as PSG, involve wearable sensors with professional installations, and thus are limited to clinical usage. Recent work in using smartphone sensors for sleep monitoring can detect several events related to sleep, such as body movement, cough and snore. Such coarse-grained sleep monitoring however is unable to detect the breathing rate which is an important vital sign and health indicator. This work presents a fine-grained sleep monitoring system which is capable of detecting the breathing rate by leveraging smartphones. Our system exploits the readily available smartphone earphone placed close to the user to reliably capture the human breathing sound. Given the captured acoustic sound, our system performs noise reduction to remove environmental noise and then identifies the breathing rate based on the signal envelope detection. Our system can further detect detailed sleep events including snore, cough, turn over and get up based on the acoustic features extracted from the acoustic sound. Our experimental evaluation of six subjects over six months time period demonstrates that the breathing rate monitoring and sleep events detection are highly accurate and robust under various environments. By combining breathing rate and sleep events, our system can provide continuous and noninvasive fine-grained sleep monitoring for healthcare related applications, such as sleep apnea monitoring as evidenced by our experimental study.

A global survey of cloud overlap based on CALIPSO and CloudSat measurements

Abstract: . Using 2B-CLDCLASS-LIDAR (radar–lidar) cloud classification and 2B-FLXHR-LIDAR radiation products from CloudSat over 4 years, this study evaluates the co-occurrence frequencies of different cloud types, analyzes their along-track horizontal scales and cloud radiative effects (CREs), and utilizes the vertical distributions of cloud types to evaluate cloud-overlap assumptions. The statistical results show that high clouds, altostratus (As), altocumulus (Ac) and cumulus (Cu) tend to coexist with other cloud types. However, stratus (St) (or stratocumulus, Sc), nimbostratus (Ns) and convective clouds are much more likely to exhibit individual features than other cloud types. On average, altostratus-over-stratus/stratocumulus cloud systems have a maximum horizontal scale of 17.4 km, with a standard deviation of 23.5 km. Altocumulus-over-cumulus cloud types have a minimum scale of 2.8 km, with a standard deviation of 3.1 km. By considering the weight of each multilayered cloud type, we find that the global mean instantaneous net CREs of multilayered cloud systems during the daytime are approximately −41.3 and −50.2 W m−2, which account for 40.1 and 42.3% of the global mean total net CREs at the top of the atmosphere (TOA) and at the surface, respectively. The radiative contributions of high-over-altocumulus and high-over-stratus/stratocumulus (or cumulus) in the all multilayered cloud systems are dominant due to their frequency. Considering the overlap of cloud types, the cloud fraction based on the random overlap assumption is underestimated over vast oceans, except in the west-central Pacific Ocean warm pool. Obvious overestimations mainly occur over tropical and subtropical land masses. In view of a lower degree of overlap than that predicted by the random overlap assumption to occur over the vast ocean, particularly poleward of 40° S, the study therefore suggests that a linear combination of minimum and random overlap assumptions may further improve the predictions of actual cloud fractions for multilayered cloud types (e.g., As + St/Sc and Ac + St/Sc) over the Southern Ocean. The establishment of a statistical relationship between multilayered cloud types and the environmental conditions (e.g., atmospheric vertical motion, convective stability and wind shear) would be useful for parameterization design of cloud overlap in numerical models.

D 3 : Abnormal driving behaviors detection and identification using smartphone sensors

Abstract: Real-time abnormal driving behaviors monitoring is a corner stone to improving driving safety. Existing works on driving behaviors monitoring using smartphones only provide a coarsegrained result, i.e. distinguishing abnormal driving behaviors from normal ones. To improve drivers' awareness of their driving habits so as to prevent potential car accidents, we need to consider a finegrained monitoring approach, which not only detects abnormal driving behaviors but also identifies specific types of abnormal driving behaviors, i.e. Weaving, Swerving, Sideslipping, Fast U-turn, Turning with a wide radius and Sudden braking. Through empirical studies of the 6-month driving traces collected from real driving environments, we find that all of the six types of driving behaviors have their unique patterns on acceleration and orientation. Recognizing this observation, we further propose a finegrained abnormal Driving behavior Detection and iDentification system, D3, to perform real-time high-accurate abnormal driving behaviors monitoring using smartphone sensors. By extracting unique features from readings of smartphones' accelerometer and orientation sensor, we first identify sixteen representative features to capture the patterns of driving behaviors. Then, a machine learning method, Support Vector Machine (SVM), is employed to train the features and output a classifier model which conducts fine-grained identification. From results of extensive experiments with 20 volunteers driving for another 4 months in real driving environments, we show that D3 achieves an average total accuracy of 95.36%.

Developer Recommendation for Crowdsourced Software Development Tasks

Abstract: Crowdsourced software development utilises an open call format to attract geographically distributed developers to accomplish various types of software development tasks. Although the open call format enables wide task accessibility, potential developers must choose from a dauntingly large set of task options (usually more than one hundred available tasks on TopCoder each day). Inappropriate developer-task matching may lower the quality of the software deliverables. In this paper, we employ content-based recommendation techniques to automatically match tasks and developers. The approach learns particular interests from registration history and mines winner history to favour appropriate developers. We measure the performance of our approach by defining accuracy and diversity metrics. We evaluate our recommendation approach by introducing 4 machine learners on 3,094 historical tasks from TopCoder. The evaluation results show that promising accuracy and diversity are achievable (accuracy from 50% to 71% and diversity from 40% to 52% when recommending reliable developers).We also provide advice extracted from our results to guide the crowdsourcing platform in building a recommender system in practice.

Combating Rumor Spread on Social Media: The Effectiveness of Refutation and Warning

Abstract: Twitter and other social media are now a major method of information exchange and dissemination. Although they can support rapid communication and sharing of useful information, they can also facilitate the spread of rumors, which contain unverified information. The purpose of the work reported here was to examine several design ideas for reducing the spread of health-related rumors in a Twitter-like environment. The results have shown that exposing people to information that refutes rumors or warns that the statement has appeared on rumor websites could reduce the spread of rumors. These results suggest that social media technologies can be designed such that users can self correct and inactivate potentially inaccurate information in their environment.

SE 2014: Curriculum Guidelines for Undergraduate Degree Programs in Software Engineering

Abstract: Revised curriculum guidelines help university faculty create or update undergraduate software engineering programs.