The Experience Sampling Method (ESM) is widely used to collect emotion self-reports to train machine learning models for emotion inference. However, as ESM studies are time-consuming and burdensome, participants often withdraw in between. This unplanned withdrawal compels the researchers to discard the dropout participants’ data, significantly impacting the quality and quantity of the self-reports. To address this problem, we leverage only the self-reporting similarity across participants (unlike prior works that apply different machine learning approaches on additional modalities) for missing self-report estimation. In specific, we propose a Multi-task Learning (MTL) framework, MUSE, that constructs the missing self-reports of the dropout participants. We evaluate MUSE in two in-the-wild studies (N1=24, N2=30) of 6-week and 8-week duration, during which the participants reported four emotions (happy, sad, stressed, relaxed) using a smartphone application. The evaluation reveals that MUSE estimates the missing emotion self-reports with an average AUCROC of 84% (Study I) and 82% (Study II). A follow-up evaluation of MUSE for an emotion inference (downstream) task reveals no significant difference in emotion inference performance when estimated self-reports are used. These findings underscore the utility of MUSE in estimating missing self-reports in ESM studies and the applicability of MUSE for downstream tasks (e.g., emotion inference).

The shift from monolithic services to microservices brings modularity and elasticity, but detecting faults and anomalies is challenging due to diverse data and evolving technology. The heterogeneous nature of this data complicates the analysis of anomaly signatures across various dimensions. Given the continuous evolution of this technology, exhaustively learning from historical data poses difficulties. To address these challenges, we present URCD, a solution designed to identify and localize faults or anomalies at the application and service level. Remarkably, URCD achieves this without explicit training on faulty data. Our approach integrates heterogeneous microservice data into a bidirectional weighted graph, leveraging a sophisticated Hyper Graph Attention Network (HGAN) model to process heterogeneous data logs generated by microservices. Our evaluation shows the optimal performance of URCD while detecting root cause of anomalies.

Road traffic congestion affects not only the commute delay but also a city's overall social, economic, and environmental growth. Existing approaches for road congestion mitigation primarily adopt a reactive approach by detecting congestion after it occurs and recommending alternate routes to the vehicles, which fails to prevent congestion cascading. In contrast, we propose a pervasive platform called ProCon that proactively infers the driving micro-behaviors that can contribute to congestion formation and assist the drivers in avoiding such maneuvers in real-time during the navigation. Thorough evaluations over multiple real-life and simulated datasets indicate that ProCon can reduce congestion for more than 60% of the scenarios on average while significantly reducing the travel time of the vehicles.

Road safety remains a critical global concern, with millions of crashes reported annually. Understanding the safety of individual road junctions is vital, especially in areas prone to road rage and reckless driving. However, current navigation systems lack detailed safety information, increasing risk for drivers and pedestrians. Recognizing this need, this paper introduces øurmethod that automatically annotates the road segments with a driving safety level to aid cautious maneuvering and safe driving practices. By leveraging onboard sensors, øurmethod identifies causal chains behind poor driving maneuvers, enabling the modeling of safety levels for various road segments. We perform a thorough evaluation of øurmethod over publicly available and collected datasets from multiple countries and observe >80% accuracy (in terms of F1-score) in correctly annotating the safety concerns. In addition, a thorough user study indicates the generalizability and usability of the proposed approach for its practical deployment considerations.

Drive tourism has become increasingly popular in the past decade; however, driving through a new city is challenging because the road and traffic environments vary significantly across cities. A driver used to driving in one city may face severe difficulty in adapting to a different driving environment, leading to road fatalities. This article develops GRIDS , an explainable model for guidelines recommendation for inter-domain driving safety, which learns the driving rules behind the changing environment and recommends the necessary personalized guidelines to a driver while driving through a new city. We develop an explainable domain adaptation model to provide customized recommendations in terms of driving guidelines, broadly categorized into four major feature categories of a driving environment. A thorough evaluation over the CARLA driving simulator shows that the recommendations generated through GRIDS can help improve driving safety.

Driving is a complex task carried out under the influence of diverse spatial objects and their temporal inter-actions. Therefore, a sudden fluctuation in driving behavior can be due to either a lack of driving skill or the effect of various on-road spatial factors such as pedestrian movements, peer vehicles' actions, etc. Therefore, understanding the context behind a degraded driving behavior just-in-time is necessary to ensure on-road safety. In this paper, we develop a system called DriCon that exploits the information acquired from a dashboard-mounted edge-device to understand the context in terms of micro-events from a diverse set of on-road spatial factors and in-vehicle driving maneuvers taken. DriCon uses the live in-house testbed and the largest publicly available driving dataset to generate human interpretable explanations against the unexpected driving events. Also, it provides a better insight with an improved similarity of 80% over 50 hours of driving data than the existing driving behavior characterization techniques.

Smartphone keyboard interaction based emotion detection systems are used widely to provide value-added services such as mental health monitoring, keyboard layout optimization, guided response generation. At the core of these services lie a machine learning model, which automatically infers emotion based on keyboard interaction pattern. To train these models, the emotion ground truth labels are typically collected as emotion self-report by conducting an Experience Sampling Method (ESM) based study. However, as responding to repetitive self-report probes is time-consuming and fatigue-inducing, efficient self-report collection approaches are essential that avoid probing at inopportune moments and reduce survey fatigue. To address this problem, we propose an active learning based framework, ALOE (Active Learning based Opportunistic Experience Sampling for Emotion Self-report Collection) that automatically decides to avoid probing at the unfavorable moments based on the typing signatures captured from smartphone keyboard interaction sessions. We bootstrap the framework with a few labeled instances (typing session) and allow the learner to probe (or query) the user only when it is least confident about an instance (typing session) and retrain accordingly. This way, we reduce the number of probes required (and therefore user engagement) and yet probe at the opportune moments. We evaluate ALOE in a 3-week in-the-wild study involving 18 participants, who record their smartphone keyboard interaction patterns and emotion self-reports during this period. The experimental results demonstrate that ALOE requires 56% less inopportune self-reports to train the probing moment detection learning model and yet detects the probing moments accurately with an average F-score of 93%.

Multiuser activity recognition has been the core of different context-aware services. However, the development of such services is often plagued by the dearth of multiuser datasets. This paper presents a strategy for generating synthetic multiuser datasets by augmenting existing real-life datasets. The described strategy exploits pre-cise time synchronization and well-known audio augmentation approaches to generate a multimodal activity recognition dataset with locomotive and acoustic signatures.

Monitoring driving behavior is essential to ensure on-road safety. Although driving is a collective, cooperative task among the drivers of the neighboring vehicles, existing platforms for driving behavior analysis solely rely on different on-road maneuvers taken by a driver. By analyzing a large volume of publicly available data over two countries and in-house collected data, this paper argues that analyzing driving behavior needs treatment over different factors which compel a driver to take maneuvers that are otherwise recommended to be avoided. Consequently, we develop DriBe. This smartphone-based pervasive sensing system utilizes video, GPS, and inertial sensor data to investigate the causes and consequences of driving maneuvers to score a driver based on a thorough understanding of their on-road driving behavior. Considering that the causality factors are very much specific to a particular driving environment (like a country), DriBe also incorporates a domain-adaptive architecture by utilizing a transfer learning framework. Thorough evaluation of DriBe with datasets from three countries shows that a score based on such causal factors provides a more accurate representation of driving behavior compared to baselines.

Severe road accidents are reported regularly across the globe due to drivers getting distracted while using their smartphones. To prevent such fatalities, one possible approach is to make the smartphone intelligent enough to detect whether it is being used by the driver, thus providing restricted access to the applications while driving. However, this problem is challenging as the driver can behave like an adversary to fool the system; therefore, additional devices or forward communication cannot be used. This paper proposes a novel approach of smartphone localization within a car by exploiting the ambient mechanical noise within the vehicle. We utilize the periodic nature of such mechanical noises to develop a simple yet satisfactorily accurate approach, called Blah, that can utilize the acoustic properties from the ambient mechanical noise within the car to detect whether the driver or the passenger is using the smartphone while the car is on the road.

An individual’s cognitive and functional abilities are commonly assessed through physical and mental status examination, observational performance measures, surveys and proxy reports of symptoms. These strategies are not ideal for early impairment detection as the individual needs to be present physically at the clinic to avail the assessments, especially for older adults who require assistance from a caregiver, and experience mobility, cognitive and functional disabilities from neurodegenerative disorders. Moreover, these strategies rely on self-reporting and proxy reports for evaluation which often leads to under-reporting of symptoms and decrease the validity of these measures. We argue that an early assessment of functional, and cognitive health impairment can be obtained from the individual’s daily activities captured through accelerometry. In this work, we postulate to learn high-level motion related representations from accelerometer data to better correlate with underlying functional and cognitive health parameters of older adults using a contrastive and multi-task learning framework. In particular, we posit a novel indicator, Impairment Indicator using the proposed multi-task learning framework that can indicate functional or cognitive decline as neurodegenerative disease progresses. An extensive 24-hour data collection from 25 older adults with the clinician in-the-loop was carried out in a retirement community center with IRB approval. We collected the activity patterns using wearables in their homes in addition to survey-based assessments and observational performance measures recorded by a clinical evaluator to infer their current cognitive and functional impairment status. Our evaluation on the acquired dataset reveals that the representations learned using contrastive learning aids in improving the detection of activities, activity performance score, and stage of dementia to 92%, 97%, and 98%, respectively.

In recent years, non-invasive human activity recognition (HAR) has gathered huge momentum using locomotive sensors. However, for effective HAR, there is a need for a significant volume of annotated data. Typically, the conventional practices for gathering HAR annotations have relied on human annotators. Nevertheless, the growing volume of data often leads to the collection of shallow annotations, which in most cases ignore the fine-grained micro-activities that constitute any complex activities of daily living (ADL). Understanding this, we, in this paper, try to develop the framework AmicroN that can automatically generate micro-activity annotations using locomotive signatures. To achieve this, in the backend, AmicroN applies change-point detection for the precise detection of activity boundaries followed by zero-shot learning with verb attributes to identify the unseen micro-activities without any external supervision. Rigorous evaluation on a publicly available Kitchen dataset shows that AmicroN can identify the micro-activities with a median F1-score of$\geq \mathbf{0.75}$for all the subjects, which can help develop novel pervasive applications.

This demo presents AmicroN which automatically generates fine-grained annotations using locomotive signatures. In the backend, AmicroN exploits short-duration macro labels already present in a pre-annotated dataset. It uses zero-shot learning to identify the finer micro-activities present within a coarse-grain macro-activity label.

Monitoring students’ engagement and understanding their learning pace in a virtual classroom becomes challenging in the absence of direct eye contact between the students and the instructor. Continuous monitoring of eye gaze and gaze gestures may produce inaccurate outcomes when the students are allowed to do productive multitasking, such as taking notes or browsing relevant content. This paper proposes Stungage – a software wrapper over existing online meeting platforms to monitor students’ engagement in real-time by utilizing the facial video feeds from the students and the instructor coupled with a local on-device analysis of the presentation content. The crux of Stungage is to identify a few opportunistic moments when the students should visually focus on the presentation content if they can follow the lecture. We investigate these instances and analyze the students’ visual, contextual, and cognitive presence to assess their engagement during the virtual classroom while not directly sharing the video captures of the participants and their screens over the web. Our system achieves an overall F2-score of 0.88 for detecting student engagement. Besides, we obtain 92 responses from the usability study with an average SU score of 74.18.

Commuter comfort in cab rides affects driver rating as well as the reputation of ride-hailing firms like Uber/Lyft. Existing research has revealed that commuter comfort not only varies at a personalized level but also is perceived differently on different trips for the same commuter. Furthermore, there are several factors, including driving behavior and driving environment, affecting the perception of comfort. Automatically extracting the perceived comfort level of a commuter due to the impact of the driving behavior is crucial for a timely feedback to the drivers, which can help them to meet the commuter’s satisfaction. In light of this, we surveyed around 200 commuters who usually take such cab rides and obtained a set of features that impact comfort during cab rides. Following this, we develop a system Ridergo which collects smartphone sensor data from a commuter, extracts the spatial time series feature from the data, and then computes the level of commuter comfort on a five-point scale with respect to the driving. Ridergo uses a Hierarchical Temporal Memory model-based approach to observe anomalies in the feature distribution and then trains a multi-task learning-based neural network model to obtain the comfort level of the commuter at a personalized level. The model also intelligently queries the commuter to add new data points to the available dataset and, in turn, improve itself over periodic training. Evaluation of Ridergo on 30 participants shows that the system could provide efficient comfort score with high accuracy when the driving impacts the perceived comfort.

Keyboard interaction patterns on a smartphone is the input for many intelligent emotion-aware applications, such as adaptive interface, optimized keyboard layout, automatic emoji recommendation in IM applications. The simplest approach, called the Experience Sampling Method (ESM), is to systematically gather self-reported emotion labels from users, which act as the ground truth labels, and build a supervised prediction model for emotion inference. However, as manual self-reporting is fatigue-inducing and attention-demanding, the self-report requests are to be scheduled at favorable moments to ensure high fidelity response. We, in this paper, perform fine-grain keyboard interaction analysis to determine suitable probing moments. Keyboard interaction patterns, both cadence, and latency between strokes, nicely translate to frequency and time domain analysis of the patterns. In this paper, we perform a 3-week in-the-wild study (N = 22) to log keyboard interaction patterns and self-report details indicating (in)opportune probing moments. Analysis of the dataset reveals that time-domain features (e.g., session length, session duration) and frequency-domain features (e.g., number of peak amplitudes, value of peak amplitude) vary significantly between opportune and inopportune probing moments. Driven by these analyses, we develop a generalized (all-user) Random Forest based model, which can identify the opportune probing moments with an average F-score of 93%. We also carry out the explainability analysis of the model using SHAP (SHapley Additive exPlanations), which reveals that the session length and peak amplitude have strongest influence to determine the probing moments.

Smartphones provide the capability to perform in-situ sampling of human behavior using Experience Sampling Method (ESM). Designing an ESM schedule involves probing the user repeatedly at suitable moments to collect self-reports. Timely probe generation to collect high fidelity user responses while keeping probing rate low is challenging. In mobile-based ESM, timeliness of the probe is also impacted by user's availability to respond to self-report request. Thus, a good ESM design must consider -probing frequency,timely self-report collection, andnotifying at opportune momentto ensure highresponse quality. We propose a two-phase ESM design, where the first phase (a) balances between probing frequency and self-report timeliness, and (b) in parallel, constructs a predictive model to identify opportune probing moments. The second phase uses this model to further improve response quality by eliminating inopportune probes. We use typing-based emotion detection in smartphone as a case study to validate proposed ESM design. Our results demonstrate that it reduces probing rate by 64 percent, samples self-reports timely by reducing elapsed time between self-report collection, and event trigger by 9 percent while detecting inopportune moments with an average accuracy of 89 percent. These design choices improve the response quality, as manifested by 96 percent valid response collection and a maximum improvement of 24 percent in emotion classification accuracy.

Annotated IMU sensor data from smart devices and wearables are essential for developing supervised models for fine-grained human activity recognition, albeit generating sufficient annotated data for diverse human activities under different environments is challenging. Existing approaches primarily use human-in-the-loop based techniques, including active learning; however, they are tedious, costly, and time-consuming. Leveraging the availability of acoustic data from embedded microphones over the data collection devices, in this paper, we propose LASO, a multimodal approach for automated data annotation from acoustic and locomotive information. LASO works over the edge device itself, ensuring that only the annotated IMU data is collected, discarding the acoustic data from the device itself, hence preserving the audio-privacy of the user. In the absence of any pre-existing labeling information, such an auto-annotation is challenging as the IMU data needs to be sessionized for different time-scaled activities in a completely unsupervised manner. We use a change-point detection technique while synchronizing the locomotive information from the IMU data with the acoustic data, and then use pre-trained audio-based activity recognition models for labeling the IMU data while handling the acoustic noises. LASO efficiently annotates IMU data, without any explicit human intervention, with a mean accuracy of $0.93$ ($\pm 0.04$) and $0.78$ ($\pm 0.05$) for two different real-life datasets from workshop and kitchen environments, respectively.

Most of the latest context-based applications capture the mobility of a user using Inertial Measurement Unit (IMU) sensors like accelerometer and gyroscope which do not need explicit user-permission for application access. Although these sensors provide highly accurate mobility context information, existing studies have shown that they can lead to undesirable leakage of location information. To evade this breach of location privacy, many of the state-of-the-art studies suggest to impose stringent restrictions over the usage of IMU sensors. However, in this paper, we show that typing and smartphone engagement patterns can act as an alternative modality to sniff the mobility context of a user, even if the IMU sensors are not sampled at all. We develop an adversarial framework, named ConType, which exploits the signatures exposed by typing and smartphone engagement patterns to track the mobility of a user. Rigorous experiments with in-the-wild dataset show that ConType can track the mobility contexts with an average micro-F1of 0.87 (±0.09), without using IMU data. Through additional experiments, we also show that ConType can track mobility stealthily with very low power and resource footprints, thus further aggravating the risk.

Network embedding, that aims to learn low-dimensional vector representation of nodes such that the network structure is preserved, has gained significant research attention in recent years. However, most state-of-the-art network embedding methods are computationally expensive and hence unsuitable for representing nodes in billion-scale networks. In this paper, we present LouvainNE, a hierarchical clustering approach to network embedding. Precisely, we employ Louvain, an extremely fast and accurate community detection method, to build a hierarchy of successively smaller subgraphs. We obtain representations of individual nodes in the original graph at different levels of the hierarchy, then we aggregate these representations to learn the final embedding vectors. Our theoretical analysis shows that our proposed algorithm has quasi-linear run-time and memory complexity. Our extensive experimental evaluation, carried out on multiple real-world networks of different scales, demonstrates both (i) the scalability of our proposed approach that can handle graphs containing tens of billions of edges, as well as (ii) its effectiveness in performing downstream network mining tasks such as network reconstruction and node classification.

Although public transport vehicles such as buses have always been an economical means of commuting in the cities of many developing countries, it is always considered as a secondary mode of transport owing to poor infrastructure, chaotic and reckless driving habits, and absence of any proper information system in buses. Based on rigorous experiments carried out over a period of two years and multiple surveys, we have tried to learn the problems faced by bus commuters. As a solution, in this article, we develop a novel energy-efficient system which would help commuters navigate through their journey safely. Along with making them aware of any upcoming points of concerns (PoCs) such as sudden bumps, sharp turns, and bad roads, we also inform commuters about the expected time of arrival at the destination. The system makes use of several landmarks such as speed breakers, turns, and bus stops on a trail stored in a specialized data structure, the probabilistic timed automata. We conducted extensive experiments using 25 volunteers over 50 trails. The system showed an average localization error of only 50 m and mean estimated time of arrival (ETA) error of 2.5 mins and a fairly high alert prediction accuracy while consuming significantly less energy when compared to GPS.

Event-based online social platforms, such as Meetup and Plancast, have experienced increased popularity and rapid growth in recent years. In EBSN setup, selecting suitable venues for hosting events, which can attract a great turnout, is a key challenge. In this paper, we present a deep learning based venue recommendation system DeepVenue which provides context driven venue recommendations for the Meetup event-hosts to host their events. The crux of the proposed model relies on the notion of similarity between multiple Meetup entities such as events, venues, groups, etc. We develop deep learning techniques to compute a compact descriptor for each entity, such that two entities (say, venues) can be compared numerically. Notably, to mitigate the scarcity of venue related information in Meetup, we leverage on the cross domain knowledge transfer from popular LBSN service Yelp to extract rich venue related content. For hosting an event, the proposed DeepVenue model computes a success score for each candidate venue and ranks those venues according to the scores and finally recommend the top k venues. Our rigorous evaluation on the Meetup data collected for the city of Chicago shows that DeepVenue significantly outperforms the baselines algorithms. Precisely, for 84 percent of events, the correct hosting venue appears in the top 5 of the DeepVenue recommended list.

Characteristics of typing on smartphone keyboards among different individuals can elicit emotion, similar to speech prosody or facial expressions. Existing works on typing based emotion recognition rely on feature engineering to build machine learning models, while recent speech and facial expression based techniques have shown the efficacy of learning the features automatically. Therefore, in this work, we explore the effectiveness of such learning models in keyboard interaction based emotion detection. In this paper, we propose an end-to-end framework, which first uses a sequence-based encoding method to automatically learn the representation from raw keyboard interaction pattern and subsequently uses this representation to train a multi-task learning based neural network (MTL-NN)to identify different emotions. We carry out a 3-week in-the-wild study involving 24 participants using a custom keyboard capable of tracing users' interaction pattern during text entry. We collect interaction details like touch speed, error rate, pressure and self-reported emotions (happy, sad, stressed, relaxed) during the study. Our analysis on the collected dataset reveals that the representation learnt from the interaction pattern has an average correlation of 0.901 within the same emotion and 0.811 between different emotions. As a result, the representation is effective in distinguishing different emotions with an average accuracy (AUCROC)of 84%.

Typing based interfaces are common across many mobile applications, especially messaging apps. To reduce the difficulty of typing using keyboard applications on smartphones, smartwatches with restricted space, several techniques, such as auto-complete, auto-suggest, are implemented. Although helpful, these techniques do add more cognitive load on the user. Hence beyond the importance to improve the word recommendations, it is useful to understand the pattern of use of auto-suggestions during typing. Among several factors that may influence use of auto-suggest, the role of emotion has been mostly overlooked, often due to the difficulty of unobtrusively inferring emotion. With advances in affective computing, and ability to infer user's emotional states accurately, it is imperative to investigate how auto-suggest can be guided by emotion aware decisions. In this work, we investigate correlations between user emotion and usage of auto-suggest i.e. whether users prefer to use auto-suggest in specific emotion states. We developed an Android keyboard application, which records auto-suggest usage and collects emotion self-reports from users in a 3-week in-the-wild study. Analysis of the dataset reveals relationship between user reported emotion state and use of auto-suggest. We used the data to train personalized models for predicting use of auto-suggest in specific emotion state. The model can predict use of auto-suggest with an average accuracy (AUCROC) of 82% showing the feasibility of emotion-aware auto-suggestion.

The growth in the market for cab companies like Uber has opened the door to high-income options for drivers. However, in order to boost their income, drivers many a time resort to accepting trips which increases their stress resulting in poor driving quality and accidents in serious cases. Every driver handles stress differently and the trip recommendation thus needs to be on a personalized level. In this paper, we explore historical trip data to compute the driving stress and its impact on various driving behavioral features, captured through vehicle-mounted GPS and inertial sensors. We utilize a Multi-task Learning based Neural Network model to learn both the common features and the personalized features from the driving data to predict the stress level of a driver. We further establish a causal relationship between the stress level of a driver and his driving behavior. Finally, we develop a trip recommendation system for cab drivers to avoid stress driving. The models have been tested over both a publicly available dataset with 6 drivers for 500 minutes of driving data and an in-house collected dataset from 8 drivers over 1700 trips for 5 months. We observe that the proposed model gives an average prediction accuracy of 94% with low false-positive rates. We also observed that the driving behavior is improved when a driver akes a recommended trip.

Identification of influential users in online social networks allows to facilitate efficient information diffusion to a large part of the network and thus benefiting diverse applications including viral marketing, disease control, and news dissemination. Existing methods have mainly relied on the network structure only for the detection of influential users. In this paper, we enrich this approach by proposing a fast, efficient, and unsupervised algorithm SmartInf to detect a set of influential users by identifying anchor nodes from a temporal sequence of retweets in Twitter cascades. Such anchor nodes provide important signatures of tweet diffusion across multiple diffusion localities and, hence, act as precursors for detection of influential nodes.1The set of influential nodes identified by SmartInf has the capacity to expose the tweet to a large and diverse population, when targeted as seeds thereby maximizing the influence spread. Experimental evaluation on empirical datasets from Twitter shows the superiority of SmartInf over state-of-the-art baselines in terms of infecting larger population; further, our evaluation shows that SmartInf is scalable to large-scale networks and is robust to missing data. Finally, we investigate the key factors behind the improved performance of SmartInf by testing our algorithm on a synthetic network using synthetic cascades simulated on this network. Our results reveal the effectiveness of SmartInf in identifying a diverse set of influential users that facilitate faster diffusion of tweets to a larger population.1We use the terms “influential nodes” and “influential users” interchangeably.

In an organization, individuals preferto form various formal and informal groups for mutual interactions. Therefore, ubiquitous identification of such groups and understanding their dynamics are important to monitor activities, behaviors, and well-being of the individuals. In this paper, we develop a lightweight, yet near-accurate, methodology, called GroupSense, to identify various interacting groups based on collective sensing through users' smartphones. Group detection from sensor signals is not straightforward because users in proximity may not always be under the same group. Therefore, we use acoustic context extracted from audio signals to infer the interaction pattern among the subjects in proximity. We have developed an unsupervised and lightweight mechanism for user group detection by taking cues from network science and measuring the cohesivity of the detected groups regarding modularity. Taking modularity into consideration, GroupSense can efficiently eliminate incorrect groups, as well as adapt the mechanism depending on the role played by the proximity and the acoustic context in a specific scenario. The proposed method has been implemented and tested under many real-life scenarios in an academic institute environment, and we observe that GroupSense can identify user groups with on an average 0:9(±0:14) F1-Score even in a noisy environment.

Twitter is currently a popular microblogging platform for the dissemination of information by users in the form of messages such as tweets. Such tweets are shared with followers of the seed user who in turn may reshare it with their own set of followers. Long chain of such retweets form cascades. In this paper, we aim to estimate the influence tree of cascades denoting the who-influenced-whom relationship among retweeting users by leveraging on temporal pattern of its retweets. We use a principled methodology to construct the ground truth influence trees of cascades using standard diffusion models. We define diverse structural metrics to quantify the structural characteristics of different influence trees. Based on empirical observations from ground truth, we develop CasCon, an unsupervised model that leverages on temporal pattern of retweets obtained from time series of cascades and underlying follower network of Twitter to construct influence trees. Further, we provide an analytical formulation of CasCon by deriving the degree distribution of these predicted influence trees and use it to approximate the structural metrics. Our evaluation shows that CasCon exhibits superior performance compared to state-of-theart baseline algorithms in selecting influencers of high quality based on standard influence measures in Twitter as well as in correctly (re)constructing the ground truth influence trees with high accuracy. Finally, we validate the analytical formulation of CasCon on the influence tree structures of both synthetic and real cascades; experimental results demonstrate its effectiveness in closely resembling ground truth influence trees for empirical cascades with high retweet count.

With the increase in road fatalities due to various factors like aggressive driving and road rage, quantifying and monitoring the stress level of a driver is an important task for the preparation of driving rosters for the cab companies. Stress monitoring using physiological sensors is a costly and obstructive task, while stress factors impact differently for different individuals based on their personality traits. In this paper, we develop a learning-based model to predict the stress level of a driver and its effect on his driving behavior, solely based on spatio-temporal driving data collected through GPS and inertial sensors. We further establish a correlation between the stress level of a driver and his driving behavior; thus, we develop a complete system to infer stress profiling and its impact on driving behavior based on spatio-temporal driving data. The model has been tested over a publicly available dataset with 6 drivers for 500 minutes of driving data. We observe that the proposed model gives an average prediction accuracy of 79% with low false-positive rates.

In day to day life, people meet strangers while commuting in public transports, roaming around in a shopping mall, waiting at airport boarding areas etc., and thus form passively encountering groups. Detection and analysis of such groups are essential for providing services like targeted advertisements, supply chain management, information broadcasting and so on. However, identifying such groups is challenging because of the underlying dynamics, where an encounter between two subjects is entirely instantaneous without having a specific pattern. This problem has two steps - (a) identification of subjects in proximity and (b) detecting groups from the proximity information. In this paper, we develop an unsupervised model to identify subjects in proximity based on WiFi signal information and assign a proximity score to each pair of subjects based on a novel metric defining the degree of proximity. With the help of these concepts from network science, we then utilize a community detection mechanism to infer the passively encountering groups from the proximity score. The proposed model has been implemented and deployed over an academic institute campus. A study over 25 subjects for six months reveals that the proposed model can detect passively encountering groups with more than 90\% accuracy, even with heterogeneous devices under various real-life scenarios.

A large population of users gets affected by sudden slowdown or shutdown of an enterprise application. System administrators and analysts spend considerable amount of time dealing with functional and performance bugs. These problems are particularly hard to detect and diagnose in most computer systems, since there is a huge amount of system generated supportability data (counters, logs etc.) that need to be analyzed. Most often, there isn't a very clear or obvious root cause. Timely identification of significant change in application behavior is very important to prevent negative impact on the service. In this paper, we present ADELE, an empirical, data-driven methodology for early detection of anomalies in data storage systems. The key feature of our solution is diligent selection of features from system logs and development of effective machine learning techniques for anomaly prediction. ADELE learns from system's own history to establish the baseline of normal behavior and gives accurate indications of the time period when something is amiss for a system. Validation on more than 4800 actual support cases shows~83% true positive rate and~12% false positive rate in identifying periods when the machine is not performing normally. We also establish the existence of problem “signatures” which help map customer problems to already seen issues in the field. ADELE's capability to predict early paves way for online failure prediction for customer systems.

Recent context detection techniques in smartphones leverage on the embedded motion sensors, which in turn increases the potential of side-channel attacks. We in this paper propose an alternative modality for obtaining mobility context using smartphone keyboard interaction patterns using a personalized framework. Experimental results show that the framework can predict the mobility context at an average F1 score (both micro and macro) greater than 0.6 across all subjects.

Typing characteristics on smartphones can provide clues for emotion detection. Collecting large volumes of typing data is also easy on smartphones. This motivates the use of Deep Neural Network (DNN) to determine emotion states from smartphone typing. In this work, we developed a DNN model based on typing features to predict four emotion states (happy, sad, stressed, relaxed) and investigate its performance on a smartphone. The evaluation of the model in a 3-week study with 15 participants reveals that it can reliably detect emotions with an average accuracy of 80% with peak CPU utilization less than 15%.

Passenger comfort is a major factor influencing a commuter's decision to avail public transport. Existing studies suggest that factors like overcrowding, jerkiness, traffic congestion etc. correlate well to passenger's (dis)comfort. An online survey conducted with more than 300 participants from 12 different countries reveals that different personalized and context dependent factors influence passenger comfort during a travel by public transport. Leveraging on these findings, we identify correlations between comfort level and these dynamic parameters, and implement a smartphone based application, ComfRide, which recommends the most comfortable route based on user's preference honoring her travel time constraint. We use a 'Dynamic Input/Output Automata' based composition model to capture both the wide varieties of comfort choices from the commuters and the impact of environment on the comfort parameters. Evaluation of ComfRide, involving 50 participants over 28 routes in a state capital of India, reveals that recommended routes have on average 30% better comfort level than Google map recommended routes, when a commuter gives priority to specific comfort parameters of her choice.

In Affective Computing, different modalities, such as speech, facial expressions, physiological properties, smart-phone usage patterns, and their combinations, are applied to detect the affective states of a user. Keystroke analysis i.e. study of the typing behavior in desktop computer is found to be an effective modality for emotion detection because of its reliability, non-intrusiveness and low resource overhead. As smartphones proliferate, typing behavior on smartphone presents an equally powerful modality for emotion detection. It has the added advantage to run in-situ experiments with better coverage than the experiments using desktop computer keyboards. This work explores the efficacy of smartphone typing to detect multiple affective states. We use a qualitative and experimental approach to answer the question. We conduct an online survey among 120 participants to understand the typing habits in smartphones and collect feedback on multiple measurable parameters that affect their emotion while typing. The findings lead us to design and implement an Android based emotion detection system, TapSense, which can identify four different emotion states (happy, sad, stressed, relaxed) with an average accuracy (AUCROC) of 73% (maximum of 94%) based on typing features only. The analysis also reveals that among different features, typing speed is the most discriminative one.

Twitter has recently become one of the most popular online social networking websites where users can share news and ideas through messages in the form of tweets. As a tweet gets retweeted from user to user, large cascades of information diffusion are formed over the Twitter follower network. Existing works on cascades have mainly focused on predicting their popularity in terms of size. In this paper, we leverage on the temporal pattern of retweets to model the diffusion dynamics of a cascade. Notably, retweet cascades provide two complementary information: (a) inter-retweet time intervals of retweets, and (b) diffusion of cascade over the underlying follower network. Using datasets from Twitter, we identify two types of cascades based on presence or absence of early peaks in their sequence of inter-retweet intervals. We identify multiple diffusion localities associated with a cascade as it propagates over the network. Our studies reveal the transition of a cascade to a new locality facilitated by pivotal users that are highly cascade dependent following saturation of current locality. We propose an analytical model to show co-occurrence of first peaks and cascade migration to a new locality as well as predict locality saturation from inter-retweet intervals. Finally, we validate these claims from empirical data showing co-occurrence of first peaks and migration with good accuracy; we obtain even better accuracy for successfully classifying saturated and non-saturated diffusion localities from inter-retweet intervals.

Experience Sampling Method (ESM) is widely used in idiographic approaches to collect within-person patterns. Planning a suitable survey schedule while designing an ESM based experiment is challenging as it must balance between survey fatigue of users, and the timeliness and accuracy of the responses provided by users. Even with the proliferation of ESM experiments, survey scheduling typically remains confined to use of fixed schedules, where periodic probes are sent to user, or event-based schedules, where depending on the number of events, large number of probes may interrupt user frequently. We propose a novel survey scheduling scheme, Low-Interference High-fidelity (LIHF) ESM schedule, which is designed to reduce interference while retaining fidelity of user response. We integrated LIHF into an ESM application, called TapSense, that is used to infer user's emotion from typing characteristics on smartphone keypad. Conducting a 2-week field study involving 9 users, using proposed metrics we show that using LIHF there is 26% reduction in survey fatigue, 50% improvement in triggering survey probes in timely manner, and 8% improvement in predicting emotion states based on typing patterns compared to typical ESM scheduling techniques.

Community detection in single layer, isolated networks has been extensively studied in the past decade. However, many real-world systems can be naturally conceptualized as multilayer networks which embed multiple types of nodes and relations. In this paper, we propose algorithm for detecting communities in multilayer networks. The crux of the algorithm is based on the multilayer modularity index Q_M, developed in this paper. The proposed algorithm is parameter-free, scalable and adaptable to complex network structures. More importantly, it can simultaneously detect communities consisting of only single type, as well as multiple types of nodes (and edges). We develop a methodology to create synthetic networks with benchmark multilayer communities. We evaluate the performance of the proposed community detection algorithm both in the controlled environment (with synthetic benchmark communities) and on the empirical datasets (Yelp and Meetup datasets); in both cases, the proposed algorithm outperforms the competing state-of-the-art algorithms.

City transit maps are one of the important resources for public navigation in today's digital world. However, the availability of transit maps for many developing countries is very limited, primarily due to the various socio-economic factors that drive the private operated and partially regulated transport services. Public transports at these cities are marred with many factors such as uncoordinated waiting time at bus stoppages, crowding in the bus, sporadic road conditions etc., which also need to be annotated so that commuters can take informed decision. Interestingly, many of these factors are spatio-temporal in nature. In this paper, we develop CityMap, a system to automatically extract transit routes along with their eccentricities from spatio-temporal crowdsensed data collected via commuters' smart-phones. We apply a learning based methodology coupled with a feature selection mechanism to filter out the necessary information from raw smart-phone sensor data with minimal user engagement and drain of battery power. A thorough evaluation of CityMap, conducted for more than two years over 11 different routes in 3 different cities in India, show that the system effectively annotates bus routes along with other route and road features with more than 90% of accuracy.

The penetration of smartphones and advancement of the location-based services have jointly contributed to the popularity of location-based social networking platforms such as Foursquare and Yelp. Business owners are important stakeholders of such location-based social network (LBSN) platforms, for whom setting up new businesses at the right location has immense importance in making a business popular. Given a wide variety of businesses, a single metric may not correctly capture the business popularity, rather, popularity measure should be specific to geographical regions, business category etc. In this paper, we systematically define popularity metrics and use them to label diverse businesses available in Yelp dataset. Notably, information diffusion plays a major role in popularizing businesses, word of mouth communication across customers facilitates information spread. This paper concentrates on two key modalities that exhibit a strong signature of business popularity: social connections across customers, showing high correlation with corresponding business popularity, and geographical proximity, which reveals two types of customers, namely, local visitor and foreign visitor observed through a comprehensive study on Yelp dataset. Depending on the business type and category, either local or foreign visitors play important role in making a business popular. Finally, we leverage on these signatures to predict business popularity and develop a recommendation model suggesting top regions to business owners for starting popular businesses. We show that both the models outperform the state-of-the-art baseline algorithms.

This paper explores the feasibility of automatically discriminating users from the activity as well as temporal information of their daily routine. We observe that everyone pursues a daily semi-regular activity pattern. Based on this observation, we have developed a system UDAT and experimented on Microsoft Geolife as well as UDAT datasets. With Geolife transportation activity log and UDAT motion-static activity log, the system achieves 73.3% and 80.68% accuracy, respectively. Although the overall system accuracy is moderate, the system achieves the highest accuracy when the users belong to the different activity buckets. This signifies the utility of two-phase classification for user discrimination.

Typing based communication applications on smartphones, like WhatsApp, can induce emotional exchanges. The effects of an emotion in one session of communication can persist across sessions. In this work, we attempt automatic emotion detection by jointly modeling the typing characteristics, and the persistence of emotion. Typing characteristics, like speed, number of mistakes, special characters used, are inferred from typing sessions. Self reports recording emotion states after typing sessions capture persistence of emotion. We use this data to train a personalized machine learning model for multi-state emotion classification. We implemented an Android based smartphone application, called TapSense, that records typing related metadata, and uses a carefully designed Experience Sampling Method (ESM) to collect emotion self reports. We are able to classify four emotion states - happy, sad, stressed, and relaxed, with an average accuracy (AUCROC) of 84% for a group of 22 participants who installed and used TapSense for 3 weeks.

Success of Meetup groups is of utmost importance for the members who organize them. Given a wide variety of such groups, a single metric may not be indicative of success for different groups; rather, success measure should be specific to the interest of a group. In this paper, accounting for the group diversity, we systematically define Meetup group success metrics and use them to generate labels for our machine learnt models. We crawl the Meetup dataset for three US cities namely New York, Chicago and San Francisco over a period of 8 months. The data study reveals the key players (such as core members, new members etc.) behind the success of the Meetup groups. This study leverages semantic, syntactic, temporal and location based features to discriminate between successful and unsuccessful groups. Finally, we present a model to predict success of the Meetup groups with high accuracy (0.81 with AUC = 0.86). Our approach generalizes well across groups, categories and cities. Additionally, the model performs reasonably well for new groups with little history (cold start problem), exhibiting high accuracy for the cross city validation.

In this paper, we investigate the role of mentions on tweet propagation. We propose a novel tweet propagation model SIR_MF based on a multiplex network framework, that allows to analyze the effects of mentioning on final retweet count. The basic bricks of this model are supported by a comprehensive study of multiple real datasets and simulations of the model show a nice agreement with the empirically observed tweet popularity. Studies and experiments also reveal that follower count, retweet rate & profile similarity are important factors in gaining tweet popularity and allow to better understand the impact of the mention strategies on the retweet count. Interestingly, we analytically identify a critical retweet rate regulating the role of mention on the tweet popularity. Finally, our data driven simulation demonstrates that the proposed mention recommendation heuristic "Easy-Mention" outperforms the benchmark "Whom-To-Mention" algorithm.

Public bus services in many cities in countries like India are controlled by private owners, hence, building up a database for all the bus routes is non-trivial. In this paper, we leverage smart-phone based sensing to crowdsource and populate the information repository for bus routes in a city. We have developed an intelligent data logging module for smart-phones and a server side processing mechanism to extract roads and bus routes information. From a 3 month long study involving more than 30 volunteers in 3 different cities in India, we found that the developed system, CrowdMap, can annotate bus routes with a mean error of 10m, while consuming 80% less energy compared to a continuous GPS based system.

Public transport in suburban cities (covers 80% of the urban landscape) of developing regions suffer from the lack of information in Google Transit, unpredictable travel times, chaotic schedules, absence of information board inside the vehicle. Consequently, passengers suffer from lack of information about the exact location where the bus is at present as well as the estimated time to be taken to reach the desired destination. We find that off-the-shelf deployment of existing (non-GPS) localization schemes exhibit high error due to sparsity of stable and structured outdoor landmarks (anchor points). Through rigorous experiments conducted over a month however, we realize that there are a certain class of volatile landmarks which may be useful in developing efficient localization scheme. Consequently, in this paper, we design a novel generalized energy-efficient outdoor localization scheme - UrbanEye, which efficiently combines the volatile and non-volatile landmarks using a specialized data structure, the probabilistic timed automata. UrbanEye uses speed-breakers, turns and stops as landmarks, estimates the travel time with a mean accuracy of ±2.5 mins and produces a mean localization accuracy of 50 m. Results from several runs taken in two cities, Durgapur and Kharagpur, reveal that UrbanEye provides more than 50% better localization accuracy compared to the existing system Dejavu [1], and consumes significantly less energy.

This paper explores the feasibility of automatically extracting passwords from a user's daily activity logs, such as her Facebook activity, phone activity etc. As an example, a smartphone might ask the user: "Today morning from whom did you receive an SMS?" In this paper, we observe that infrequent activities (i.e., outliers) can be memorable and unpredictable. Building on this observation, we have developed an end to end system ActivPass and experimented with 70 users. With activity logs from Facebook, browsing history, call logs, and SMSs, the system achieves 95% success (authenticates legitimate users) and is compromised in 5.5% cases (authenticates impostors). While this level of security is obviously inadequate for serious authentication systems, certain practices such as password sharing can immediately be thwarted from the dynamic nature of passwords. With security improvements in the future, activity-based authentication could fill in for the inadequacies in today's password-based systems.

In superpeer based networks, the superpeers are discovered through the process of bootstrapping, whereby resourceful peers get upgraded to superpeers. However, bootstrapping is influenced by several factors like limitation on the maximum number of connections a peer can have due to bandwidth constraints, limitation on the availability of information of existing peers due to cache size constraints and also by the attachment policy of the newly arriving peers to the resourceful peers. In this paper, we derive closed form equations that model the effect of these factors on superpeer related topological properties of the networks. Based on the model, we observe that the cache parameters and the preferentiality parameters must be suitably tuned so as to increase the fraction of superpeers in the network. Finally, we perform an empirical analysis of social networks like Twitter and Facebook using our model to obtain and derive insights for suitably bootstrapping superpeer topology.

WiFi clients must associate to a specific Access Point (AP) to communicate over the Internet. Current association methods are based on maximum Received Signal Strength Index (RSSI) implying that a client associates to the strongest AP around it. This is a simple scheme that has performed well in purely distributed settings. Modern wireless networks, however, are increasingly being connected by a wired backbone. The backbone allows for out-of-band communication among APs, opening up opportunities for improved protocol design. This paper takes advantage of this opportunity through a coordinated client association scheme where APs consider a global view of the network, and decide on the optimal client-AP association. We show that such an association outperforms RSSI based schemes in several scenarios, while remaining practical and scalable for wide-scale deployment. We also show that optimal association is a NP-Hard problem and our max-flow based heuristic is a promising solution.

Wireless clients must associate to a specific Access Point (AP) to communicate over the Internet. Current association methods are based on maximum Received Signal Strength Index (RSSI) implying that a client associates to the strongest AP around it. This is a simple scheme that has performed well in purely distributed settings. Modern wireless networks, however, are increasingly being connected by a wired backbone. The backbone allows for out-of-band communication among APs, opening up opportunities for improved protocol design. This paper takes advantage of this opportunity through a coordinated client association scheme where APs consider a global view of the network, and decides on the optimal client-AP association. We show that such an association outperforms RSSI based schemes in several scenarios, while remaining practical and scalable for wide-scale deployment. Although an early work in this direction, our basic analytical framework (based on a max-flow formulation) can be extended to sophisticated channel and traffic models. Our future work is focussed towards designing and evaluating these extensions.

In this paper, we develop an analytical framework which explains the emergence of superpeer networks on execution of the commercial peer-to-peer bootstrapping protocols by incoming nodes. Bootstrapping protocols exploit physical properties of the online peers like resource content, processing power, storage space, connectivity etc as well as take the finiteness of bandwidth of each online peer into consideration. With the help of rate equations, we show that execution of these protocols results in the emergence of superpeer nodes in the network - the exact degree distribution is evaluated. We validate the framework developed in this paper through extensive simulation. The analysis of the results shows that the amount of superpeers produced in the network depends on the protocol as well as the properties of the joining nodes. Interestingly, our analysis reveals that increase in the amount of resource and the number of resourceful nodes do not always help to increase the fraction of superpeer nodes in the network. The impact of the frequent leaving of the peers on the topology of the emerging network is also evaluated. As an application study, we show that our framework can explain the topological configuration of commercial Gnutella networks. The developed model can almost perfectly match the degree distribution of Gnutella network.

In this paper, we develop an analytical framework to explain the appearance of bimodal degree distribution in existing popular superpeer networks like Gnutella. The framework leads to several interesting and important inferences related to network topology and node/link dynamics. Beyond the simulation based study, we validate our framework through almost exact matching of the topological structure of Gnutella network constructed from real data.

This paper proposes a fair association scheme between clients and APs in WiFi network, exploiting the hybrid nature of the recent WLAN architecture. We show that such an association outperforms RSSI based schemes in several scenarios, while remaining practical and scalable for wide-scale deployment.

In this paper, we develop an analytical framework to measure the vulnerability of superpeer networks against attack. Two different kinds of attacks namely deterministic and degree dependent attack have been introduced here. We formally model the superpeer networks with the help of bimodal structure and different attacks with the help of graph dynamics. Our analysis shows that fraction of superpeers and their connectivity have profound impact upon the stability of the network. The results obtained from the theoretical analysis are validated through simulation. The agreement between the simulation results and theoretical predictions is almost perfect.

Many previous works of Peer-to-Peer traffic characterization and modeling focused their attention on the distribution of query contents. However, few has been done towards a better understanding of the time series distribution of these queries, which is vital for system performance. To remedy this situation, this paper characterizes query traffic by using automatic time series analysis to evaluate different linear models(Box-Jenkins models and some simple windowed-mean models) for predicting the number of duplicated queries from 10 minutes to 2 hours into the future. Both the predictive power and the computational costs of these models are evaluated over 318,942,450 real world Gnutella queries collected over 3 months. We find the number of duplicated queries is consistently predictable. Simple, practical models like AR perform well on prediction. To show that these characteristics have a wide range of potential applications, we propose two enhancement to existing search results caching and load balancing algorithms. Our simulation study shows that our methodology works quite well in both scenarios in terms of efficiency and effectiveness. The main contribution of this paper lies in: (1) proposing new measurement techniques on Gnutella, (2) characterizing and modeling peer-to-peer query traffic with Box-Jenkins Models, (3) presenting a general enhancement to existing performance optimization algorithm in P2P systems.

No abstract available.

In this paper, we present a polynomial time algorithm that gives an optimal solution to the routing and wavelength assignment (RWA) problem in a tree topology. One of the major design issues in wavelength-division multiplexed networks is the assignment of the limited number of wavelengths among network stations so that greater capacity can be achieved. The problem of RWA is known to be NP-hard problem. Many researchers have tackled the problem of RWA with a number of efficient heuristic algorithms. This paper presents an algorithm that optimally assigns a single wavelength to maximize one-hop traffic in a tree topology. The algorithm uses dynamic programming and is shown to be optimal with a time complexity of O(N/sup 4/). We also propose a heuristic scheme to use our optimal algorithm for wavelength assignment in a general graph. The heuristic works on the tree subgraphs of a given graph and the remaining spare wavelengths can be assigned with an existing RWA policy.