Preference mining techniques for customer behavior analysis

The thesis has studied a number of critical problems in data mining for customer behavior analysis and has proposed novel techniques for better modeling of the customers’ decision making process, more efficient analysis of their travel behavior, and more effective identification of their emerging preference.

Network dynamics of mobile social networks

With the rapid development of smartphones and mobile Internet technology, we witness an overwhelming growth of mobile social networks (MSN), which is a type of social network, forming virtual communities among people with similar interests or commonalities. In MSNs, users play a crucial role for their development, deployment and success. Understanding the MSN user behavior therefore attracts interests of different entities - ISPs, service providers, and researchers. However, it is hard to gather a comprehensive real data set, little is known and even less has been published about MSN user activities. In this paper, we focus on analyzing MSN user behavior from the perspective of ISP network, which is seldom reported in literature. Based on the real data set collected from the mobile network gateway of a major mobile carrier who has more than five million subscribers, we present an in-depth user behavior analysis of four popular social networks. We study the MSN user behavior from six aspects: user requests, active online time, sessions, inter-session, the number of requests in a session, and inter-request. We found that power law and lognormal are two popular features of the studied objects, and exposed some interesting findings as well. We hope our work could be helpful for ISPs, MSN content providers, and researchers. © 2014 IEEE.

Exploring probabilistic follow relationship to prevent collusive peer-to-peer piracy

P2P collusive piracy, where paid P2P clients share the content with unpaid clients, has drawn significant concerns in recent years. Study on the follow relationship provides an emerging track of research in capturing the followee (e.g., paid client) for the blocking of piracy spread from all his followers (e.g., unpaid clients). Unfortunately, existing research efforts on the follow relationship in online social network have largely overlooked the time constraint and the content feedback in sequential behavior analysis. Hence, how to consider these two characteristics for effective P2P collusive piracy prevention remains an open problem. In this paper, we proposed a multi-bloom filter circle to facilitate the time-constraint storage and query of P2P sequential behaviors. Then, a probabilistic follow with content feedback model to fast discover and quantify the probabilistic follow relationship is further developed, and then, the corresponding approach to piracy prevention is designed. The extensive experimental analysis demonstrates the capability of the proposed approach.

Time use clusters of New Zealand adolescents are associated with weight status, diet and ethnicity

OBJECTIVE: To describe New Zealand adolescent time use clusters and correlate cluster profiles. METHODS: Data were from the cross-sectional 2008/2009 National Survey of Children and Young People's Physical Activity and Dietary Behaviours, which surveyed a random sample of 10-16 year-old New Zealanders (study subset n=679). Time use data were collected using the Multimedia Activity Recall for Children and Adults, and collapsed into 17 age-adjusted variables for sex-specific cluster analysis. Cluster associations with socio-demographic, anthropometric, physical activity and dietary variables were analysed. RESULTS: Three time use clusters were discovered for both boys and girls. For boys, the Techno-active cluster was characterised by high levels of team sports and TV; the Quiet movers cluster by transport (active and passive) and quiet time; and the Social studious cluster by reading, study activities and social interaction. The boys' clusters were associated with ethnicity. The girls'Social sporty cluster was characterised by sports and social interaction; the Screenie tasker cluster by TV, computer, chores and work; and the Super studious cluster by reading, study and school-based activities. The girls' time use cluster membership was associated with weight status and serves of extra foods. CONCLUSIONS: Distinct sex-specific time use clusters and correlate profiles exist among NZ adolescents. IMPLICATIONS: These findings may assist the development of targeted time use interventions to improve adolescent health and well-being.

Weibull analysis of the tensile behavior of fibers with geometrical irregularities

This paper further develops the conventional Weibull/weakest-link model by incorporating the within-fiber diameter variation. This is necessary for fibers with considerable geometrical irregularities, such as the wool and other animal fibers. The strength of wool fibers has been verified to follow this modified Weibull/weakest-link distribution. In addition, the modified Weibull model can predict the gauge length effect more accurately than the conventional model.

Cyclic deformation of advanced high-strength steels : mechanical behavior and microstructural analysis

The fatigue properties of multiphase steels are an important consideration in the automotive industry. The different microstructural phases present in these steels can influence the strain life and cyclic stabilized strength of the material due to the way in which these phases accommodate the applied cyclic strain. Fully reversed strain-controlled low-cycle fatigue tests have been used to determine the mechanical fatigue performance of a dual-phase (DP) 590 and transformation-induced plasticity (TRIP) 780 steel, with transmission electron microscopy (TEM) used to examine the deformed microstructures. It is shown that the higher strain life and cyclic stabilized strength of the TRIP steel can be attributed to an increased yield strength. Despite the presence of significant levels of retained austenite in the TRIP steel, both steels exhibited similar cyclic softening behavior at a range of strain amplitudes due to comparable ferrite volume fractions and yielding characteristics. Both steels formed low-energy dislocation structures in the ferrite during cyclic straining.

Constitutive analysis of hot deformation behavior of a Ti6Al4V alloy using physical based model

The effect of deformation parameters on the flow behavior of a Ti6Al4V alloy has been studied to understand the deformation mechanisms during hot compression. Cylindrical samples with partially equiaxed grains were deformed in the α+β phase region at different thermo-mechanical conditions. To develop components with tailored properties, the physically based Estrin and Mecking (EM) model for the work hardening/dynamic recovery combined with the Avrami equation for dynamic recrystallization was used to predict the flow stress at varying process conditions. The EM model revealed good predictability up to the peak strain, however, at strain rates below 0.01s^-1, a higher B value was observed due to the reduced density of dislocation tangles. In contrast, the flow softening model revealed higher value of constants a and b at high strain rates due to the reduction in the volume fraction of dynamic recrystallization and larger peak strain. The predicted flow stress using the combined EM+Avrami model revealed good agreement with the measured flow stress resulted in very low average absolute relative error value. The microstructural analysis of the samples suggests the formation of coarse equiaxed grains together with the increased β phase fraction at low strain rate leads to a higher flow softening.

Hierarchical analysis of avian re-nesting behavior: mean, across-individual, and intra-individual responses

© 2015, Springer-Verlag Berlin Heidelberg. Anti-predator behavior is a key aspect of life history evolution, usually studied at the population (mean), or across-individual levels. However individuals can also differ in their intra-individual (residual) variation, but to our knowledge, this has only been studied once before in free-living animals. Here we studied the distances moved and changes in nest height and concealment between successive nesting attempts of marked pairs of grey fantails (Rhipidura albiscapa) in relation to nest fate, across the breeding season. We predicted that females (gender that decides where the nest is placed) should on average show adaptive behavioral responses to the experience of prior predation risk such that after an unsuccessful nesting attempt, replacement nests should be further away, higher from the ground, and more concealed compared with replacement nests after successful nesting attempts. We found that, on average, females moved greater distances to re-nest after unsuccessful nesting attempts (abandoned or depredated) in contrast to after a successful attempt, suggesting that re-nesting decisions are sensitive to risk. We found no consistent across-individual differences in distances moved, heights, or concealment. However, females differed by 53-fold (or more) in their intra-individual variability (i.e., predictability) with respect to distances moved and changes in nest height between nesting attempts, indicating that either some systematic variation went unexplained and/or females have inherently different predictability. Ignoring these individual differences in residual variance in our models obscured the effect of nest fate on re-nesting decisions that were evident at the mean level.

Integrating network analysis, sensor tags, and observation to understand shark ecology and behavior

Group living in animals is a well-studied phenomenon, having been documented extensively in a wide range of terrestrial, freshwater, and marine species. Although social dynamics are complex across space and time, recent technological and analytical advances enable deeper understanding of their nature and ecological implications. While for some taxa, a great deal of information is known regarding the mechanistic underpinnings of these social processes, knowledge of these mechanisms in elasmobranchs is lacking. Here, we used an integrative and novel combination of direct observation, accelerometer biologgers, and recent advances in network analysis to better understand the mechanistic bases of individual-level differences in sociality (leadership, network attributes) and diel patterns of locomotor activity in a widespread marine predator, the lemon shark (Negaprion brevirostris). We found that dynamic models of interaction based on Markov chains can accurately predict juvenile lemon shark social behavior and that lemon sharks did not occupy consistent positions within their network. Lemon sharks did however preferentially associate with specific group members, by sex as well as by similarity or nonsimilarity for a number of behavioral (nonsimilarity: leadership) and locomotor traits (similarity: proportion of time swimming "fast," mean swim duration; nonsimilarity: proportion of swimming bursts/transitions between activity states). Our study provides some of the first information on the mechanistic bases of group living and personality in sharks and further, a potential experimental approach for studying fine-scale differences in behavior and locomotor patterns in difficult-to-study organisms.

mHealth technologies to influence physical activity and sedentary behaviors: behavior change techniques, systematic review and meta-analysis of randomized controlled trials

BACKGROUND: mHealth programs offer potential for practical and cost-effective delivery of interventions capable of reaching many individuals.

PURPOSE: To (1) compare the effectiveness of mHealth interventions to promote physical activity (PA) and reduce sedentary behavior (SB) in free-living young people and adults with a comparator exposed to usual care/minimal intervention; (2) determine whether, and to what extent, such interventions affect PA and SB levels and (3) use the taxonomy of behavior change techniques (BCTs) to describe intervention characteristics.

METHODS: A systematic review and meta-analysis following PRISMA guidelines was undertaken to identify randomized controlled trials (RCTs) comparing mHealth interventions with usual or minimal care among individuals free from conditions that could limit PA. Total PA, moderate-to-vigorous intensity physical activity (MVPA), walking and SB outcomes were extracted. Intervention content was independently coded following the 93-item taxonomy of BCTs.

RESULTS: Twenty-one RCTs (1701 participants-700 with objectively measured PA) met eligibility criteria. SB decreased more following mHealth interventions than after usual care (standardised mean difference (SMD) -0.26, 95 % confidence interval (CI) -0.53 to -0.00). Summary effects across studies were small to moderate and non-significant for total PA (SMD 0.14, 95 % CI -0.12 to 0.41); MVPA (SMD 0.37, 95 % CI -0.03 to 0.77); and walking (SMD 0.14, 95 % CI -0.01 to 0.29). BCTs were employed more frequently in intervention (mean = 6.9, range 2 to 12) than in comparator conditions (mean = 3.1, range 0 to 10). Of all BCTs, only 31 were employed in intervention conditions.

CONCLUSIONS: Current mHealth interventions have small effects on PA/SB. Technological advancements will enable more comprehensive, interactive and responsive intervention delivery. Future mHealth PA studies should ensure that all the active ingredients of the intervention are reported in sufficient detail.

Isotemporal substitution analysis for sedentary behavior and body mass index

This study examined the prospective associations of reallocating time spent in different types of sedentary behavior, physical activity, and sleep with body mass index (BMI) in children using isotemporal substitution analysis. Methods Chinese children in grades 1-3 were recruited to a cohort study in 2009 and were followed up over a 2-yr period. Reports were gathered from the parents on children's sedentary behavior, sleep duration, and sociodemographic variables. The reported sedentary behavior types were then grouped into three categories: screen time (e.g., watching TV), academic-related activities (e.g., doing homework), and other sedentary behaviors (e.g., sitting and talking). Moderate-to-vigorous physical activity (MVPA) and light-intensity physical activity were assessed by ActiGraph accelerometry (ActiGraph, Pensacola, FL). Isotemporal substitution models were performed to examine the effects of time allocation on BMI. Results A total of 672 children (359 boys, mean age at recruitment = 7.6 yr) who had provided valid accelerometer data for at least one assessment time point were included in the analysis. Controlling for covariates and total behavior time, isotemporal substitution models indicated that the displacement of 30 min·d -1 of other sedentary behaviors with equal amounts of screen time (B = 0.12; 95% confidence interval, 0.04-0.20) or academic-related activities (B = 0.13; 95% confidence interval, 0.04-0.21) was associated with higher BMI. Reallocating 30 min·d -1 of MVPA with each of the sedentary behavior variables resulted in increased BMI. Conclusions The substitution of screen time or academic-related activities with other sedentary behaviors or MVPA was associated with lower BMI in Chinese children.

Behavioral versus demographic predictors of early adoption : a critical analysis and comparative test

Predicting which consumers will be amongst the first to adopt an innovative product is a difficult task but is valuable in allowing effective and efficient use of marketing resources. This paper examines the accuracy of predictions made about likely first adopters based on the most widely accepted theory and compares them to predictions made by examining the relevant past behavior of consumers. A survey of over 1000 consumers examined adoption of an innovative technology: compact fluorescent lightglobes. The results show that variables which were derived from a utility and awareness perspective were a more accurate and managerially useful predictor than the demographic variables derived from the widely accepted theory based on the work of Rogers. It is suggested that these alternative variables could be utilized more readily by marketing managers in many circumstances.

Spatial and temporal distribution of solute leaching in heterogeneous soils: analysis and application to multisampler lysimeter data

Accurate assessment of the fate of salts, nutrients, and pollutants in natural, heterogeneous soils requires a proper quantification of both spatial and temporal solute spreading during solute movement. The number of experiments with multisampler devices that measure solute leaching as a function of space and time is increasing. The breakthrough curve (BTC) can characterize the temporal aspect of solute leaching, and recently the spatial solute distribution curve (SSDC) was introduced to describe the spatial solute distribution. We combined and extended both concepts to develop a tool for the comprehensive analysis of the full spatio-temporal behavior of solute leaching. The sampling locations are ranked in order of descending amount of total leaching (defined as the cumulative leaching from an individual compartment at the end of the experiment), thus collapsing both spatial axes of the sampling plane into one. The leaching process can then be described by a curved surface that is a function of the single spatial coordinate and time. This leaching surface is scaled to integrate to unity, and termed S can efficiently represent data from multisampler solute transport experiments or simulation results from multidimensional solute transport models. The mathematical relationships between the scaled leaching surface S, the BTC, and the SSDC are established. Any desired characteristic of the leaching process can be derived from S. The analysis was applied to a chloride leaching experiment on a lysimeter with 300 drainage compartments of 25 cm2 each. The sandy soil monolith in the lysimeter exhibited fingered flow in the water-repellent top layer. The observed S demonstrated the absence of a sharp separation between fingers and dry areas, owing to diverging flow in the wettable soil below the fingers. Times-to-peak, maximum solute fluxes, and total leaching varied more in high-leaching than in low-leaching compartments. This suggests a stochastic–convective transport process in the high-flow streamtubes, while convection–dispersion is predominant in the low-flow areas. S can be viewed as a bivariate probability density function. Its marginal distributions are the BTC of all sampling locations combined, and the SSDC of cumulative solute leaching at the end of the experiment. The observed S cannot be represented by assuming complete independence between its marginal distributions, indicating that S contains information about the leaching process that cannot be derived from the combination of the BTC and the SSDC.