Detection Of Behavioral Patterns Research Articles

User behavior pattern detection in unstructured processes – a learning management system case study

ABSTRACTProcess mining methodologies are designed to uncover underlying business processes, deviations from them, and in general, usage patterns. One of the key limitations of these methodologies is that they struggle in cases in which there is no structured process, or when a process can be performed in many ways. Learning Management Systems are a classic case of unstructured processes since each learner follows a different learning process. In this paper, we address this limitation by proposing and validating the user behavior pattern detection (UBPD) methodology which is based on detecting very short user activities and clustering them based on shared variance to construct a more meaningful behavior. We develop and validate this methodology by using two datasets of unstructured processes from different implementations of a learning management system. The first dataset uses a gamified course where users have the freedom to choose how to use the system, and the second dataset uses data from a massive online open course, where again, system usage is based on personal learning preferences. The key contribution of the methodology is its ability to discover user-specific usage patterns and cluster users based on them, even in noisy systems with no clear process. It provides great value to course designers and teachers trying to understand how learner interact with their system and sets the foundation for additional research in this class of systems.

Analytic System to Evaluate Efficient Driving Programs in Professional Fleets

Interest in energy cost saving and in global warming have persuaded transport companies to apply measures to reduce fuel consumption. Efficient driving is one of the most employed solutions as it allows savings in fuel consumption of around 10% with a minimal investment. The drawback is that efficient driving is a learning process, and it greatly depends on the drivers’ behavior, which in turn is closely related to their motivation. If drivers are not really involved or after some time their interest decreases, efficiency improvements would disappear. Thus, an efficient driving program should make drivers’ motivation one of the main targets. One option could be the implementation of reward programs. However, these should be based on a clear individual evaluation process, as an unfair system could lead to discomfort, complaints, and repudiation. In this paper, we propose an analytic system, based on the detection of efficient and inefficient behavioral patterns, to evaluate the individual driver’s progression in efficient driving with the aim of being the basis of a reward program. The system receives relevant, driving related, vehicle information every 1.5 s, allowing a precise searching of patterns. It has been tested successfully in 16 bus companies, analyzing the performance of 880 professional drivers. To accurately illustrate the analytic process, three detailed driver analyses have been included as a case study. Results of this applied research on the eco-driving field show that the proposed system identifies efficient and inefficient actions that are used to fairly evaluate the drivers’ performance.

Performance and Home Advantage in Handball

The main objective of this research was to delve into the concept of playing handball at home from a classical perspective of previous studies. The emergence of regularities or certain patterns of play can be explained by the location of matches. Through an observational methodology and a nomothetic, monitoring and multidimensional design, thirty-nine European elite handball matches were selected (N = 39). A mixed ‘ad hoc’ registration system was designed. Records were made of the last three minutes of home and visiting games of which images were available. Two observers with observational methodology experience participated in the study. The quality of the records was dealt with in an intraobserver and interobserver way. Two different techniques were applied for the treatment of the data: a) detection of behavioral patterns by Theme (p < .005) (Magnusson, 2000) and b) Chaid decision trees (p < .05) using SPSSv.24. The results showed the existence of T-patterns according to the location (different T-patterns: home: 1085; away: 1242) of the matches. The categories most involved in effective launches were unveiled. On the other hand, the Chaid model also showed the effect of location (p < .001): home-win (node 9) (n = 149, 69%), away-win (node 10) (n = 15, 16%) in handball. Crossing analysis offered enriched interpretations to advance in the home advantage concept. From this study, guidelines can be drawn that may help handball coaches to build training tasks as differences in behavioral patterns between home or away play in handball were identified. Thus, designing tasks considering a match location has requires specificity.

Automatic Extraction of Behavioral Patterns for Elderly Mobility and Daily Routine Analysis

The elderly living in smart homes can have their daily movement recorded and analyzed. As different elders can have their own living habits, a methodology that can automatically identify their daily activities and discover their daily routines will be useful for better elderly care and support. In this article, we focus on automatic detection of behavioral patterns from the trajectory data of an individual for activity identification as well as daily routine discovery. The underlying challenges lie in the need to consider longer-range dependency of the sensor triggering events and spatiotemporal variations of the behavioral patterns exhibited by humans. We propose to represent the trajectory data using a behavior-aware flow graph that is a probabilistic finite state automaton with its nodes and edges attributed with some local behavior-aware features. We identify the underlying subflows as the behavioral patterns using the kernel k -means algorithm. Given the identified activities, we propose a novel nominal matrix factorization method under a Bayesian framework with Lasso to extract highly interpretable daily routines. For empirical evaluation, the proposed methodology has been compared with a number of existing methods based on both synthetic and publicly available real smart home datasets with promising results obtained. We also discuss how the proposed unsupervised methodology can be used to support exploratory behavior analysis for elderly care.

Comparing the behavioural organization of head pecking, severe feather pecking, and gentle feather pecking in domestic turkeys

Detection of behavioural temporal patterns (T-patterns) allows for insight into the organization of behaviours that is not apparent through quantitative analysis of the durations and frequencies of discrete behaviours. In this study, a T-pattern was defined as a reoccurring sequence of ≥2 behavioural event types22Behavioural event types: In this study, a behavioural event type was classified as each occurrence of a behavioural event and both the start and end of each occurrence of a behavioural state. linked by a non-random critical period of time and occurring more often than expected by chance. This study aimed to identify differences in the behavioural distribution and organization of two forms of injurious pecking, head pecking (HP) and severe feather pecking (SFP), and non-damaging gentle feather pecking (GFP) in domestic turkeys using two analytical methods T-pattern analysis and a conventional assessment of behavioural frequencies and durations. Video observations of pecking behaviour were obtained from focal samples of domestic male turkeys between 5 and 15 weeks of age. Each focal sampling observation recorded all behaviours shown by a single turkey 120 s (s) before and after the turkey performed HP, GFP, or SFP (241 s total observation length per pecking event). Pecking type had no effect on the number of T-patterns and the frequency of T-pattern occurrences. However, both T-pattern detection and the conventional analysis showed turkeys in observations containing HP tended to be more active with shorter lying durations (H = 7.36, df = 2, P = 0.025), less frequent standing (H = 5.61, df = 2, P = 0.061), and less structured T-patterns (total T-pattern durations: H = 5.996, df = 2, P = 0.050, mean durations: H = 5.34, df = 2, P = 0.068) than turkeys engaged in GFP or SFP. The analyses also showed that turkeys that performed SFP tended to show more frequent GFP than turkeys that performed HP (H = 5.79, df = 2, P = 0.055). However, discrepancies in organization of feeding and foraging behaviour between the two analytical methods highlighted the susceptibility of both methods to behavioural variation between individuals when using small sample sizes. Future research should use T-pattern detection to evaluate age, sex, and environmental differences in the organization of the injurious pecking behaviour in domestic turkeys.

Detection of Ludic Patterns in Two Triadic Motor Games and Differences in Decision Complexity.

The triad is a particular structure in which an ambivalent social relationship takes place. This work is focused on the search of behavioral regularities in the practice of motor games in triad, which is a little known field. For the detection of behavioral patterns not visible to the naked eye, we use Theme. A chasing games model was followed, with rules, and in two different structures (A↔B↔C↔A and A → B → C → A) on four class groups (two for each structure), for a total of 84, 12, and 13 year old secondary school students, 37 girls (44%) and 47 boys (56%). The aim was to examine if the players' behavior, in relation to the triad structure, matches with any ludic behavior patterns. An observational methodology was applied, with a nomothetic, punctual and multidimensional design. The intra and inter-evaluative correlation coefficients and the generalizability theory ensured the quality of the data. A mixed behavioral role system was used (four criteria and 15 categories), and the pattern detection software Theme was applied to detect temporal regularities in the order of event occurrences. The results show that time location of motor responses in triad games was not random. In the “maze” game we detected more complex ludic patterns than the “three fields” game, which might be explained by means of structural determinants such as circulation. This research points out the decisional complexity in motor games, and it confirms the differences among triads from the point of view of motor communication.

Automatic detection and labeling of self-stimulatory behavioral patterns in children with Autism Spectrum Disorder.

An infrastructure to record, detect and label the behavioral patterns of children with Autism Spectrum Disorder (ASD) has been developed. The system incorporates 2 different sensor platforms which are wearable and static. The wearable system is based on accelerometer which detects behavioral patterns of a subject, while the static sensors are microphones and cameras which captures the sounds, images and videos of the subjects within a room. The video also provides ground truth for wearable sensor data analysis. The system labels the segment of video data upon detection of the autistic behavior. That is, it stores the time of the video when the activities are detected. Time-Frequency methods are used to extract features and Hidden Markov Model (HMM) are used for analyzing the accelerometer signal. Using these methods, we are able to achieve 91.5% of classification rate for behavioral patterns studied in this paper which is used to label and save data.

Assessing machine learning classifiers for the detection of animals’ behavior using depth-based tracking

There is growing interest in the automatic detection of animals’ behaviors and body postures within the field of Animal Computer Interaction, and the benefits this could bring to animal welfare, enabling remote communication, welfare assessment, detection of behavioral patterns, interactive and adaptive systems, etc. Most of the works on animals’ behavior recognition rely on wearable sensors to gather information about the animals’ postures and movements, which are then processed using machine learning techniques. However, non-wearable mechanisms such as depth-based tracking could also make use of machine learning techniques and classifiers for the automatic detection of animals’ behavior. These systems also offer the advantage of working in set-ups in which wearable devices would be difficult to use. This paper presents a depth-based tracking system for the automatic detection of animals’ postures and body parts, as well as an exhaustive evaluation on the performance of several classification algorithms based on both a supervised and a knowledge-based approach. The evaluation of the depth-based tracking system and the different classifiers shows that the system proposed is promising for advancing the research on animals’ behavior recognition within and outside the field of Animal Computer Interaction.

Activity inference engine for real-time cognitive assistance in smart environments

Recent research in ambient intelligence allows wireless sensor networks to perceive environmental states and their changes in smart environments. An intelligent living environment could not only provide better interactions with its ambiance, inside electrical devices and everyday objects, but also offer smart services, even smart assistance to disabled or elderly people when necessary. This paper proposes a new inference engine based on the formal concept analysis to achieve activity prediction and recognition, even abnormal behavioral pattern detection for ambient-assisted living. According to occupants’ historical data, we explore useful frequent patterns to guide future prediction, recognition and detection tasks. Like the way of human reasoning, the engine could incrementally infer the most probable activity according to successive observations. Furthermore, we propose a hierarchical clustering approach to merge activities according to their semantic similarities. As an optimized knowledge discovery approach in hierarchical ambient intelligence environments, it could optimize the prediction accuracies at the earliest stages when only a few observations are available.

Detection and Selection of Behavioral Patterns Using Theme: A Concrete Example in Grassroots Soccer.

Observational methodology provides a rigorous yet flexible framework for capturing behaviors over time to allow for the performance of subsequent diachronic analyses of the data captured. Theme is a specialized software program that detects hidden temporal behavioral patterns (T-patterns) within data sets. It is increasingly being used to analyze performance in soccer and other sports. The aim of this study was to show how to select and interpret T-patterns generated by the application of three “quantitative” sort options in Theme and three “qualitative” filters established by the researchers. These will be used to investigate whether 7-a-side (F7) or 8-a-side (F8) soccer is best suited to the learning and skills development needs of 7- and 8-year-old male soccer players. The information contained in the T-patterns generated allowed us to characterize patterns of play in children in this age group. For both formats, we detected technical-tactical behaviors showing that children of this age have difficulty with first-touch actions and controlling the ball after a throw-in. We also found that ball control followed by a pass or a shot at the goal are common in the central corridor of the pitch. Further, depth of play is achieved by ball control, followed by dribbling and a pass or shot. In F8, we saw that depth of play was achieved through ball control, followed by dribbling and passing of one or more opponents leading to a pass or shot. However, in F7, we saw that players succeeded in advancing from their goal area to the rival goal area through a sequence of actions.

Detection of behavioural patterns in Olympic male taekwondo athletes

The relationship between athletes’ behaviours is relevant for developing defensive and offensive strategies. Traditionally, sequential analysis has been conducted to assess relationships between two behaviours (one considered as focal and the other as conditioned). The aim of the study was to examine Olympic taekwondo athletes’ behavioural patterns by conducting sequential analyses of tactical actions in bouts. Seventy-five male matches of the London Olympic Games in 2012 were studied to analyse tactical actions: attack (direct and indirect), counterattack (anticipatory, simultaneous and posterior), defensive (cuts, blocks and dodges), and opening actions. A sequential analysis of 2 lags (i.e., actions) in both prospective and retrospective perspectives were conducted by using five focal behaviours (i.e., those considered as tactical scoring actions: direct and indirect attack, anticipatory, simultaneous and posterior counterattacks). The results showed different tactical sequences according to the type of action considered as focal in the analysis. Specifically, direct attacks and simultaneous counterattacks were the most frequently action used in competition. Sequential analyses confirmed the interchangeable pattern of attacking and counterattacking actions. Those analyses can provide an insight for athletes’ patterns and tactical strategy training in taekwondo based on empirical data.

Extraction of geometric and prosodic features from human-gait-speech data for behavioral pattern detection: Part I

In this work, we extract prosodic features from subjects while they are talking while walking using human-gait-speech data. These human-gait-speech data are separated into 1D data (human-speech) and 2D data (human-gait) using the adaptive-lifting scheme of wavelet transform. We carry out extraction of prosodic features from human-speech data, such as speech duration, pitch, speaking rate, and speech momentum, using five different natural languages (Hindi, Bengali, Oriya, Chhattisgarhi, and English) for the detection of behavioral patterns. These behavioral patterns form real-valued measured parameters, stored in a knowledge-based model called the human-gait-speech model. Extraction of geometrical features from human-gait data, such as step length, energy or effort, walking speed, and gait momentum, is carried out for the authentication of behavioral patterns. In this paper, the data of 25 subjects of different ages, talking in five different natural languages while walking, are analyzed for the detection of behavioral patterns.

Two Strikes

There are large racial disparities in school discipline in the United States, which, for Black students, not only contribute to school failure but also can lay a path toward incarceration. Although the disparities have been well documented, the psychological mechanisms underlying them are unclear. In two experiments, we tested the hypothesis that such disparities are, in part, driven by racial stereotypes that can lead teachers to escalate their negative responses to Black students over the course of multiple interpersonal (e.g., teacher-to-student) encounters. More generally, we argue that race not only can influence how perceivers interpret a specific behavior, but also can enhance perceivers’ detection of behavioral patterns across time. Finally, we discuss the theoretical and practical benefits of employing this novel approach to stereotyping across a range of real-world settings.

From Surveillance to Assistance: Optimizing Care of Dementia Patients by Intelligent Behavioural Analysis. Introduction into Opdemiva as a Diagnosis-related Ambient Assisted Living (AALl) - System.

Introduction In the course of demographic change an increasing number of dementia patients is predicted. More elderly desire to remain longer self-determined at their familiar domestic surroundings while declining disposability of nursing staff is prognosticated. Modern AAL-systems offer the opportunity to tackle this challenge. Objectives In order to meet the needs of dementia patients, a prospective AAL-system should discern people's activities of daily living (ADLs). Regarding to their complexity, in order to generate sophisticated assistance, the system needs to integrate harmonically into the domestic environment without any interference on ongoing activities of those concerned. Aims OPDEMIVA ascertains possibilities in real-time-recognizing of ADLs applying innovative contactless optical measurement sensors with focusing on the detection of high-risk behavioural patterns. Methods An interdisciplinary research team of technical and health-care experts was involved in the mutual project. Based on examination of the list of ADLs and preconceiving both aspects of medical importance and technical feasibility, a subset of relevant behavioural patterns was identified. An appropriate technical system based on optical smart sensor technology for ADLs detection was developed and preliminarily tested. Results In preliminary tests several fundamental ADLs, e.g. movement and recreation could be detected with high reliability while possible deficits like motionlessness over a defined period became derivable. Conclusions Smart sensor based detection of dementia-related high-risk behavioural patterns requires extensive detailed examination of capabilities and needs of persons concerned. Patient's privacy and ethic concerns are of high relevance. Results gained as yet imperatively demand further research.

Extracting patterns of behavior from a network of binary sensors

According to the World Health Organization and the US Department of Health, older adults are one of the most rapidly expanding population segments of many developed countries. As a result, significant research in the area of assistive technologies for the elderly has been undertaken to help manage this growing demographic segment. Our contribution to this effort is based on leveraging environmental sensors to track occupants for extended durations in their homes in order to extract and track various behavioral patterns in order to identify atypical activities. To aid in our detection of behavioral patterns, we extract additional intermediate indicators of mobility and behavior such as estimates of an occupant’s position as well as other parameters such as guest visitations. In our initial research, we did not associate these instances of atypical behaviors with any specific health events. Our primary goal was to see if it was possible to detect atypical behaviors using only the limited data that can be extracted using passive infrared sensors. We compared the days when atypical patterns were detected with the similar analysis of a human annotated activity log and determined that there is a strong correlation between patterns extracted from our derived behavioral indicators and those from the annotated activities. We also compared our systems result’s across multiple datasets to verify that our approach could be applied to different sensor configurations and occupants.

On the elements activating the transmission of specialized knowledge in verbs

In this paper we propose an analysis of phraseological verbs in specialized discourse in order to establish syntactic and semantic characteristics likely to indicate the transmission of specialized knowledge. By adopting and adapting theoretical and descriptive proposals existing for general language, we consider elements from the relation between the verb and its arguments (such as argument structure and thematic roles) and aspects from the arguments themselves (such as semantic features and termhood). Through the detection of combined features and behaviour patterns of occurrences, we are able to more specifically establish which characteristics allow us to recognise the terminological use of phraseological verbs and which kind of information is the more relevant to its conveyance.

Detection of Communicative Behavior Patterns AMONG College Students during Lab Practice

The aim of this work is to detect the existence of communicative interaction patterns from the conversations among Biology students during lab practice. Observational methodology was used by creating a category system as an observation tool, and continuing with a process of re-categorization. A total of six sessions were observed and each one last one hour. It was used a software called SDIS-GSEQfor the inter-sessional sequential analysis. The results highlight that it is important for the students the use of question-answer (IAE) in its different modalities: in probability of occurrence, first are the categories of Persuading and Proposing, the next significant sequential probabilities are: Evaluating, Confirming, and Confusing; followed in sequential occurrence order by: Arguing, Classifying, Correcting, Clarifying, and Suggesting and lasting in sequential inhibitory occurrence are: Creating an opinion, Directing, and Evaluating. These are the particular ways in which students encourage their participation during laboratory practice, opening their minds to their classmates’ feedback, which invites them to revalue their initial questions and answers and encourages new discussion topics and increase learning opportunities.

A real-time model predictive control for building heating and cooling systems based on the occupancy behavior pattern detection and local weather forecasting

Current research studies show that building heating, cooling and ventilation energy consumption account for nearly 40% of the total building energy use in the U.S. The potential for saving energy through building control systems varies from 5% to 20% based on recent market surveys. This papers introduces and illustrates a methodology for integrated building heating and cooling control to reduce energy consumption and maintain indoor temperature set-point, based on the prediction of occupant behavior patterns and local weather conditions. Advanced machine learning methods including Adaptive Gaussian Process, Hidden Markov Model, Episode Discovery and Semi-Markov Model are modified and implemented into this study. A Nonlinear Model Predictive Control (NMPC) is designed and implemented in real-time based on dynamic programming. The experiment test-bed is setup in the Solar House, with over 100 sensor points measuring indoor environmental parameters, power consumption and ambient conditions. The experiments are carried out for two continuous months in the heating season and for a week in the cooling season. The results show that there is a 30.1% measured energy reduction in the heating season compared with the conventional scheduled temperature set-points, and 17.8% energy reduction in the cooling season.

Detection of Behavior Patterns in Single Women in Mexico

The proposal of this work is to detect the existence of patterns of interactive sequential discursive behavior in three single Mexican adult women from retrospective analysis of the interview between them and the interviewer. The applied methodology was observational. It was developed iteratively and thoroughly, a category system as an instrument of observation, and then it was undergone a recategorization process. The number of sessions observed was three, of three hours a duration. Intersessional sequential analysis was used through the SDIS-GSEQ software. Thus we find that the first significant categories in order of occurrence are: Being for her self, Paradox and Independence. The following in significant probability of occurrence position are: Dependency, self-satisfaction, Complementary, Comprehension andNormal. Finally, the categories that are inhibitory occurrence probabilistically negative are: Responsible, Myself, Self-understanding. Finding out these women are learning to live for themselves within a process of emancipation from their families and social structures.

Activity-informed Dynamic Data Driven Simulation

In previous work we developed dynamic data driven simulation (DDDS) that assimilates real time sensor data using Sequential Monte Carlo (SMC) methods. This paper builds on previous work and presents a framework that adds a real time behavior pattern detection layer on top of data assimilation for dynamic data driven simulation. The real time behavior pattern detection layer uses Hidden Markov Model (HMM) to detect the behavior patterns of a system in real time and uses the detected behavior pattern to inform the simulation model for more accurate simulation. We apply the proposed framework to a smart environment application and discuss how to recognize behavior pattern from spatial-temporal sensor data using Coupled HMM (CHMM).