Human Interaction Behavior Research Articles

Using the Instance-Based Learning Paradigm to Model Energy-Relevant Occupant Behaviors in Buildings

Human interactive behavior is accountable for most of the variance between the observed and predicted energy consumption of buildings, and is accordingly acknowledged as a major field of research into limiting building-related energy consumption. A thorough understanding of occupant behavior is critical to facilitate a more reliable prediction of energy consumption and identifying means by which pro-environmental behaviors can be promoted. Insights and models from psychology and sociology appear to be best suited to improving such understanding, and this article contributes to this end by developing and testing a cognitive model that serves as the core of a numerical human-building interaction model. The proposed implementation builds on instance-based learning, a well-established cognitive modeling paradigm, is integrated into a thermodynamic building model, and complemented by perception models for the approximation of the thermal and olfactory perception of the environment. The model successfully learns to interact plausibly with a set of elements of a model room—a heating system, a window, and the actor’s clothing—in order to establish predefined room conditions. Accumulation of context-specific instances in the declarative memory, which are retrieved and blended in a decision situation, provide the model with the flexibility to adapt its actions to very different climatic contexts, represented by the locations Stuttgart, Madrid, Stockholm, and Melbourne. Moreover, the model manages to find appropriate compromises if need satisfaction requires contradictory actions, such as in situations where satisfaction of the olfactory need requires opening the window and satisfaction of the thermal need requires keeping it closed. Despite its obvious complexity, the model must be considered to be a basic model, which restricts the immediate comparability of its results to human behavior data. However, the successfully applied plausibility checks clearly indicate the value of the cognitive approach to modeling human-building interaction.

Dynamic property of consumer-based brand competitiveness (CBBC) in human interaction behavior

Purpose The purpose of this paper is to identify potential competitive relationships among brands by analyzing the dynamic clicking behavior of consumers. Design/methodology/approach Consumer sequential online click data, collected from JD.com, is used to analyze the dynamic competitive relationship between brands. It is found that the competition intensity across categories of products can differ considerably. Consumers exhibit big differences in purchasing time of durable-like goods, that is, the purchasing probability of such products changes considerably over time. The local polynomial regression model (LPRM) is used to analyze the relationship between brand competition of durable-like goods and the purchasing probability of a particular brand. Findings The statistical results of collective behaviors show that there is a 90/10 rule for the category durable-like goods, implying that ten percent of the brands account for 90 percent market share in terms of both clicking and purchasing behavior. The dynamic brand cognitive process of impulsive consumers displays an inverted V shape, while cautious consumers display a double V shaped cognitive process. The dynamic consumers’ cognition illustrates that when the brands capture a half of the click volume, the brands’ competitiveness reaches to its peak and makes no significant different from brands accounting for 100 percent of the click volume in terms of the purchasing probability. Research limitations/implications There are some limitations to the research, including the limitations imposed by the data set. One of the most serious problems in the data set is that the collected click-stream is desensitized severely, restricting the richness of the conclusions of this study. Second, the data set consists of many other consumer behavioral data, but only the consumer’s clicking behavior is analyzed in this study. Therefore, in future research, the parameters brand browsing by consumers and the time of browsing in each brand should be added as indicators of brand competitive intensity. Practical implications The authors study brand competitiveness by analyzing the relationship between the click rate and the purchase likelihood of individual brands for durable-like products. When the brand competitiveness is less than 50 percent, consumers tend to seek a variety of new brands, and their purchase likelihood is positively correlated with the brand competitiveness. Once consumers learn about a particular brand excessively among all other brands at a period of time, the purchase likelihood of its products decreases due to the thinner consumer’s short-term loyalty the brand. Till the brand competitiveness runs up to 100 percent, consumers are most likely to purchase a brand and its product. That indicates brand competitiveness maintain 50 percent of the whole market is most efficient to be profitable, and the performance of costing more to improve the brand competitiveness might make no difference. Originality/value There are many studies on brand competition, but most of these research works analyze the brand’s marketing strategy from the perspective of the company. The limitation of this research is that the data are historical and failure to reflect time-variant competition. Some researchers have studied brand competition through consumer behavior, but the shortcoming of these studies is that it does not consider sequentiality of consumer behavior as this study does. Therefore, this study contributes to the literature by using consumers’ sequential clicking behavior and expands the perspective of brand competition research from the angle of consumers. Simultaneously, this paper uses the LPRM to analyze the relationship between consumer clicking behavior and brand competition for the first time, and expands the methodology accordingly.

Hidden Markov models for human interactive behavior

In order to meet the requirements in the cooperation and competition experiments for an individual patient in clinical application, two human interactive behavior key-press models based on hidden Markov model (HMM) were proposed. To validate the cooperative and competitive models, a verification experimental task was designed and the data were collected. The correlation of the score and subjects' participation level has been used to analyze the reasonability verification. Behavior verification was conducted by comparing the statistical difference in response time for subjects between human-human and human-computer experiment. In order to verify the physiological validity of the models, we have utilized the coherence analysis to analyze the deep information of prefrontal brain area. Reasonability verification shows that the correlation coefficient for the training data and the testing data is 0.883 1 and 0.578 6 respectively based on cooperation model, and 0.813 1 and 0.617 8 respectively based on the competition model. The behavioral verification result shows that the cooperation and competition models have an accuracy of 71.43% respectively. The results of physiological validity show that the deep information of prefrontal brain area could been extracted based on the cooperation and competition models, and reveal the consistency of coherence between the double key-press cooperative and competitive experiments, respectively. Above all, the high consistency is obtained between the cooperatio/competition model and the double key-press experiment by the behavioral and physiological evaluation results. Consequently, the cooperation and competition models could be applied to clinical trials.

Human Interactive Behavior: A Bibliographic Review

As a common human behavior, interaction is everywhere in human life. With the rise of big data and human–computer interaction in the 21st century, more and more researchers from different industries and disciplines pay great attention to the human interactive behavior research. From the perspective of computer science, scholars try to use computer technology to make research more meaningful. To the best of our knowledge, this paper is the first study to investigate the potential rules of human interactive behavior in the view of computer science, based on 16 top-tier journals of human interactive behavior from Microsoft Academic Graph dataset. We put forward a topic extraction and clustering model based on word2vec to infer key topics, which can be widely used in different fields of research. We find that the growth of human interactive behavior is in an uptrend on the whole. Besides, the cooperative relationship between authors and countries/regions is closer over time. We also make the mensurable evolution analysis of topics by a statistical method. Some topics are hot all the time, while some are unpopular as time goes by. Finally, we do rankings in the field of human behavior research from a new perspective. All these findings help researchers observe potential patterns and the topic evolution in half a century, which may shed dazzling light on the exploration of human interactive behavior.

A Study of Arabic Social Media Users—Posting Behavior and Author’s Gender Prediction

Social media opens up numerous possibilities to study human interaction and collective behavior in an unprecedented scale. It opened a whole new venue for research under the name “social computing”. Researchers are interested in profiling individuals (e.g., gender, age group), groups, community, and networking. We are interested in studying the collective behavior of Arabic social media users. Most studies covering Arabic social media has focused on the sentiment analysis of, say tweets. This study, however, looks into who and when users interact with the Arabic social media. Specifically, there are two objectives of this work. First, studying the demographic posting behavior of social media users from two different perspectives: gender and educational level. Second, author profiling. Identifying author’s gender of a social media post. We use Saudi Arabia, a very prolific country when it comes to social media in general, as a backdrop for this study. The results in this study are based on mining huge amount of metadata of a popular local social media forum covering the period 2011–14 inclusive. The extracted features (normalized list of k highest scoring words, and likewise for stems) from the posts were used to train classifiers to identify the author’s gender. We used two different classifiers, Support Vector Machine (SVM) with linear kernel and 1-NN (1-nearest neighbor), and experimented with different sizes for the list of features. When the number of features (size of the features vector) is small (≤ 50) both classifiers perform equally well in identifying the author’s gender, but we risk overfitting the data. The classifiers achieved their best result when using 100 features. The 1-NN classifier delivered a better performance, achieving a balanced accuracy of 93.16% vs 87.33% for the SVM in predicting the author’s gender. And for a larger set of features, SVM delivered a better performance and more stable behavior than 1-NN, but still nowhere close to its best performance. We used t test to confirm our assessment that the difference between the performance of both classifiers is statistically significant. Based on that, we recommend using 100 features, and we get our best result using 1-NN with a balanced accuracy of 93.16%.

Appearance based pedestrians’ gender recognition by employing stacked auto encoders in deep learning

Pedestrians’ gender is a soft attribute which is useful in many areas of computer vision including human robot interaction, intelligent surveillance and human behavior analysis. Apart from its importance, pedestrians’ gender prediction is one of the challenging methodologies in image processing. In this article, a deep learning approach is presented to classify a pedestrian as a male or a female. As a pre-processing step, pedestrian parsing is performed by a deep decompositional neural network method. The outcome of this network is a binary mask that maps the pedestrian full body from the input image. The pedestrian body image is then extracted by applying the generated pedestrian mask to the input image. This pre-processed image is then supplied to the stacked sparse auto encoder with soft max classifier for prediction. The proposed network is trained and tested separately on different pedestrians’ views such as frontal views, back views and mixed views. The training is performed on PETA dataset. The experiments for testing are performed on MIT and PETA datasets (containing images other than train images). The accuracy values on MIT dataset are calculated as 82.9%, 81.8% and 82.4% on frontal, back and mixed views respectively. The mean AUC value by proposed scheme on PETA dataset is found as 91.5%±4. The performance measures and comparisons with existing works depict the robustness and applicability of proposed methodology.

Real-time gaze estimation via pupil center tracking

Abstract Automatic gaze estimation not based on commercial and expensive eye tracking hardware solutions can enable several applications in the fields of human computer interaction (HCI) and human behavior analysis. It is therefore not surprising that several related techniques and methods have been investigated in recent years. However, very few camera-based systems proposed in the literature are both real-time and robust. In this work, we propose a real-time user-calibration-free gaze estimation system that does not need person-dependent calibration, can deal with illumination changes and head pose variations, and can work with a wide range of distances from the camera. Our solution is based on a 3-D appearance-based method that processes the images from a built-in laptop camera. Real-time performance is obtained by combining head pose information with geometrical eye features to train a machine learning algorithm. Our method has been validated on a data set of images of users in natural environments, and shows promising results. The possibility of a real-time implementation, combined with the good quality of gaze tracking, make this system suitable for various HCI applications.

Pneumatic Artificial Muscle Optimal Control with Simulated Annealing

Abstract Pneumatic artificial muscles are soft actuators which resemble human skeletal muscles. They are attractive for many applications where safety, human interaction or biomimetic behavior is required. However, its use is not yet widespread because of the difficulties of working with them, such as high nonlinearities, difficult to model and control accurately. This work proposes a simple method for controlling the pneumatic muscle, which is using a PID controller tuned with a simulated annealing optimization algorithm.

Human Behavior Recognition Based On CNN

The technology of human behavior recognition is a cutting-edge topic which covers many areas-including computer vision, pattern recognition and artificial intelligence, and it have been widely used in areas such as commercial products, medical testing and military fields. According to the principle of Convolution Neural Networks (CNN), this paper presents a CNN human behavior recognition method based on small sample, which can identify a variety of human behaviors and human interaction behaviors respectively. The experimental result shows that the method can accurately identify the behavior of single-person and double-person interaction, and the recognition rate is above 92%, which proves the effectiveness of the method.

Let's do this together: Bi-Manu-Interact, a novel device for studying human haptic interactive behavior.

Our area of interest is robotic-based rehabilitation after stroke, and our goal is to help patients achieve optimal motor learning during high-intensity repetitive movement training through the assistance of robots. It is important, that the robotic assistance is adapted to the patients' abilities, thereby ensuring that the device is only supporting the patient as necessary ("assist-as-needed"). We hypothesize that natural and learning-effective human-machine interaction can be achieved by programming the robot's control, so that it emulates how a physiotherapist adaptively supports the patients' limb movement during stroke rehabilitation. This paper introduces the design of a novel interactive device Bi-Manu-Interact. This device is suited to be used as an experimental setup for the investigation of haptic human-human interaction and for collecting data to model therapists' haptic behavior. In this paper, we present mechanical and sensory specifications as well as tasks visualizations for future investigations. Results of a pilot clinical evaluation of the Bi-Manu-Interact with nine stroke patients are also presented in this work.

A New Framework of Human Interaction Recognition Based on Multiple Stage Probability Fusion

Visual-based human interactive behavior recognition is a challenging research topic in computer vision. There exist some important problems in the current interaction recognition algorithms, such as very complex feature representation and inaccurate feature extraction induced by wrong human body segmentation. In order to solve these problems, a novel human interaction recognition method based on multiple stage probability fusion is proposed in this paper. According to the human body’s contact in interaction as a cut-off point, the process of the interaction can be divided into three stages: start stage, execution stage and end stage. Two persons’ motions are respectively extracted and recognizes in the start stage and the finish stage when there is no contact between those persons. The two persons’ motion is extracted as a whole and recognized in the execution stage. In the recognition process, the final recognition results are obtained by the weighted fusing these probabilities in different stages. The proposed method not only simplifies the extraction and representation of features, but also avoids the wrong feature extraction caused by occlusion. Experiment results on the UT-interaction dataset demonstrated that the proposed method results in a better performance than other recent interaction recognition methods.

Welding torch attitude-based study of human welder interactive behavior with weld pool in GTAW

Skilled human welder adjusts welding parameters including the welding torch attitude, moving speed and position to control weld quality and avoid weld defects based on real time observed varying weld pool surface for precision joining using gas tungsten arc welding (GTAW). However, welder behavior/adjustment appears to be a complex reactive response. To understand this complex response and develop intelligent robotic arc welding, a new scheme that correlated welding torch attitude to weld pool surface to study welder behavior in torch adjustment as response to weld pool was proposed and realized. The torch attitude and the 3D weld pool surface were synchronously measured using a wire inertial measuring unit (IMU) and a laser vision-based approach. An image processing algorithm was developed to extract the characteristic parameters of the weld pool surface from the laser stripes reflected by the specular pool surface. The improved quaternion-based unscented Kalman filter was used to estimate the torch orientation from its inertial measurement data, showing that the torch attitude has been obtained with an acceptable error in the order of 1°(x axis and y axis) and 2°(z axis). Several experiments were performed and the correlation of the corresponding data was detailed. It indicates that the change of torch attitude represents the welder's operating skills, welding experience and smart decision. The three characteristic parameters reflect the welder's reactive response on the torch adjustments. The curvature radius of laser stripes can predict the changing trends of the weld pool surface, providing the needed information for welders to make a smart decision. The proposed scheme is feasible for measuring and analyzing the welder's skills and experience.

K Reasons Why Ergonomics Cannot Make Interactive Devices to Be User-Friendly (k>3)

Interactive interfaces are encroaching on ever more devices in our daily life activities. They appear on set top boxes, TV sets, hand-held telephones, washing machines, kitchen ovens, home thermostats, the car navigation unit and parking meters, to mention only a few. Despite the fact that practically all of them feature a menu driven interface that has been around since 1995, their usability has not improved and with the continuous increase there is even a tendency to become more inscrutable, rather than less. At the same time human factors engineering and ergonomics have become serious disciplines, widely endorsed and the associated expertise is easily available and accessible. Apparently, all scientific knowledge about human interactive behaviour does not find its way in creating transparent interfaces in many products for private use and public systems. It is argued here that this situation is caused by a number of reasons, that that are almost impossible to eliminate. In this paper five reasons are discussed, some of which are related, such that with another definition fewer reasons might emerge, but three reasons seem to be the minimum. The reasons originate from the software life cycle, cognitive models and on beliefs of the stakeholders. Acknowledging that in the current industrial and economic context the current generation of interactive interfaces will necessarily have a problematic usability, some measures for improvement are mentioned, e.g. regulation or standardization, like the familiar ones existing for electrical, radiation and medical safety.

Cave use and palaeoecology at Maludong (Red Deer Cave), Yunnan, China

Maludong is a Terminal Pleistocene fossil-bearing cave located on the northern edge of the Southeast Asian tropical zone, southeastern Yunnan Province. Hominins from the cave include remains with affinities to archaic hominins and others with an apparent mixture of archaic and modern traits all deriving from deposits dating from the Bølling-Allerød interstadial. The sedimentary sequence of the cave appears to be largely anthropogenic in origin and records a nearly continuous record of fire lasting close to 1000years. The fauna comprise only extant taxa and point to a rather biased sample with a preponderance of artiodactyls and carnivores, many of which show evidence for anthropogenic breakage and burning. A new analysis of the mammal fauna recovered during excavation, palaeohabitat reconstruction and stable isotope analysis of deer teeth and bones indicates the cave was located within or close to a closed forest environment. The mammal taxa also indicate a large body of water existed in the vicinity of Maludong, suggesting the modern lakes Datun Hai and Chang Qiao Hai were much larger at the time. Maludong may document an entirely novel ecological and behavioural scenario involving archaic and modern human interaction, economic exploitation and ceremonial behavior involving secondary burial practices.

‘It looks like a human!’ The interrelation of social presence, interaction and agency ascription: a case study about the effects of an android robot on social agency ascription

This paper will focus on the establishment of a social event that refers in our case study to a social android robot simulating interaction with human users. Within an open-field study setting, we observed a shifting of interactive behavior, presence setting and agency ascription by the human users toward the robot, operating in parallel with the robots' different modulations of activity. The activity modes of the robot range from a simple idling state, to a reactive facetrack condition and finally an interactive teleoperated condition. By relating the three types of activity modes with three types of (increasing social) presence settings--ranging from co-location, to co-presence and finally to social presence--we observed modulations in the human users' interactive behavior as well as modulations of agency ascription toward the robot. The observations of the behavioral shift in the human users toward the robot lead us to the assumption that a fortification of agency ascription toward the robot goes hand in hand with the increase in its social activity modes as well as its display of social presence features.

Games Played under Fuzzy Constraints

Psychological experiment studies reveal that human interaction behaviors are often not the same as what game theory predicts. One of important reasons is that they did not put relevant constraints into consideration when the players choose their best strategies. However, in real life, games are often played in certain contexts where players are constrained by their capabilities, law, culture, custom, and so on. For example, if someone wants to drive a car, he/she has to have a driving license. Therefore, when a human player of a game chooses a strategy, he/she should consider not only the material payoff or monetary reward from taking his/her best strategy and others' best responses but also how feasible to take the strategy in that context where the game is played. To solve such a game, this paper establishes a model of fuzzily constrained games and introduces a solution concept of constrained equilibrium for the games of this kind. Our model is consistent with psychological experiment results of ultimatum games. We also discuss what will happen if Prisoner's Dilemma and Stag Hunt are played under fuzzy constraints. In general, after putting constraints into account, our model can reflect well the human behaviors of fairness, altruism, self-interest, and so on, and thus can predict the outcomes of some games more accurate than conventional game theory.

XML을 이용한 3D 가상 제품의 HMI 행동양태 모델링과 시뮬레이션 방안

Received 2 December 2014; received in revised form 8 January 2015; accepted 16 January 2015ABSTRACTIn the virtual prototyping (VP) of digital products, it is important to provide the people involvedin product development with the visualization and interaction of the products, and the simula-tion of their human machine interaction (HMI) behaviors in interactive 3D virtual environ-ments. Especially, for the HMI behavior simulation, it is necessary to represent them properlyand to play them back effectively according to user interaction in the virtual environments. In aconventional approach to HMI behavior simulation, user interface (UI) designers use UI designsoftware tools to generate the HMI behavior of a digital product of interest. Due to lack ofreusability of the HMI behavior, VP developers need to analyze and integrate it into a VP sys-tem for its simulation in a 3D virtual environment. As this approach hinders the effective com-munication between the UI designers and the VP developers, it is easy to create errors andthereby it takes significant time and effort especially when it is required to represent the HMIbehavior to the finest level of detail. In order to overcome the shortcomings of the conven-tional approach, we propose an approach for representing the HMI behavior of a digital prod-uct using XML (eXtensible Markup Language) and for reusing it to perform the HMI behaviorsimulation in 3D virtual environments. Based on the approach, a VP system has been devel-oped and applied for the design evaluation of various products. A case study about the designevaluation is given to show the usefulness of the proposed approach.Key Words: 3D Virtual Prototypes, HMI Behavior Modeling, HMI Behavior Simulation, VirtualPrototyping, XML

Learning interaction protocols by mimicking understanding and reproducing human interactive behavior

Understanding human interactive behavior is a key technology required for future robots. To achieve this goal, the robot should be able to recognize key patterns in human–human interactions. Moreover, the robot should be able to generate similar behaviors during its interaction with human partners. In this paper, an unsupervised system is proposed that allows the robot to build a generative model of the interaction protocol using interaction records. A system is evaluated in a guided navigation task and is shown to successfully learn the underlying interaction protocol.

Virtual city: An individual-based digital environment for human mobility and interactive behavior

Increasing computing capability and high-resolution digital tracing of human behavior make large-scale computational models for individual-based realistic simulation available. Reconstructing a virtual computational environment is crucial for designing and implementing individual interactions in an artificial society as human beings behave in the real world. In this paper, we propose a methodology to recreate a virtual city by utilizing statistical data and geographic information. The synthetic population and physical environment are baseline components of the virtual city. Individual-based modeling is used to specify individuals’ demographic characteristics, and each individual is endowed with heterogeneous social attributes. Various physical environments are generated with geographic locations and mapped with individuals to support daily mobility, migration, and interaction. A series of algorithms are proposed to bridge the gap between macroscopic data and microscopic models, and guarantee equivalence between them. Based on the methodology, we reconstructed a virtual city of Beijing, and presented the statistical analysis of population structure, spatial distribution of physical environments, human travel characteristics, and spatial topologies of social networks. Our synthetic population can represent individual actors in the form of households and household members, and the synthetic population is statistically equivalent to a real population. The proposed methodology is efficient to recreate a synthetic virtual city and can serve as a base for computational experiments.

A Continuous Re-Authentication Approach Using Ensemble Learning

Insider threat becomes a more and more serious problem to organizations. Identify theft is a common attack method among various attack methods from insider. And it is very hard to detection since the attacker acts as a legal identity. Existing researches focus on Human Computer Interaction (HCI) behavior such as keystroke dynamics or mouse dynamics to detection this kind of attack. Based on an obvious observable that different applications show different HCI behavioral patterns (rich-key-operations or rich-mice-operations), we demonstrate that single authentication method is not efficient always in full time work period. In this paper, we provide an ensemble re-authentication approach to detection identify theft in real time, which combine the classification of keystroke-classifier and mice-classifier to determine whether the current operations is produced by the real legitimate user or not. Our experiment proves this new approach is efficient and can provide consistent detection ability with high accuracy.