Context‐aware movement analytics: implications, taxonomy, and design framework

Abstract

Movement of an entity is greatly affected by its internal and external contexts. Such consequential influence has created new paradigms for context-aware movement data mining and analysis. The significance of incorporating contextual information and movement data is becoming quite evident because of the growing interest in context-aware movement analysis. Despite such importance, there is limited consensus among researchers on the definition of context and context-aware system design in movement studies. Therefore, this paper comprehensively reviews current concepts of context and provides a definition and a taxonomy for context in movement analysis. The paper proceeds by providing a definition of context-aware systems in the movement area after a complete review and comparison of the present definitions present in the literature. Inspired by related works, the paper further suggests a holistic three-layer design framework tailored to context-aware systems in movement studies to examine in greater depth the techniques applied during the design stages. The paper outlines the challenges and emergent issues in future research directions in context-aware movement analysis. The present study is an attempt to bridge the gap between solely using context and developing context-aware systems, thus paving the way for further research in movement applications. © 2017 Wiley Periodicals, Inc. How to cite this article:WIREs Data Mining Knowl Discov 2018Discov , 8:e1233. doi: 10.1002Discov /widm.1233 INTRODUCTIONM ovement is intrinsically a continuous phenomenon and is normally recorded as discrete snapshots at different temporal resolutions. The three fundamental sets pertinent to movement are space S (i.e., set of locations/places), time T (i.e., set of instants/intervals), and object O (i.e., set of entities). 1 In this paper, by movement we mean the change in the spatial location of one or more entities (e.g., people, animals, and vehicles) over time, not a change in entities' geometries (e.g., an individual's body parts). Such changes can be represented as a function μ: T ! S, which maps any time to a location in space. The purpose of analyzing movement data in the majority of research fields (e.g., geography, ecology, sociology, and transportation) has focused on the development of insights into the behavior of moving entities. However, the movement behavior of entities is generally associated with their internal states and external factors. 2 In other words, understanding entities' movement greatly depends on recognizing the characteristics attributed to the entity itself (e.g., the entities' purposes and intentions for movement, the entities' perceptions of space) as well as the discovery of geographical circumstances during the movement (e.g., weather conditions and traffic). We, respectively, term these influential variables internal and external contexts in this paper. of 19Because of the progress in sensing, imaging, tracking, and navigation technologies and infrastructures in recent years, unprecedented amount...

Knowledge Graphs for Analyzing Thermal Comfort: A Proposal

“…Here, a number of problems that arises while acquiring data from highly sensitive environments need to be addressed [5,8,11]: errors during the measurement of context parameters, data loss, speed, size and meaning [5,11,13].…”

“…Traditionally, data that is needed to make an informed decision does not come from a single source, but from several potentially heterogeneous data sources. This heterogeneity relates to differences among the sources with respect to formats, amount of data issued, acquisition speed, meaning, among others [5,9,11,13]. In the case of the analysis of thermal comfort, it is possible to identify the need for physical and physiological data that comes naturally from heterogeneous sources [11,17,18].…”

“…Physical Data and its Sources. These are the result of the measurement of environmental aspects, for example temperature, humidity, air quality, airspeed, intensity of light, noise level, among others [9][10][11]13]. These can be obtained through sensors such as: a) DHT11, for measuring temperature and humidity, b) MQ-135, for measuring air quality, c) KY-038, for measuring noise, and d) LDR, for measuring slight intensity.…”

“…Physiological Data and its Sources. These are the result of measuring aspects such as heart rate, or brain activity, among other parameters [9][10][11]13]. These data is usually obtained using devices with integrating sensors, such as smartwatches, smartphones, headbands, among others.…”

“…It has been recognized that several physical (temperature, humidity, air quality) and physiological factors (heart rate, stress) influence the thermal comfort of an individual. The physical factors refer to environmental features, while the physiological are people characteristics [9,11,12]. The factors described above can be found in the ANSI/ASHRAE 55-2017 standard [9,13] and some recommendations of the World Health Organization (WHO) [10], and have been traditionally used for the design of spaces in static contexts.…”