There is a high interest in context-aware applications that intelligently support user tasks by acting autonomously on behalf of users. Behavior of context-aware applications depends not only on their internal state and user interactions but also on the context sensed during their execution. Such context is of a variety of types and includes human activities.

While some models of context information exist, many research issues related to context information modeling are still not fully addressed. Existing context models vary in types of context information they can represent. While some models take the user’s current situation (e.g. “in a meeting”) into account, others consider the physical environment, e.g., locations. A more generic approach to context modeling is needed in order to capture various features of context information including a variety of types of context information, dependencies between context information, quality of context information and context histories.

Besides modeling aspects, several research problems fall in the area of context recognition (i.e., automatically detecting high-level context information from raw sensor data). Among the huge amount of types of context, human activities play a major role. Indeed, human activity recognition (HAR) is a hot topic that the community has been tackling for more than a decade. Open research problems in HAR are mainly related to real-world deployments, e.g., continual learning, labeled data scarcity problem, privacy issues, or activity discovery. While our workshop considers HAR as one of the most important contexts, contributions that tackle other context-awareness research problems are clearly welcome as well.

In addition, to ease software engineering problems encountered in programming context-aware applications, appropriate abstractions are necessary to support discovery and reuse of context information as well as scalable methods of context processing and management.

This workshop’s aim is to advance the state of the art in context modeling, recognition, and reasoning and also discuss fundamental issues in context processing and management. The goal is to identify concepts, theories and methods applicable to context modeling and context reasoning as well as system-oriented issues related to the design and implementation of context-aware systems. Particular attention will be paid to hybrid approaches for context modeling and reasoning. For instance, the combination of machine learning and knowledge-based methods for context reasoning (e.g., activity recognition) is a recent and promising research direction that still needs to be investigated. In particular, the following topics are of interest to this workshop:

• Context modeling techniques and domain-specific context models

• Ontology-based approaches to context modeling and reasoning and

• Ontologies of activities and context

• Hybrid context models and advanced issues in context modeling, including issues of information quality, ambiguity, and provenance

• Context reasoning algorithms, their complexity and accuracy

• Discovery, reuse, privacy, security and trust of context information

• Distributed and scalable context management

• Tool support for context modeling and development of context model-based applications

• High level activity recognition from sensor data

• Machine Learning and Computer Vision for Activity Recognition and Context Reasoning

• Reference Datasets and Benchmarks for Activity Recognition and Context Reasoning

• Innovative context-aware applications

• Context modelling, reasoning, and management for Internet of Things

• Research in the area of Human Activity Recognition

Each accepted paper requires a full PerCom registration! No registration is available for workshops only.

Workshop papers will be included and indexed in the IEEE digital libraries (Xplore).