BICA 2017 - 1st International Workshop on Behavioral Implications of Contextual Analytics (co-located with IEEE PerCom 2017)

A combination of multi-modality sensors, powerful processors, and faster memory access have enabled both collection and processing of large amount of information on mobile phones. The available information is used to infer the current contextual state of the individual. In recent years, tremendous amount of progress has been made in the area of detecting specific contexts, however, significant challenges remain in terms of context modeling, context computation, and implication of current or historical contexts on both individual and app behavior.
Context modeling requires advances in efficient representation of the semantic dependence between contexts. Context computation requires access to large amounts of personal data that users are often unwilling to share with cloud-based services for reasons of privacy. Furthermore, the context information itself is dynamic, and its relevance heavily dependent on the timeliness of the computation process. Thus, near real-time computing techniques, that can be realized on the phone itself (without the data leaving the phone) are required. Finally, context information is consumed by apps to customize their behavior for providing better human experience. Therefore, there is a significant implication of the context on app behavior and indirectly on user behavior, including user privacy concerns, that needs careful exploration.
As part of this workshop, we will consider original and unpublished research articles that take a holistic approach towards addressing the above challenges. Today, one can leverage, recent advances in machine learning, especially the ability to execute deep learning models on phones, sophisticated data privacy mechanisms, and computationally efficient graphical models for context determination in ways that were not feasible earlier.
Topics of interest include (but are not limited to):
context computation from hard and soft sensors on mobile phones
context modeling using graphical and ontological models
use of machine learning, deep learning techniques for context computation on phones
bootstrapping context computation on phones (incremental learning from user data)
system challenges in computing contexts
privacy and security implications of sharing and computing contexts
effect of contexts on app-behavior and indirectly on users