On the Context-Aware Servicing of User Needs: Extracting and Managing Context Information Supported by Rules and Predictions

The desired context-aware servicing of user needs assumes adequately capturing the user situation, which in turn is often done using sensors. In most cases, the sensor-driven extraction of context information is done counting on pre-defined rules that concern Boolean expressions directly referring to data values for the sake of evaluating the user situation. Further, sensors would be of limited use when considering context indicators (such as intentions) that are not “physical”. Inspired by those challenges, we address the training-data-driven extraction of context information, opting for considering Bayesian Modeling and particularly the Naïve Bayesian Classification Approach because it is: (i) effective as it concerns predictions that are based on training data; (ii) rarely misleading in comparatively “simple” cases, which holds for most real-life cases, as opposed to natural-science-related cases where numerous possible outcomes may apply to any situation; (iii) easily applicable in terms of hardware and software capabilities. Hence, we study the adequacy and usefulness of applying probabilistic approaches together with rules, in establishing and managing the extraction of context information that in turn is needed for the appropriate context-aware servicing of user needs.