Challenges for Migration of Rule-Based Reasoning Engine to a Mobile Platform

Abstract

Abstract. Research in the area of context-awareness has recently been revolutionized by the rapid development of mobile devices like smart phones and tablets, which became omnipresent in daily human life. Such devices are valuable sources of information about their user location, physical and social activity, profiles and habits, etc. However, the information that can be obtained is not limited to the hardware sensors that the device is equipped with, but can be extended to every sensor that is available in a communication range of a device. Although the concept of multiple sensors and devices, exchanging information and working together as one big pervasive system is not new, there is still a lot of research that has to be done to allow building such systems efficiently. In this paper the prototype of a rule-based inference engine for mobile devices is described and evaluated. The most important challenges connected with migration from desktop to mobile environment were defined, and a comparison of Prolog-based platforms, as a portable environments for mobile context-aware systems were presented. We consider implementation using a portable Prolog compiler on Android platform.

HEARTDROID—Rule engine for mobile and context‐aware expert systems

“…Additionally, it supports modelling of temporal dependencies between contexts to test weather several events occurred within predefined time span. However, the reasoning engine is Prolog, and the context acquisition layer is based on ContextToolkit, which both are very inefficient on mobile platforms (Bobek, Nalepa, & Ślażyński, ). Furthermore, the solution does not allow for uncertainty handling.…”

“…Every time a new model is loaded in the working memory module, it is analysed to retrieve point in time up to which rules in such model refer. In order to save memory, which is of great value in mobile systems (Bobek et al, ), only the states up to this point are stored. At the beginning, the state registry is empty which may lead to errors in rules that includes referrals to the non‐existing data.…”