An Efficient Approach for Realizing RCCAR to Resolve Context Conflicts in Context-Aware Systems Using Influential Context Elements

The first step for determining the quality of context (QoC) in context aware systems (CASs) is to make sure that there are no conflicts among the values of each context element collected from different resources/sensors. This work is an extension of our approach RCCAR (Resolving Context Conflicts Using Association Rules) which proposes a solution to resolve context conflicts by exploiting the previous context to predict the valid values between conflicted ones using Association Rules (AR) technique. The contributions of this paper are twofold: firstly, an algorithm is proposed for implementing the RCCAR approach; secondly, a novel solution that improves the efficiency of RCCAR is proposed. This solution uses the decision tree technique to compute the most influential context elements before applying RCCAR. A number of experiments were conducted using the Weka tool. Results show that the enhanced RCCAR is more efficient than the original RCCAR.