No abstract

With the advent of break-through sensing technology, performing data capturing and analysis for knowledge engineering has become more opportunistic. The task of efficiently analyzing sensor based data for effective decision making poses a significant challenge. Conventional prediction and recommender systems lack comprehensive analysis of all parameters and aspects, thus compromising prediction results. At the decision-making level, traditional knowledge driven prediction systems deploy classical ontology for knowledge representation and analysis. However, classical ontologies are not considered as powerful tools due to their inability to handle vagueness in data for real-world applications. On the contrary, fuzzy ontology deals with the issue of hazy and uncertain data for effective analysis to give promising results. This work presents interval type 2 fuzzy ontological knowledge model that predicts water quality of sensor based water samples and providing solutions with respect to the corresponding quality state. The proposed knowledge model constitutes of two newly developed ontologies: water sensor observations ontology (crisp ontology to model sensor observational data) and water quality ontology (interval type 2 fuzzy ontology for modeling the water quality prediction process). The inference mechanism is based on interval type-2 fuzzy partitioning and computation. Besides water quality prediction and providing solutions, the proposed model handles the issue of interoperability and exchange of consensual knowledge among multiple disciplines. The proposed knowledge model is validated with real-life water sensor based parameterized data captured from various geographically dispersed monitoring stations with approximately 50,000 samples at each station.

No abstract

No abstract

No abstract