Uncertain Ontology Model for Knowledge Representation and Information Retrieval Using Decision Rules

EasyChair Preprint 9442

Abstract

Knowledge plays a vital role for an effective operation and decision making. Relevance of Information Retrieval (IR) depends on an efficient Knowledge Representation (KR). Ontology constitutes rich-set of knowledge formalism for KR but inconsistency, vagueness, incompleteness etc., are major limitations shows uncertainty. In this paper, we have presented Rough Bayesian (RB) approach for uncertain ontology. Under this work, we have presented decision rules to determine attributes reduction, estimate the outcomes of a set of queries and decision class for PIMA Indians Diabetes Ontology. The model identifies a group of relational rules, reduct calculations, minimal rules and utilizes it in inferences and query retrieval with the help of probabilistic BN. The model captures the ontology's knowledge as a whole, and accurately predicts the average of belief with 91 % accuracy for four queries in terms of precision, recall and accuracy and 98 % accuracy of decision class on approximation.

Keyphrases: Information Retrieval, Ontology, decision rule, knowledge representation

@booklet{EasyChair:9442,
  author    = {Sanjay Kumar Anand and Suresh Kumar},
  title     = {Uncertain Ontology Model for Knowledge Representation and Information Retrieval Using Decision Rules},
  howpublished = {EasyChair Preprint 9442},
  year      = {EasyChair, 2022}}