Dempster-shafer theory and linguistic intuitionistic fuzzy number-based framework for blending knowledge from knowledge repositories: An approach for knowledge management. (1st August 2022)

• Record Type: Journal Article
• Authors: Anjaria, Kushal
• Is Part Of: Expert systems with applications. Volume 199(2022)
• Publication Date: 2022-08-01
• Subjects: Knowledge management -- Knowledge blending -- Linguistic Intuitionistic fuzzy number -- Dempster-shafer evidence theory -- Weight combination; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2022.117142
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: Provides a framework to blend the knowledge from various knowledge repositories. Address the problems related to the subjective and objective weights of the knowledge repositories. Handles the issues related to unknown knowledge attribute weights. Presents the utility of the proposed approach using a case study. Compare the proposed method with the existing techniques. Abstract: Based on Dempster-Shafer Evidence Theory (DSET) and Linguistic Intuitionistic Fuzzy Numbers (LIFN), this study presents a paradigm for Knowledge Blending (KB) that supports the Knowledge Management (KM) process. The KB framework highlights and analyzes knowledge blending, the emergence of new insights, the ordering of knowledge repositories, and selected knowledge entities that are part of such a knowledge combination. The KB method enables the analysis of knowledge repositories using linguistic terms and quantitative variables and the identification of the ordering of knowledge repositories and specific entities. Ordering knowledge repositories and entities during KB is tricky since knowledge providers employ a range of adjectives, rankings, interpretations, and linguistic terms, rendering the KB process ambiguous. To address the issue, we leverage DSET to describe the LIFNs extracted from knowledge sources as Basic Probability Assignments (BPA). Then, we merge subjective and objective weights obtained from knowledge repositories. To amend evidence from knowledge repositories, we apply the combined weight. Finally, the comprehensive evaluation value of each option is quantified using the combination rule of evidence, which aids in the blending of knowledge repositories and entities acquired from the knowledge repositories. We discuss a case study to explain the proposed method and compare it to other approaches in order to demonstrate its viability and usefulness. The case study highlights the value of KM during the knowledge structuring and auditing stages of the KM process.