CRIXP: Cultural Knowledge Representation for Inherent eXPlainability

Read full text Summarise with VIRAT
Authors: Abhisek Sharma, Sarika Jain
Cited by: 0
References: 26

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Cultural awareness plays a vital role in making AI systems beneficial for a globalized audience. This cultural awareness is lacking in modern decision-support systems and in the current state of the work, where explainability is the paramount requirement now from all AI systems. To address this, the article introduces CRIXP, a framework with decisive cultural knowledge and the capability to explain the reasoning behind its suggestions. CRIXP comprises: 1) a cultural ontology (CRO); 2) a cultural knowledge graph (CKG); 3) a general-purpose adaptable algorithm providing recommendations and explanations; and 4) a CKG enrichment module. The cultural traits of all the locales are consolidated and represented as an ontology and a knowledge graph. A cultural shift suggestion system has been demonstrated for recommendations and explanations. The response generated is compared with ChatGPT. CRIXP is found to outperform ChatGPT by providing more relevant recommendations with inherent explainability.

Fields of Study & Subjects

Explainable Artificial Intelligence (XAI) Semantic Web and Ontologies Ethics and Social Impacts of AI

Publication details

Journal / source
IEEE Transactions on Computational Social Systems
Volume
13
Issue
3
Pages
2779–2788
Publication date
30 March 2026
Language
EN