Paper Number

ECIS2026-1949

Paper Type

CRP

Abstract

Government organisations lack structured methods to evaluate knowledge management system designs before implementation. Existing approaches emphasise operational outcomes rather than design characteristics, limiting early detection of weaknesses. This design science research develops a conceptual framework integrating Explainable AI (XAI) and Knowledge Graphs (KG) for structured design value assessment in digital government. The framework applies five criteria across the knowledge management lifecycle: capture and codification, sharing and collaboration, accessibility and findability, renewal and updating, and citizen-centric approaches. Mathematical formulation specifies calculation procedures combining XAI explanation capabilities with KG-based knowledge representation. A proof-of-concept demonstration using constructed synthetic data illustrates computational feasibility, producing design value scores ranging from 52% to 87% across criteria. The study contributes by distinguishing design value from operational value, integrating XAI-KG technologies for transparent evaluation, and providing a replicable framework for early-stage design assessment. Future research requires empirical validation through organisational case studies in operational government contexts.

Share

COinS
 
Jun 14th, 12:00 AM

Integrating Xai And Knowledge Graphs For Design Value Assessment In Digital Government

Government organisations lack structured methods to evaluate knowledge management system designs before implementation. Existing approaches emphasise operational outcomes rather than design characteristics, limiting early detection of weaknesses. This design science research develops a conceptual framework integrating Explainable AI (XAI) and Knowledge Graphs (KG) for structured design value assessment in digital government. The framework applies five criteria across the knowledge management lifecycle: capture and codification, sharing and collaboration, accessibility and findability, renewal and updating, and citizen-centric approaches. Mathematical formulation specifies calculation procedures combining XAI explanation capabilities with KG-based knowledge representation. A proof-of-concept demonstration using constructed synthetic data illustrates computational feasibility, producing design value scores ranging from 52% to 87% across criteria. The study contributes by distinguishing design value from operational value, integrating XAI-KG technologies for transparent evaluation, and providing a replicable framework for early-stage design assessment. Future research requires empirical validation through organisational case studies in operational government contexts.