Open Access Repository

KEM-DT: a knowledge engineering methodology to produce an integrated rules set decision tree classifiers

Downloads

Downloads per month over past year

Ali, M ORCID: 0000-0002-4107-7122, Lee, S and Kang, BH ORCID: 0000-0003-3476-8838 2018 , 'KEM-DT: a knowledge engineering methodology to produce an integrated rules set decision tree classifiers', in Proceedings from the International Conference on Ubiquitous Information Management and Communication , pp. 1-5 , doi: 10.1145/3164541.3164640.

[img] PDF (The 12th International Conference on Ubiquitous Information Management and Communication (IMCOM 2018), Jan 5-7, 2018,Langkawi, Malaysia)
KEM-DT- A Knowl...pdf | Request a copy
Full text restricted

Abstract

In artificial intelligence, knowledge engineering is one of the key research areas in which knowledge-based systems are developed to solve the real-world problems and helps in decision making. For constructing a rule-based knowledge base, normally single decision tree classifier is used to produce If-Then rules (i.e. production rules). In the health-care domain, these machine generated rules are normally not well accepted by domain experts due to knowledge credibility issues. Keeping in view these facts, this paper proposes a knowledge engineering methodology called KEM-DT, which generates classification models of multiple decision trees, transforms them into production rules sets, and lastly, after rules verification and validation from an expert, integrates them to construct an integrated as well as a credible rule-based knowledge base. Finally, in order to realize the KEM-DT methodology, a Data-Driven Knowledge Acquisition Tool (DDKAT) is developed.

Item Type: Conference Publication
Authors/Creators:Ali, M and Lee, S and Kang, BH
Keywords: knowledge engineering, decision tree, classication model, model translation, production rule
Journal or Publication Title: Proceedings from the International Conference on Ubiquitous Information Management and Communication
DOI / ID Number: 10.1145/3164541.3164640
Copyright Information:

Copyright 2018 ACM

Related URLs:
Item Statistics: View statistics for this item

Actions (login required)

Item Control Page Item Control Page
TOP