Elsevier

Knowledge-Based Systems

Volume 19, Issue 6, October 2006, Pages 396-403
Knowledge-Based Systems

Integrating fuzzy data mining and fuzzy artificial neural networks for discovering implicit knowledge

https://doi.org/10.1016/j.knosys.2006.04.003Get rights and content

Abstract

This study proposes a knowledge discovery model that integrates the modification of the fuzzy transaction data-mining algorithm (MFTDA) and the Adaptive-Network-Based Fuzzy Inference Systems (ANFIS) for discovering implicit knowledge in the fuzzy database more efficiently and presenting it more concisely. A prototype was built for testing the feasibility of the model. The testing data are from a company’s human resource management department. The results indicated that the generated rules (knowledge) are useful in supporting the company to predict its employees’ future performance and then assign proper persons for appropriate positions and projects. Furthermore, the convergence of ANFIS in the model was proven to be more efficient than a generic fuzzy artificial neural network.

Introduction

Recent advancements in database, artificial intelligence, and computer technologies have given rise to the analysis of data, extraction of knowledge (data mining), and understanding of extracted knowledge. Many computational methods developed for these purposes involve a form of pattern recognition or classification. Often, these methods make use of classical solutions to these problems, and are therefore limited in their applications [1]. Inaccuracies and the redundancy of rules are two of the major limitations. Several studies point out that for generic artificial neural networks (ANNs), providing high-quality training input data is not as simple as “cleaning” inaccuracies from data or eliminating [2]. Generally, it requires a significant length of time for the convergence of generic ANN. On the other hand, the conventional data-mining algorithms tend to make a redundancy in the rules. For solving the above problems, this study proposes a knowledge discovery model that integrates the modification of the fuzzy transaction data-mining algorithm (MFTDA) and the adaptive-network-based fuzzy inference systems (ANFIS). In practice, this combination of methods is more accurate and comprehensive [3]. From the fuzzy database, the MFTDA is used to mine the association rules that are implicit knowledge from the fuzzy database and can be used for the training of the ANFIS. A prototype was built for testing the feasibility of the model. The testing data were from a company’s human resource management department. The results indicated that the generated rules (knowledge) stored in a knowledge base are useful in supporting the company to predict its employees’ future performances and then select proper persons for appropriate positions and projects. Furthermore, the convergence of ANFIS in the model was proven to be more efficient than a generic fuzzy artificial neural network.

The rest of this paper is organized as follows. Section 2 explores some necessary background information (literature review) while the rules generating process (the knowledge discovery model) is presented in Section 3. Section 4 uses a prototype for testing the feasibility of the knowledge discovery model and Section 5 details our conclusions.

Section snippets

Data mining

Data mining is known as “knowledge discovery in databases” [4]. It is the process of discovering interesting patterns in databases that are meaningful in decision-making [5]. Thus, data mining is a discipline of growing interest and importance. It is also an area of application that has the potential to provide significant competitive advantage to an organization [6]. Data mining is also a particular kind of machine learning (learning from examples). Its process must be automatic or

The knowledge discovery model

The knowledge discovery model is shown in Fig. 2. The major components of the model include a knowledge discovery interface, a fuzzy engine, a rule generator, a fuzzy database, a fuzzy rule base (a knowledge base), and an ANFIS. The details of the components are as follows.

A Prototype with an illustrative example

A prototype of the knowledge discovery model in human resource management is developed for the testing of feasibility. The simplified data of an international trading company, company A, are adopted for the prototype. Company A with primary computerization uses basic information systems to process the operational transaction data and payroll data. The major tasks of the human resource management department in company A include offering on-job training courses, recruiting, assigning employees to

Conclusion

The knowledge discovery model using MFTDA could automatically find specific association rules that were implicit knowledge in the fuzzy database and could be used for the training of the ANFIS. After the ANFIS has been trained, the generated rules and the trained ANFIS are deployed for supporting a company in predicting its employees’ future performances and then selecting proper persons for appropriate positions and projects. The implication of the contributions of the knowledge discovery

References (20)

There are more references available in the full text version of this article.

Cited by (61)

  • Adaptive Nonstationary Fuzzy Neural Network

    2024, Knowledge-Based Systems
  • Soft Computing Framework for Digital Healthcare System

    2023, Lecture Notes on Data Engineering and Communications Technologies
View all citing articles on Scopus
View full text