Human Resources Industrial

Placement Problem in an Industrial Environment

Shamshul Bahar Yaakob and Junzo Watada

Graduate School of Information, Production and Systems, Waseda University, 2-7 Hibikino,

Wakamatsu-ku, Kitakyushu-shi, Fukuoka 808-0135 Japan

********@****.******.**, ******@***.***.**.**

Abstract. A problem of workers evaluation and placement in an industrial en-

vironment is studied in this paper; an effect of workers relationship on their

placement is newly included in this paper. Evaluating workers suitability is

important when decision makers select proper candidates under various evalua-

tion criteria among available human resources and jobs. For problems of this

type, an analysis using a fuzzy number approach promises to be potentially

effective. In order to make a more convincing and accurate decision, the rela-

tionship between jobs is included in the workers assignment in an industrial

environment. The fuzzy suitability evaluation is performed by means of aggre-

gating the decision makers fuzzy assessment. Examples of typical application

are also presented: the results demonstrate that the workers relationship is an

important factor and our method is effective for the decision making process.

Keywords: Fuzzy sets, Decision making, Workers relationship, Workers

placement.

1 Introduction

The evaluation of workers is important for decision makers (DMs) to select proper

workers under various evaluation criteria in an industrial environment [12], [3]. The

aim of this research is to help the DMs make more effective selections from optional

candidates [10]. The workers' placement is concerned with seeking the optimal

matching between the workers and jobs within the constraints of available human

resources and jobs [5], [4]. In non-'fuzzy' conventional approaches for workers

placement approaches, the evaluation of workers suitability tends to use exact values.

Kim et al discussed it from personal network [14], [15]. However, due to the vague-

ness of job demands as well as the complexity of human attributes, the exact evalua-

tion of workers' suitability is quite difficult. The fuzzy theory developed by Zadeh [7],

[8] and the concept of fuzzy numbers for example Dubois and Prade [1] can be ap-

plied to improve the assessments and the expressions for the assessment results in an

industrial environment. Liang and Wang [3] and Kim et al [16] applied the concepts

of combining the fuzzy set theory and weighted complete bipartite graphs to develop

a polynomial time algorithm for solving personnel placement in a fuzzy environment.

In an industrial environment, an evaluation of workers relationship, i.e., a group

evaluation is also important as well as individual evaluation. In this paper, we develop

a new method in which the workers relationship is included to determine the optimal

I. Lovrek, R.J. Howlett, and L.C. Jain (Eds.): KES 2008, Part III, LNAI 5179, pp. 111 118, 2008.

Springer-Verlag Berlin Heidelberg 2008

112 S.B. Yaakob and J. Watada

workers placement. Triangular fuzzy numbers [3] are used to describe the suitability

of workers and the approximate reasoning of linguistic values [8], [9]. The operations

of fuzzy addition, subtraction and multiplication derived based on the extension prin-

ciple [11] are used to implement our algorithm.

In the following sections, triangular fuzzy numbers are briefly reintroduced, and

the inclusion of the relationship among the workers is proposed and discussed. Typi-

cal examples are also presented in order to demonstrate the effectiveness of our

proposal.

2 Fuzzy Numbers and Linguistic Variables

The aim of the fuzzy set theory is to deal with problems which have a source of

vagueness. The membership function fz(x) of fuzzy set Z represents the degree of

membership or the grade of x in the fuzzy set Z. The larger fz(x) is, the stronger the

belonging degree of x in Z. Under a fuzzy environment fuzzy numbers are useful in

promoting the representation and the information processing. A fuzzy number z in

(real line) is triangular, if its membership function

fz: [0, 1] is defined as follows:

(x - a)/(b - a), a

Contact this candidate