A Study on Case of Ergonomic Evaluation using Data Mining Techniques

Objective: The aim of this study is to analyze case of ergonomic evaluation using data mining techniques. Background: It is important to discover the hidden trends and patterns in large data. Data mining model is composed of four fundamental categories which are classification, clustering, combination and continuity. Decision Tree is to model the relationships that exist in data and find the rules from the relationships. Self-organizing feature map (SOFM) is a type of artificial neural network that is trained using unsupervised learning to produce two-dimensional, discretized representation of the input space of the training samples. Method: It is to evaluate perceived discomfort of working postures in terms of upper body when an external load varies. Eighteen subjects participated in an experiment of appraising perceived discomfort of varying upper body postures with three levels of external loads given. Results: The ANOVA results showed that the perceived discomfort of upper body postures was significantly affected by external load and upper body posture were significant. Also, it is able to understand the most affected variable and extract the rules on relationships between external load and joint angle of human body. In addition, working postures of upper body with respect to perceived discomfort are divided into eight groups using SOFM. Conclusion: It is need to use data mining techniques to analyze ergonomic evaluation data because it is possible to apply various cases of ergonomics.