Thematic analysis of sustainable ultra-precision machining by using text mining and unsupervised learning method

Abstract

Sustainable manufacturing is one key research area to reduce environmental damages and resource waste nowadays. As a cutting-edge manufacturing method, ultra-precision machining (UPM) plays an increasingly significant role to achieve sustainable manufacturing because of its rapidly increasing demand. The purpose of this paper is to discover and evaluate the main themes of current works about sustainable UPM. By utilizing the latent Dirichlet allocation (LDA) method to analyze the abstracts of the relevant publications, four main themes of sustainable UPM were identified. The percentage of each documents’ content contributing to these four themes was also extracted. According to the documents’ contribution data, the publications can be classified into four groups by using the K-means algorithm. It shows that the machining process is the most focused theme in this field and the majority of works about surface structure involved multiple topics. And the social aspect of sustainable UPM needs extensive investigation in the future. In this paper, the thematic analysis was conducted for the first time in the area of sustainable UPM. And the LDA-unpreserved learning approach was also proposed in this work originally. This work provides an overall map of sustainable UPM literature to help researchers select the topics which have not been discussed.