A hybrid data mining model in analyzing corporate social responsibility

作者：Pai Ping Feng^*; Chen Lei Chun; Lin Kuo Ping

来源：Neural Computing & Applications, 2016, 27(3): 749-760.

DOI：10.1007/s00521-015-1893-0

• Support vector machines
• Classification
• Rule generation
• Corporate social responsibility

Over the past two decades, corporate social responsibility (CSR) has received worldwide attention. Publication of CSR reports has become the trend for domestic and foreign enterprises. In the constantly changing and competitive corporate environment, public attention has come to be focused on how enterprises play the role of corporate citizen, and how they achieve a balance of profitable, environmental and charitable activities. However, most quantitative CSR studies to date have concentrated on traditional statistical approaches. The data mining technique has not been widely explored in this area. Thus, this investigation proposes a hybrid data mining CSFSC model, which stands for the first letters of CFS, SMOTE, FCM, SVMOAO and C5.0, integrating data-preprocessing approaches, a classification method and a rule generation mechanism for analyzing CSR data. The data-preprocessing approaches include correlation-based feature selection (CFS), the synthetic minority over-sampling technique (SMOTE) and the fuzzy c-means (FCM) clustering algorithm. The support vector machine one-against-one (SVMOAO) method was employed as a classifier for performing multiclassification, and the C5.0 decision tree algorithm was utilized to generate rules from the results of the SVMOAO model. In this study, CSR data collected from China's listed firms in 2010 were used to test the performance of the proposed model. The empirical results showed that the designed CSFSC model yields satisfactory classification accuracy, and can provide rules for decision makers. Therefore, the presented CSFSC model is a feasible and effective alternative in analyzing CSR data.

• 出版日期2016-4