A study on support method of consulting service using text mining-Application to real problem-

摘要

This study was conducted to produce a supporting method for consulting service companies so that the companies can respond to client demands irrespective of the consultant expertise. Occurrence of future problems in client companies is predicted using text mining with data taken from a consulting company. To do so, correspondence analysis and Data Envelopment Analysis (DEA) discriminant analysis are used. As described herein, the proposed method is improved using Intermediate Distribution Frame (IDF) values and standardization for the application in a real problem. For computer experiments to verify the effectiveness of the proposed method, prediction is made about cancellation problems. The prediction is compared to predictions produced by the consultant.