企業信用評等是一個複雜且昂貴的過程,如何有效評估企業信用評等,便成為金融市場上的重要問題,傳統上大多數現有研究僅使用財務指標來預測公司的績效或信用等級,本研究之目的是利用財務和非財務指標來探討企業信用評等的評量方法,除了傳統的財務指標之外,本研究另外加入了企業新聞的情緒分析,並利用專利探勘技術評估其創新指標,綜合這些財務及非財務指標利用機器學習分類器建立公司信用評等的分類模型。研究結果顯示,本研究提出的模型可以通過集成學習分類器有效分類企業信用等級,研究結果可以幫助企業從財務和非財務指標中識別出重要因素,從而提高其信用評級,並可協助金融市場準確進行企業信用評等。
In recent years, finance and corporate innovation has gained increasing attention and emerged as a significant subject of research from researchers and practitioners. Corporate credit rating is a complex and expensive process. However, most existing studies only use financial indicators to predict firm performance or credit ratings. The goal of this research is to examine the relationship between corporate innovation, financial news sentiment analysis, and a firm’s financial performance, as well as its credit rating structures by using financial and non-financial indicators. In this work, we propose a predictive model while utilizing machine learning classifiers to extract important features to predict corporate credit ratings. The experimental results show our proposed model can effectively predict the corporate credit rating by ensemble learning classifiers. Our research can help corporates identify the significant factors from financial and non-financial indicators to improve their credit ratings.