[目的/意義]針對傳統企業競爭對手評價研究中數據源單一和評價指標片面性的不足,文章率先提出了基于多源信息融合的競爭對手評價方法。[方法/過程]基于顧客價值領先戰略,從企業和消費者的視角出發,選取企業財務報表和電商平臺上的消費者評論作為信息源。通過構建財務特征和情感特征,依托BP神經網絡,分別建立基于財務特征和綜合特征的競爭對手評價模型。[結果/結論]采用仿真實驗對提出的方法進行驗證,證明基于綜合特征建立的模型相比基于單一財務特征的模型更有效。該方法為大數據背景下的企業競爭對手評價提供了一種新的研究思路。

[Purpose/significance] Due to the singleness of data source and inadequacy of evaluation indexes in traditional enterprise competitor evaluation research,this paper firstly puts forward the competitor evaluation method based on multi-source information fusion.[Method/process] Based on the strategy of customer value leadership,the financial statements and consumer reviews on the e-commerce platform are used as information sources from the perspective of enterprises and consumers.Then a comprehensive feature system is built by integrating financial features and emotional features.Subsequently,competitor evaluation models with financial features and comprehensive features are built based on BPNN respectively.[Result/conclusion] The proposed method is validated by simulation experiments,which indicates that the model with comprehensive features outperforms the model with pure financial features.The proposed approach provides a new research idea for the evaluation of enterprise competitors in big data environment.