中文摘要：

基于DEA的企业自主创新能力评价，能消除企业自主创新的基础条件差异带来的影响，客观地反映企业在自主创新过程中的主观努力程度。构造企业自主创新能力的投入产出指标体系，利用模糊多粒度语言对定性指标进行量化；利用C^2R模型判断企业自主创新能力的相对有效性，利用“投影”分析非有效DMU的成因，利用改进的DEA模型确定创新能力相对有效的企业创新资源投入产出的“扩大系数”；以某地区5个开展自主创新的企业为例，利用该方法对自主创新能力进行了评价，并提出了有效和非有效DMU自主创新能力提升的策略。实践表明，该方法能有效地评价和提升自主创新能力。

英文摘要：

Independent innovation remains a vital way for enterprises to enhance their core competence. Evaluation of enterprises＇ independent innovation via DEA diminishes the influences due to difference of enterprises＇ environment, and well reflects the effort enterprises make towards independent innovation. The establishment of the input-output index system lays the foundation for the evaluation of independent innovation. Fuzzy Multi-granularity Linguistic Term Sets is used to evaluate the qualititative indexes. Traditional C2R model is employed to evaluate enterprises＇ capability of independent innovation,＂ project＂ is used to analyze the causes of ineffectiveness of DMU, and the modified DEA model is used to determine the ratio of augmentation of the resources of innovation that help enhance enterprises＇ capability of independent innovation. Evaluation of enterprises＇ independent innovation via DEA is expected to provide theoretic guide for enterprises to locate themselves effectively as well as enhancing their independent innovation capability. Finally, five independent innovation enterprises in a region are selected as a case to measure and analyze every enterprise＇ s independent innovation capability and enhancing measures are discussed for all enterprises. By an empirical analysis, the method can evaluate and enhance independent innovation capability effectively.