目的克服相合系数法或相关系数法在化橘红指纹图谱分析中的误判问题。方法提出了信息熵确定权重的相似度算法。结果该算法明显提高了对不同产地的23批次化橘红药材样品的识别有效度。结论本法为化橘红品种的快速识别提供了科学的参考。 Objective To overcome the problem of misjudgment in the analysis of orange-red fingerprints by the method of coincidence coefficient or correlation coefficient. The method proposed a similarity algorithm to determine the weight of information entropy. Results The algorithm significantly improves the efficiency of identification of 23 samples of tangerine herbs in different areas. Conclusion This method provides a scientific reference for the rapid identification of the orange cultivars.