学者需要在海量的数据中提取有助于学术研究的信息和知识，在面对大量数据时难以从繁杂的论文数据中提取学者需要的有用信息。现有系统中准确推荐学术论文高度依赖于相关领域知识，知识图谱可以对领域知识进行表示和计算，提高信息检索效率。将知识图谱应用于学术论文推荐领域，是比较合理且可行的。本文研究了基于知识图谱的学术论文推荐方法，利用学术领域知识图谱增强推荐系统的能力。一方面，对论文涉及的核心概念进行识别和链接；另一方面，基于知识图谱对学者研究兴趣进行表示。在此基础上，研究了个性化的推荐方法，对相应论文研究者推荐可能感兴趣的论文列表，实现了个性化对不同论文研究者进行相应推荐的效果。本文的方法在真实数据上进行实验分析，推荐的准确率经过验证达到了 80%以上，体现了本文推荐系统的能力。

Today's scholars need to extract information and knowledge that is helpful to academic research from massive data, and extract useful information that scholars need from the complex paper data. In the existing systems, the accurate recommendation of academic papers highly depends on the relevant domain knowledge. Knowledge mapping can represent and calculate the domain knowledge, and improve the efficiency of information retrieval. It is reasonable and feasible to apply knowledge mapping to the field of academic paper recommendation. This paper studies the method of academic paper recommendation based on knowledge map, and uses the knowledge map of academic field to enhance the ability of recommendation system. On the one hand, identify and link the core concepts of the thesis; On the other hand, the research interest of scholars is expressed based on knowledge map. On this basis, the personalized recommendation method is studied to recommend the list of papers that may be of interest to the corresponding researchers, which achieves the effect of personalized recommendation for different researchers. The method is tested on real data, and the accuracy of recommendation isIII over 80%, which reflects the ability of recommendation system.