碎片化科研创新点动态挖掘研究

Dynamic Mining of Fragmented Scientific Research Innovation Points

从海量科技信息中挖掘出科研创新点碎片已成为大数据环境下知识挖掘与服务的一个关键问题，也仍然是迄今为止非结构化知识发现的一个难题。文章提出一种碎片化科研创新点动态挖掘方法。通过对学术成果的要素和条件分析，建立学术成果创新要素的关键变量和语义关系，给出学术成果创新点的本体模型；基于模型的理论指导，实现科技文献中科研创新点碎片的动态挖掘系统。该方法有利于过滤海量科技文献的创新点，发现文献中的知识关联关系，提高文献知识挖掘的效率，为科研工作者快速方便地直接获取科研动态信息提供技术支持。

Innovation fragments excavated from the mass of information in science and technology have become a key issue in large data mining and knowledge services, which remains a problem so far in unstructured knowledge discovery. This paper presents a fragmented innovation dynamic mining method. Through the analysis of the elements and conditions of academic achievement, we establish key variables and semantic relationships of innovative elements in academic achievements, and give an ontology model of innovation in academic achievement. Based on theoretical models, we achieve a dynamic mining system of science and technology research and innovation literature debris. This method is conducive to innovation filtration of massive scientific literature. We also find the association between knowledge of scientific literature, improve the efficiency of knowledge mining literature, and help researchers access dynamic information quickly and easily.