基于大数据挖掘的科技项目查重模型研究

Study on Detection Model of Similar Scientific Project Based on Big Data Mining

科技项目查重是避免重复立项、重复建设的重要措施之一,目前缺乏行之有效的方法。文章提出基于大数据挖掘和多源信息整合的项目查重方法,以科技项目的基本信息、发表论文信息、关键词、负责人信息和承担机构等要素构建的大数据网络为研究对象,利用多源信息整合方法构建科技项目的相似度判别模型,并采用Hadoop框架实现海量数据的快速挖掘。文章介绍项目查重模型,重点讨论需要解决的关键问题,为解决项目查重问题提供一种全新的思路和方法。

Checking out similar project is very important to avoid repetition in project approval. However, there is no way to find out similar project effectively tbr the moment. This paper proposes a novel method of detecting scientific projects similitude based on big data mining and multi-source information integration. Using that method, the authors studied the huge data network consisting of the intbrmation about the project, published papers, experts and institutions, as well as the keywords; built up a detection model of project similitude by integrating multi-source information; and adopted the Hadoop to speed up big data mining. This paper presents the detection model of project similitude and its key issues; in hope of providing brand-new thinking and methods for detecting similar projects in scientific project management.