知识图谱驱动的广东省自然资源大数据挖掘模型构建框架

Framework for Knowledge Graph-driven Construction of Natural Resources Big Data Mining Model in Guangdong Province

新一轮机构改革,形成了全国自然资源统一管理的新体系,在机构改革前各部门分别建设的"烟囱式"数据与应用系统,正在集成为自然资源大数据以及构建自然资源大数据挖掘的框架,然而这主要是基于专家的知识领域逐个构建,尚未形成通用的框架。本文提出知识图谱驱动的自然资源大数据挖掘模型构建框架,将专家的知识固化在自然资源大数据挖掘模型的知识图谱表达中,提高了挖掘模型的复用能力,并在广东省国有土地对经济发展保障潜力评价的应用案例中得到有效验证,具有广阔的应用前景。

The latest governmental organization reform led to a new administrative system of big ministry of natural resources.The isola-ted data sets and application systems constructed by natural resources related departments before the merger of the natural resources ministry,are being integrated to the natural resources big data and the framework of natural resources big data mining,which mainly relying on experts’domain knowledge case by case without a common framework.This paper is intended to propose a knowledge-graph driven framework for constructing natural resources big data mining models,which is applied to describe and store expert’s knowledge.The framework improves the reuse of expert’s knowledge.Tested on the model of evaluating potential capabilities of the state-owned land of Guangdong Province supporting economic development,the effectiveness of this framework is demonstrated.This framework is widely applicable in natural resources big data mining.