基于领域本体的油气勘探开发知识获取模式及实现  

Oil and gas exploration and production knowledge processing mode and implementation based on domain ontology

作　　者：王晓丽[1,2] 邓达康 孟祥龙 唐先明[1] 郭攀红 林畅松[3] 

机构地区：[1]中国石油化工股份有限公司石油勘探开发研究院,北京100083 [2]中国地质大学(北京)能源学院,北京100083 [3]中国地质大学(北京)海洋学院,北京100083

出　　处：《东北石油大学学报》2016年第4期74-79,87,共7页Journal of Northeast Petroleum University

基　　金：国家科技支撑计划项目(2012BAH34F04)

摘　　要：油气勘探开发行业的信息数据资源涉及多学科、多地域,结构复杂,类型多样。提出一种基于领域本体的油气勘探开发知识获取模式,提取知识源的领域命名实体,将海量知识与已知的领域本体关联起来,实现快速知识获取;通过改进Trie索引树词典构造方法、创建节点失败指针优化正向最大匹配分词方法,快速匹配并提取已登录领域命名实体;采用领域命名实体规则特征和聚类统计分析方法,识别未登录领域命名实体(包括疑似命名实体和新增命名实体)。该模式能够提高油气勘探开发领域知识源的命名实体提取效率和准确率,有效解决油气勘探开发行业获取知识的难题。The information resources of oil and gas exploration and production involve multi-disciplinary and multi-regional,structures are complex,data types are various.How to get the required knowledge quickly is very difficult.This paper proposes a knowledge acquisition mode based on the domain ontology of oil and gas exploration and production,which allows the massive knowledge associated with the known domain ontology through extracting the domain named entity from knowledge texts.The mode enables rapid knowledge acquisition.The difficulties of the mode are how to extract known entity quickly and how to extract unknown entities accurately and effectively associated with knowledge sources.The model improves Trie algorithm of the dictionary,adds the failure pointer in the Trie creation and adopts the forward maximum match method to solve the first difficulty.On the other hand,the mode combines the rules and statistical methods to identify suspected ontology.The model improves the effi-ciency of matching participle and solves the recognition problem of the new entity.The model has obtained the good effect in the knowledge.

关 键 词：勘探开发 领域本体 领域命名实体 知识获取 

分 类 号：TE19[石油与天然气工程—油气勘探] TP311[自动化与计算机技术—计算机软件与理论]