基于知识关联的多层本体立方体设计与实现——以金融证券领域为例  

Design and Implementation of Multilayer Ontology Cube Based on Knowledge Association——Taking the Financial Securities as an Example

作　　者：刘政昊 Liu Zhenghao(School of Information Management,Wuhan University,Wuhan 430072,China;Institute of Big Data,Wuhan University,Wuhan 430072,China;Center for Studies of Information Resources,Wuhan University,Wuhan 430072,China)

机构地区：[1]武汉大学信息管理学院,湖北武汉430072 [2]武汉大学大数据研究院,湖北武汉430072 [3]武汉大学信息资源研究中心,湖北武汉430072

出　　处：《现代情报》2022年第1期72-86,共15页Journal of Modern Information

基　　金：国家自然科学基金重大研究计划“大数据驱动的管理与决策研究”重点支持项目“基于知识关联的金融大数据价值分析、发现及协同创造机制”(项目编号:91646206);科技创新2030—“新一代人工智能”重大项目“亿级节点时序图谱实时智能分析关键技术与系统”(项目编号:2020AAA0108505)。

摘　　要：[目的/意义]结合金融证券行业特征,借鉴层次式设计思路和数据立方体概念,提出多层领域本体立方体模型并完成构建。[方法/过程]复用FBIO本体进行知识建模;利用LDA主题建模与BIRCH层次聚类完成概念提取;基于依存句法和深度学习框架的知识抽取完成本体实例扩充;通过维度分类和基于概率的实体空间向量表示增强语义关联性。[结果/结论]多层构建方式和立方体结构增加了知识内在关联,为金融概念知识提供多层次、细粒度的知识组织方式;也为本体构建提供新的思路。[Purpose/Significance]Based on the characteristics of financial securities industry,a multilevel domain ontology cube model was proposed and constructed by referring to the concept of hierarchical design and data cube.[Method/Process]FBIO was used ontology for knowledge modeling;LDA topic modeling and Birch hierarchical clustering were used to complete concept extraction.Ontology instance expansion was completed by knowledge extraction based on dependency syntax and deep learning framework.Semantic relevance was enhanced through dimension classification and probabilistic entity space vector representation.[Result/Conclusion]The multi-level structure and the cube structure increased the internal correlation of knowledge,and provided a multi-level and fine-grained knowledge organization mode for financial concept knowledge.It also provides new ideas for ontology construction.

关 键 词：多层领域本体 本体立方体 金融证券 知识关联 层次聚类 知识抽取 

分 类 号：G254[文化科学—图书馆学]