计算机科学与探索 ›› 2019, Vol. 13 ›› Issue (2): 214-225.DOI: 10.3778/j.issn.1673-9418.1712057

• 数据挖掘 • 上一篇    下一篇

数据、信息和知识三层图谱架构的推荐服务设计

邵礼旭1,段玉聪1+,周长兵2,高洪皓3,陈世展4   

  1. 1. 海南大学 信息科学技术学院,南海资源利用海洋国家重点实验室,海口 570228
    2. 中国地质大学 信息工程学院,北京 100083
    3. 上海大学 计算机工程与科学学院,上海 200444
    4. 天津大学 计算机科学与技术学院,天津 300072
  • 出版日期:2019-02-01 发布日期:2019-01-25

Design of Recommendation Services Based on Data, Information and Knowledge Graph Architecture

SHAO Lixu1, DUAN Yucong1+, ZHOU Zhangbing2, GAO Honghao3, CHEN Shizhan4   

  1. 1. State Key Laboratory of Marine, Resource Utilization in the South China Sea, College of Information and Technology, Hainan University, Haikou 570228, China
    2. School of Information Engineering, China University of Geosciences, Beijing 100083, China
    3. School of Computer Engineering and Science, Shanghai University, Shanghai 200444, China
    4. College of Computer Science and Technology, Tianjin University, Tianjin 300072, China
  • Online:2019-02-01 Published:2019-01-25

摘要: 海量的学习资源会引起用户在学习过程中产生认知过载和资源迷航,对数据、信息和知识等形态的资源的原始表述的自然语言的机器理解、自动处理、自动综合和自动分析等成为了巨大的挑战。从应对自动增量式结合经验知识和减少人工专家交互负担等两方面考虑,从资源建模、资源处理、处理优化和资源管理等角度进行研究,基于对现有知识图谱(knowledge graph)概念的拓展提出了一种三层可自动抽象调整的解决架构。该架构借助从数据图谱上以实体综合频度计算为核心的分析到信息图谱和知识图谱上的自适应的自动抽象的资源优化过程支持经验知识引入和高效自动语义分析。该框架借助对应5W(who/when/where,what and how)问题的分类接口衔接用户的学习需求等资源化描述,为用户提供个性化学习服务推荐。

关键词: 资源建模, 知识图谱, 服务推荐, 语义建模

Abstract: Faced with complex data, information and knowledge resources, users are easy to get lost and overloaded in the process of learning. Automated machine understanding, automated processing, automatic synthesis, and automatic analysis of natural language, such as the original representation of the resources of these data, information and knowledge, have become a huge challenge. This paper studies resource modeling, resource processing, processing optimization and resource management from the aspects of coping with automatic incremental knowledge and reducing the interaction burden of artificial experts and puts forward a three-tier solution architecture which can automatically abstract and adjust resources based on the concept extension of existing Knowledge Graph. This architecture recursively supports integration of empirical knowledge and efficient automatic semantic analysis of resource elements through frequency focused profiling on data graph and optimal search through abstraction on information graph and knowledge graph. The proposed architecture is supported by the 5W (who/when/where, what and how) to interface users?? learning requirements which can provide users with personalized learning service recommendation.

Key words: resource modeling, knowledge graph, service recommendation, semantic modeling