语义物联网中的情景不一致性消解策略

发布时间：2018-02-14 09:23

本文关键词： 语义物联网情景不一致性本体粗糙集形式概念分析证据论　出处：《大连海事大学》2017年硕士论文　论文类型：学位论文

【摘要】：在物联网中,由于物的信息具有多样化的描述形式且机器不能够完全理解这些物的信息(物联网的内在矛盾),进而很难有效地为普通用户提供语义服务。语义物联网能够消解物联网的内在矛盾,为用户提供相应的语义服务。它在物联网的基础上纳入"语义协同",即基于本体的语义标注和基于本体的语义理解。情景感知是语义物联网的核心组成部分,情景不一致性是情景感知的关键问题,导致用户无法获得正确的情景进而无法进行情景推理。本文综合利用粗糙集理论、本体、形式概念分析以及证据论,提出一种适用于语义物联网的情景不一致性消解策略,利用本体来标注情景和情景模式并建立常识库,找出不一致的情景,利用粗糙集理论从常识库的决策表中获得约简属性,再利用形式概念分析和证据论找出最佳的情景。通过与近期相关研究(基于语义的情景不一致性消除算法、基于约束的情景不一致性消除算法以及基于证据论的情景不一致性消除算法)的评估对比以及大量实验证明本文所提出的策略具有较高的正确率和可行性,本文所提出的方法既具有语义性,也适应语义物联网的动态环境。
[Abstract]:In the Internet of things, Because the information of objects has a variety of description forms and the machine can not fully understand the information of these objects (the inherent contradiction of the Internet of things, it is very difficult to provide semantic services to ordinary users effectively. The semantic Internet of things can be dispelled. The inherent contradiction of the Internet of things, It includes semantic collaboration on the basis of the Internet of things, namely Ontology-based semantic annotation and Ontology-based semantic understanding. Situational perception is the core component of semantic Internet of things. Situational inconsistency is a key problem in situational perception, which results in the user being unable to obtain the correct situation and then unable to infer the situation.In this paper, rough set theory, ontology, formal conceptual analysis and evidence theory are synthetically used. In this paper, a scenario inconsistency resolution strategy suitable for semantic Internet of things is proposed. Ontology is used to annotate scenarios and situational patterns, and common sense database is established to find inconsistent scenarios, and rough set theory is used to obtain reduction attributes from decision tables of common sense database. Then we use formal conceptual analysis and evidence theory to find out the best scenario. The evaluation and comparison of Constraint-based scenario inconsistency cancellation algorithm and evidence-based scenario inconsistency cancellation algorithm and a large number of experiments show that the proposed strategy has a high accuracy and feasibility. The method proposed in this paper is both semantic and adaptive to the dynamic environment of semantic Internet of things.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.44;TN929.5

【参考文献】