面向电子商务领域的知识库构建技术研究

发布时间：2018-06-15 09:43

本文选题：知识库 + 概念格　；参考：《华北理工大学》2017年硕士论文

【摘要】：基于电子商务的迅猛发展,提供的导航、检索等功用已很难满足顾客的需求,如何获取数据中存在价值的信息,又怎么依凭这些信息给予恰当的处理,提供令用户满意的服务,成为了各大电商企业所关注的问题。知识库的构建是提升知识的获取、共享、维护和运用能力的有效措施。知识库作为基于知识的系统,其的构建有利于提高预测模型和推荐系统的质量,将是电商企业为消费者群体提供个性化、定制化和智能化服务的核心支撑。目前基于领域的知识库构建研究并不多,且知识库构建的自动化程度并不高。应用概念格理论、本体、知识工程等技术,在一定程度上能够提高知识库构建的自动化程度。首先,依据知识库理论以及电子商务领域现状分析,设计了面向电子商务领域的知识库体系结构。其次,研究了基于完备和不完备信息下的知识获取方法,在完备信息情况下,基于概念格理论、本体论,以电子商务平台的用户个人信息、商品属性信息等作为数据源,构建概念格,进而构建领域本体,提取关联规则;在不完备信息情况下,在区间概念格的构建思想上,提出了一种近似概念格来解决不完备信息下的知识获取问题,通过与偏大近似概念格和偏小近似概念格这两种策略进行对比,验证了它的可行性和有效性。最后,研究了不同本体的合并以及随着时间发生变化的规则更新的问题,本体合并主要通过将本体解析后形成的形式背景进行概念格构建、概念格的合并,然后利用概念格到本体的映射规则将概念格转化为本体;变化的规则更新则是针对于不同时间段产生的关联规则发生了改变的状况,给出了规则的变化模式以及识别判定方法,并针对不同的规则变化模式采取了相应的更新策略。
[Abstract]:Based on the rapid development of electronic commerce, the functions of navigation and retrieval have been difficult to meet the needs of customers. How to obtain valuable information in the data, and how to rely on these information to give appropriate processing to provide satisfactory services to users, Has become the major e-commerce enterprises concerned about the problem. The construction of knowledge base is an effective measure to improve the ability of acquiring, sharing, maintaining and applying knowledge. As a knowledge-based system, the construction of knowledge base is helpful to improve the quality of prediction model and recommendation system, and it will be the core support for e-commerce enterprises to provide personalized, customized and intelligent services for consumers. At present, there are few researches on domain-based knowledge base construction, and the automation of knowledge base construction is not high. The application of concept lattice theory ontology knowledge engineering and so on can improve the automation degree of knowledge base construction to a certain extent. Firstly, according to the knowledge base theory and the analysis of the current situation in the field of electronic commerce, the knowledge base architecture oriented to the field of electronic commerce is designed. Secondly, the method of knowledge acquisition based on complete and incomplete information is studied. In the case of complete information, based on concept lattice theory, ontology, the user's personal information and commodity attribute information of electronic commerce platform are used as data sources. Construct concept lattice, then construct domain ontology, extract association rules, in the case of incomplete information, in the construction of interval concept lattice, we propose an approximate concept Ge Lai to solve the problem of knowledge acquisition under incomplete information. The feasibility and effectiveness of these two strategies are verified by comparing them with the two strategies: the larger approximate concept lattice and the smaller approximate concept lattice. Finally, the merging of different ontologies and the updating of rules over time are studied. Ontology merging is mainly done by constructing concept lattices and merging concept lattices by parsing the formal background of ontology. Then, the concept lattice is transformed into ontology by the mapping rules of concept lattice to ontology, and the changing rules are updated according to the situation that the association rules produced in different time periods have changed, and the changing pattern of rules and the method of identifying and judging are given. The corresponding updating strategy is adopted for different rule changing modes.
【学位授予单位】：华北理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F724.6;TP391.1

【参考文献】