基于描述逻辑的用户偏好建模方法研究

发布时间：2018-05-16 08:28

本文选题：描述逻辑 + 偏好　；参考：《东北大学》2013年硕士论文

【摘要】：随着互联网的不断发展,人们已经进入了信息化时代。网络上的资源和信息量呈爆炸性的增长,这些信息在丰富人们生活的同时也给使用网络的人们带来了很多问题。过量的信息使得人们想从当中准确地找到自己需要的信息变得十分困难,用户对信息的需求具有多样性,Web搜索引擎并不能根据用户的特点,为用户提供具有个性化的服务。因此如何了解用户意图,掌握用户的个性化信息,提供个性化服务成为一个亟待解决的问题。在个性化信息服务中,用户模型是个性化服务研究的重点。个性化服务首先需要建立用户模型,然后才能针对不同用户提供个性化服务。传统的用户模型大都使用关键词来表示用户兴趣,能够表达的形式简单,并且没有考虑词和词之间内在的关联,不能从语义上对信息进行理解。描述逻辑是基于对象的形式化的知识表达,具有良好的表达能力和可判定性,能够在知识库的概念之间进行语义关联并利用推理能力对描述逻辑中的知识进行推理。针对以上问题,本文提出了基于描述逻辑用户偏好建模方法。根据描述逻辑表达能力强的特点,将偏好以描述逻辑的形式进行表达,并对描述逻辑偏好表达的结构和内容进行说明和分析,对用户偏好模型进行形式化的描述,为建模提供框架结构。在描述逻辑表达的偏好基础上,对用户偏好模型进行建模和更新等各种操作。描述逻辑具有强大的推理能力,通过描述逻辑的推理和偏好之间的语义关联,从偏好模型建立和使用的角度,对模型中的冲突检测、优化表达,以及知识的获取三方面进行推理分析。根据提出的方法,本文通过给出实例,对偏好模型的建立、偏好的更新以及偏好的推理三方面进行了实验的验证,并且通过实验证明了方法的可行性。
[Abstract]:With the continuous development of the Internet, people have entered the information age. The amount of resources and information on the network increases explosively, which not only enriches people's lives, but also brings a lot of problems to people who use the network. Too much information makes it very difficult for people to find the information they need accurately. Web search engine can not provide users with personalized services according to the characteristics of users. Therefore, how to understand the user's intention, grasp the user's personalized information and provide personalized services has become a problem to be solved. In personalized information service, user model is the focus of personalized service research. The personalization service needs to establish the user model first, then can provide the personalization service to the different user. Most of the traditional user models use keywords to express user interest, which can express the interest of users simply, and do not consider the inherent relationship between words and words, so they can not understand the information from the semantic point of view. Description logic is a formal knowledge representation based on object. It has good expressive ability and decidability. It can correlate the concepts of knowledge base and use the reasoning ability to infer the knowledge in description logic. In order to solve the above problems, this paper proposes a modeling method based on description logic for user preference. According to the strong expressive ability of description logic, preference is expressed in the form of description logic, the structure and content of description logic preference expression are explained and analyzed, and the user preference model is described formally. Provide framework structure for modeling. On the basis of describing the preference of logical expression, the user preference model is modeled and updated. Description logic has powerful reasoning ability. By describing the semantic relation between reasoning and preference of logic, from the perspective of establishing and using preference model, the conflict detection and expression in the model are optimized. And the acquisition of knowledge from three aspects of reasoning analysis. According to the proposed method, an example is given to verify the establishment of preference model, the renewal of preference and the reasoning of preference, and the feasibility of the method is proved by experiments.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP393.09

【参考文献】