【作者】 王军；

【导师】 聂规划；

【作者基本信息】 武汉理工大学 ， 管理科学与工程， 2008， 博士

【摘要】 房地产企业能否准确把握客户的需求,以及能在多大程度上满足客户的个性化需求,决定了企业的竞争能力。对客户进行个性化的营销服务,企业不仅能够获得较大的利润,而且能有效的留住已有的各类客户,吸引和开发新客户。目前房地产业的个性化营销还只是处于初步的研究阶段,大部分研究只是讨论个性化营销对于企业的意义,很少涉及到如何实现房地产的个性化营销。本文提出的基于本体技术和案例推理的房地产个性化营销系统为其提供了新的思路。论文在借鉴国外现有的房地产营销、知识管理和案例推理的理论、方法和技术的基础上,探讨了房地产个性化营销系统的构建,其目的是运用案例的知识来解决房地产个性化营销的问题,并探讨了营销案例知识的共享与重用和案例的优化。论文主要研究内容和成果如下:(1)客户本体的构建。客户本体的建立能够使房地产各企业或企业各部门在一致的概念和语义基础上应用客户信息。本文在研究了本体的定义、建模语言和构建方法的基础上,对客户本体的构建进行了分析。本文首先获取了客户本体术语;然后在从多维角度产生的房地产客户分类的基础上,定义了客户本体的类及其类的属性;最后探讨了客户本体的表示方法。客户本体的建立为多案例库中的案例知识的共享奠定基础。(2)房地产本体的构建。考虑了手动构建本体的不足,房地产本体的构建采用了形式概念分析的方法。本文首先探讨了形式概念分析的理论基础、定义及其概念格的构建方法及其形式概念分析与本体间的联系;然后在此基础上,通过从房地产领域文本中获得反映该领域的概念中构造出房地产的形式背景;最后通过概念格结构化概念以及概念之间的分类关系,并将其转化为房地产本体。房地产本体能够消除不同企业甚至企业不同部门对房地产信息的理解出现的偏差和歧义,为房地产营销案例知识的共享奠定基础。(3)基于本体的房地产营销案例推理系统的设计。本文在探讨了本体和案例推理的理论及其应用的基础上,研究了如何将本体和案例推理技术应用于房地产个性化营销,设计了基于本体的房地产营销案例推理系统。该系统不仅考虑了领域专家的经验和知识的充分利用,而且能够支持多分布案例库之间的案例知识的共享与重用,解决了传统的案例推理方法只是从语法上对案例知识进行描述,并不对其语义进行描述而造成的查全率和查准率较低的难题。并且提出了新的语义相似度计算方法,采用该方法来计算新案例和从案例库中提取的案例之间的相似度,发现不同案例表达形式下各自属性所代表的真实含义,并在此基础上建立属性之间的映射关系。(4)基于CBR的住房定制研究。通过让用户一步一步地参与住房设计来提升其满意度是一项有效的营销策略,本文在研究了住房定制现状的基础上,提出将CBR用于住房定制的系统的研究,满足客户对住房的个性化需求。首先设计了基于CBR的住房定制框架;然后研究了住房案例属性的选取和表示;着重研究了基于遗传算法的住房定制案例检索和基于遗传算法的住房定制案例修改。基于遗传算法的住房定制检索解决了最近相邻法中案例属性特征权重的主观性,从而提高案例匹配的有效性;基于遗传算法的案例修改通过解决同时满足成本的最小化和客户评分的最大化的问题,提供一个令客户最满意的住房方案。该系统提供的服务能够为客户提供表达他们特定需要的机会,并根据其偏好和成本的要求为客户提供最佳的量身定制住房方案,解决了住房定制实现的难题。更多还原

【Abstract】 Whether real estate enterprises can grasp the customer’s demand accurately, and satisfy the customer’s personalized demand maximally,has decided the enterprise’s competitive ability.Personalized marketing services to the customers not only can bring profit to the enterprise,but also retain the existing customer and develop new customer effectively.At present,the personalized marketing services of real estate enterprise are only at the preliminary research stage.The most researches only discussed the significance of personalized marketing,and didn’t research how to carry out the personalized marketing of real estate.The paper proposed personalized marketing system for real estate based ontology and case-based reasoning technology which provide a new way of thinking.It was researched to construct the real estate personalization marketing system on the basis of the researches on real estate marketing knowledge and the theory and method of knowledge management and case based reasoning technology.Its aim was to use the knowledge of cases to solve the real estate personalization marketing problem.The paper also discussed the sharing and reusing of marketing case knowledge and the optimization of cases.The main research contents and results are as follows:(1) Customer ontology construction.The customer ontology can make real enterprises or departments apply customer information on the basis of consistent concept and semantic information.The paper analyses the construction of customer ontology based on the research of ontology definition,ontology modeling language and construction method.First,the customer ontology terms were obtained.Second,the class and the class’s property of the customer ontology were defined on the basis of real estate customer classification generated from multidimensional point of view.Finally,the expression of customer ontology was discussed.The establishment of customer ontology lays the foundation for case knowledge sharing in multi-case bases.(2) Real estate ontology construction.Considering the insufficiency of manual construct ontology,real estate ontology construction adopts formal concept analysis approach.First,this paper researched its theoretical basis, definition,construction method of concept latices and the relationship between formal concept analysis and ontology.Second,the real estate’s formal background was constructed based on the concepts from correlative real estate domain text.In the formal background,the objects are the documents and the properties are the ontology concepts.Finally,the concepts and classification relationships between concepts were structured in the form of concept latices and translated into real estate ontology.Real estate ontology can eliminate deviation and ambiguity which emerge from the understanding of real estate information in different enterprise or even different department and provide the basis for real estate marketing case knowledge sharing.(3) Ontology-based real estate CBR system’s design.How to use ontology and case based reasoning technology in real estate personalized marketing was designed.The system structure of real estate personalized marketing was designed on the basis of the discussion of ontology and case based reasoning theory and its application.The system not only considers the full use of domain experts’ experiences and knowledge,but also can supports sharing and reuse of case knowledge in multi-distribution case bases.So,it solved the problem of knowledge sharing generated from different case express in different enterprise or enterprise interior in traditional case based reasoning.Also a new semantic similiarity algorithm was proposed and used to compute similarity between new case and a case from case bases.So the real meaning of each property in different case expression was discovered and the mapping relation between two properties was constributed based on it.(4) CBR based housing customization research.It is an effective marketing strategy to improve customers’ satisfaction through customers’ participation in residency design step by step.This thesis researched the application of CBR in housing customization system to meet customers’ personalized demand,based on research of the current situation of housing customization.First,housing customization framework based on CBR was designed.Then the choice and expression of housing case properties were researched.Finally,the housing customization case indexing and revision based on genetic algorithm were researched.The housing customization case index based in genetic algorithm solved the subjectivity problem of case properties’ characteristic weight in the most near neighboring method and improved the validity of case match.The housing case revision based on genetic algorithm provided a satisfied housing scheme through meeting the demand of cost minimization and customer grades maximization. The system can provide customers with the chances of expressing their special needs and the best customized housing scheme according to their preferences and requirement of cost.更多还原