【作者】 曾珠；

【导师】 聂规划；

【作者基本信息】 武汉理工大学 ， 管理科学与工程， 2013， 博士

【摘要】 当前中国汽车工业的飞速发展,也带动着汽车售后服务产业的急速扩张。虽然国内汽车服务增加潜力巨大,但是整体市场起步较晚,汽车服务商良莠不齐,渠道网络组织无序,服务措施亟待完善,从业人员素质较低,专业人才较为匮乏,客户投诉频发,顾客满意率较低,问题较为突出,缺乏先进的管理思想和技术手段。售后服务基本上依赖于汽车制造商的指导和要求,缺乏主动对市场进行分析,难以适应复杂的市场环境,服务同质化严重,服务品种单一,缺乏针对性的服务,未能实现差异化的服务。本文旨在构建客户行为数据库,通过利用各种工具进行分析与统计,从而获取每个客户对汽车的使用偏好特征,结合汽车专家知识以及汽车制造商所提供的服务规则,分析客户差异化行为对汽车性能状态的影响,并结合预测模型、标准服务规则和知识库,预测出每个客户下一次最有可能接受售后服务的项目及时间,最终为汽车服务商服务推荐提供支持。针对以前研究存在的不足,本文通过对国内汽车售后服务行业现状的分析,基于进化博弈论和SD模型对汽车售后商服务推荐策略进行分析。然后,构建了客户行为指标体系,通过对客户行为本体和售后服务本体的研究,实现了基于本体的服务推荐和基于D-S证据理论的汽车售后服务案例推理。在此基础上,建立了基于集成案例推理的汽车售后服务推荐系统,用于协助汽车售后企业实现差异化服务和主动服务。论文的主要工作如下：对汽车售后服务的概念进行了界定,分析了汽车售后服务行业的特点,比较研究了常见的汽车售后服务经营模式,分析了国内汽车售后服务行业的现状,指出了存在的主要问题,并结合汽车售后服务行业发展前景分析,提出了国内汽车售后服务企业的经营策略。通过对人类行为模式的分析,提出了影响客户行为的因素体系,在此基础上对影响客户行为的生理因素、心理因素、自然环境因素和社会环境因素的构成及其与汽车售后服务的关系进行了研究,借助结构方程模型构建了客户行为指标体系,用于分析诸多因素间及行为因素与汽车售后服务间的影响机制,并通过样本数据进行了模型验证。将本体论(Ontology)和案例推理(CBR)引入服务推荐研究,利用本体描述语言OWL和本体建模工具protege构建了生理因素本体(PFO)、心理因素本体(POFO)、自然环境因素本体(NFO)、社会环境因素本体(SFO)、汽车领域本体(ADO)和汽车售后服务本体(AASO),并结合上述本体间的关联,构造了客户行为——服务本体(CBSO)模型,并根据此模型提出了基于本体的汽车售后服务方案匹配方法,并给出了计算实例。探讨了在案例之间存在着冲突情形下的汽车售后服务知识推理方法,首先运用粗糙集理论对案例库进行约简,以提取案例库的特征属性并以此形成基本的推理证据,运用决策支持强度及扩展决策支持强度的方法确定各个证据的基本概率赋值,然后运用D-S证据理论对各个证据进行合成从而实现在案例存在冲突情形下的知识推理。最后运用上述方法对湖北某汽车销售及售后服务公司的汽车刹车片的案例库进行了实例研究,并证明了该方法的有效性。对比分析了汽车售后服务推荐系统与汽车售后服务推荐、汽车售后服务系统、汽车售后服务管理系统、汽车售后维修管理系统等概念之间的区别,探讨了汽车售后服务推荐系统的基本功能,提出了汽车售后服务推荐系统的支撑技术。选择规则推理与案例推理进行集成,并给出集成系统的总体框架,给出具体的实现步骤。最后提出基于集成案例推理的汽车售后服务推荐系统框架,并对其组成结构进行了具体分析和介绍。论文最后对研究工作进行了总结,并提出了有待于进一步研究的问题和方向。更多还原

【Abstract】 Currently, with the rapid development of China’s auto industry, auto after-sales service is expanding rapidly. Although the increase in domestic auto service has great potential, the overall market started too late. Automotive service providers vary greatly and their channel network is disorganized, their service needs effective improvement measures, the quality of their employees are relatively low, professionals are even more scarced, customer complaints is frequent, customer satisfaction rate is low, and advanced management ideas and techniques are scarced. After-sales services basically rely on auto manufacturers’instructions and requirements and they are shot of initiative analysis on the market, so it is difficult for them to adapt to the complex market environment, their service lackes variety, personalized services is lacked, and service differentiation can’t be realized.This paper aims to build a customer behavior database, obtain every customer’s consuming and using preferences with various analytical and statistical tools, analyze the impact of customer differentiation behavior in automobile performance based on expert knowledge and service guide provided by manufacturers, and predict the most likely item and time for every customer’s next service based with prediction model, standard service guide and knowledge database, to provide technical support for service providers to take differentiated initiative service.In allusion to the limitations of previous researches, this paper put forward business strategy which is suitable for domestic auto after-sales enterprises with analyzing domestic auto after-sales services presentation and combining auto after-sales services industry’s characteristics. Meanwhile, it studies some theories on service mining and builds up auto after-sales mining frame based on customers behaviors differentiation. Later it constructs customers behaviors index system, through researching on customer behaviors ontology and after-sales services ontology, realizes service matching based on CBR. And constructs business intelligent decision support system based on integrated CBR. The main works are as follows:This paper defines auto after-sales services, analyses auto after-sales services industry’s features, compares common auto after-sales services business mode, analyses domestic auto after-sales services industry status, point out existed main problem, and puts forward domestic auto after-sales services enterprises’ business strategy based on auto after-sales services industry’s development prospects,.By analyzing human behavior pattern, this paper presented a factors system that influence the customers’behavior, and on the base of it, the physiological factors, psychological factors, natural environmental factors which would affect customer behavior and constitute of social environmental factors and its relationship with the automotive after-sales are analyzed. And then, this paper built the customer behavior index system with structural equation modeling to analyze the impact mechanism between many factors especially behavioral factors and inter-car service, and conducted model validation wiht sample data.By introducing Ontology and CBR into service mining research, this paper builds physiological factors ontology(PFO), psychological factors ontology(POFO), natural environment factors ontology(NFO), social environment factors ontology(SFO), auto domain ontology(ADO), auto after-sales services ontology(AASO) with OWL and protege, and combines relationship between above ontology, constructed customers behaviors service ontology model(CBSO), and then, based on this model, put forward auto after-sales services program matching method based on similarity case reasoning.This paper discussed automotive after-sales knowledge reasoning method in the case of conflict situations. First, extract characteristics of the case base properties and thus to form a basic reasoning evidence with rough set theory to reduce the case library, then determine the basic probability assignment of each evidence with decision support strength method and expansion decision support method, and then realize the knowledge reasoning within conflict in the case with D_S evidence theory for the synthesis of evidence for each case and thus. Finally, this paper performed a case study with the method described above on a car brake pads case library from a Hubei automobile sale and service company and demonstrated the effectiveness of the method.This paper analyzed the difference between the concepts of the automotive after-sales car service recommendation system, automotive after-sales car service recommendation, automotive service management systems, and automotive after-sales maintenance management system, and discussed basic functions of automotive service recommended system and proposed automotive service recommendation system supporting technology. This paper integrated rules reasoning and case-based reasoning, and put forward the overall framework of the integrated system and the specific implementation steps. Finally this paper proposed automotive after-sales recommender system framework based on integrated CBR, presented and analyzed its composition in detail.Finally, the paper summarized the research work, and proposed issues and directions to be further researched.更多还原