随着电子化人力资源管理、智能招聘系统、社交网络平台、电子就业服务等技 术在人岗匹配中应用程度的不断加深，以及数据挖掘、机器学习、知识本体、案例 推理、个性化推荐等人工智能技术以及云计算、大数据等技术在人文社科领域的广 泛应用与发展，存储和使用与人岗匹配相关的知识将变得更加智能化和个性化。同 时，伴随着知识经济时代的到来，知识型员工作为内生经济增长与企业核心竞争力 提升的主要动力，在矛盾多变的职业环境中，面临着职业相关的各种挑战，取得职 业成功的难度正在不断增加，知识型员工需要通过一种智能化的方式及时掌握劳 动力就业市场中的岗位胜任能力所需以及变化趋势，并通过积极行为的改变和终 身学习的方式来不断地提高自己的胜任能力，时刻保持能够被劳动力市场雇佣的 能力以及向更好的职业生涯迈进的能力。 同以往传统的投递简历、筛选简历等在线招聘等技术相比，基于本体的人岗匹 配案例推理技术可以为人岗语义匹配提供技术手段支持，可以使人岗匹配知识智 能集成，简化和加快人与岗位的匹配过程，使企业能够根据岗位所处的具体情境属 性特征，更加精准、动态、智能、全方位地明确知识型员工岗位所需的胜任能力特 征，并以此为依据来预测和匹配岗位最佳员工，使企业知识型员工与岗位时刻保持 最佳的配置状态。同时，也使知识型员工能够及时掌握劳动力就业市场中岗位胜任 能力所需，并不断地提高自己的胜任能力，持续保持能够被就业市场雇佣的能力。 但是，如何基于数据挖掘、机器学习、知识本体、案例推理、个性化推荐等人工智 能技术，构建基于本体（Ontology）与案例推理（Case-Based Reasoning，简称CBR） 的人岗匹配模型，如何开发智能化的人岗匹配系统，使海量的知识型员工与其岗位 的多重语义资源进行共享与重用并精准测算其相似度，实现人岗匹配智能化是企 业人力资源管理面临的一个新课题。因此，对基于本体的人岗匹配案例推理技术进 行研究具有重要的理论意义和应用价值。 本文以本体理论与案例推理理论为基础，立足于企业知识型员工人岗匹配智 能化目标，围绕人与岗位知识共享、特征识别、预测和匹配等决策存在的理论与方 法展开研究，主要研究内容如下： （1）企业知识型员工岗位胜任能力特征本体构建。针对人岗匹配领域知识的 多源、异构、不确定、语义不一致等问题，以互联网IT类知识型员工为例，利用 Python网络爬虫工具和Jieba分词技术收集和分析了海量的企业招聘信息，建立了 岗位需求胜任能力特征知识本体模型，实现了雇主需要的岗位需求胜任能力特征 知识的统一结构化表述，解决了岗位需求胜任能力特征知识的存储、组织和重用问 题，为后续进一步实现基于本体的人岗匹配案例表示、案例检索和案例推理等知识 III 共享提供了语义基础。 （2）基于本体的企业知识型员工人岗匹配案例表示。根据构建的岗位胜任能 力特征领域本体，明确了企业知识型员工人岗匹配案例表示的问题描述—情境描 述—解决方案三方面构成要素，定义了人岗匹配案例本体知识模型，并建立了基于 本体的人岗匹配案例知识建模体系——案例库，实现了对人岗匹配案例的统一结 构化规范表示，为案例相似度计算和精准匹配以及案例库的有效应用与维护提供 了基础。 （3）基于本体的企业知识型员工人岗匹配案例相似度计算与检索。根据人岗 匹配案例表示，对特定岗位问题与情境下的目标案例与案例库中的源案例进行基 于概念名称和属性的相似度计算，并根据检索结果绘制特定问题与情境下的岗位 最佳匹配者用户画像，作为目标案例的解决方案和企业评价候选人的标准，其相似 度计算与检索的质量决定了候选人隐性知识测算与案例推理系统实现的精准性和 智能性。 （4）基于案例推理检索结果的岗位候选人隐性知识测算与人才社区开发知识 共享。根据人岗匹配案例推理的检索结果，通过社交网络、工作日志、贝叶斯网络 方法，对候选人的性格偏好、工作业务行为、完成特定任务所反映出的隐性知识进 行测算与评价，并通过人才社区知识共享平台的开发来优化用户网络结构与职业 交流，为人岗匹配知识共享提供更好的途径，同时，更好地帮助知识型员工通过终 身学习的方式持续保持能够被劳动力市场雇佣的能力。 （5）基于本体的企业知识型员工人岗匹配案例推理系统的设计与实现。在上 述研究成果的基础上，结合案例推理系统的工作机制，设计与实现了完整的企业知 识型员工人岗匹配案例推理原型系统，基本完成了本文基于本体的企业知识型员 工人岗匹配的智能化、动态化、精准化和持续化等研究目的，为人岗匹配知识预测、 特征识别、匹配方案的制定以及匹配后效果评估提供决策支持服务。 关键词：本体；知识型员工；人岗匹配；案例推理系统；人力资源管理 IV ABSTRACT As the world moves towards a more technologically advanced means of recruitment, intelligent human resource management, social networking platforms such as electronic employment service technology application in post matching, these technologies along with theories such as data mining, machine learning knowledge ontology, and case-based reasoning, personalized recommendation and artificial intelligence such as cloud based computing and big data computing become increasingly embedded and widely used in the field of humanities, social science and development. Concurrently, the arrival of the era of knowledge economy, knowledge workers as the endogenous economic growth and promoting enterprise core competitiveness of main engine, in a contradictory and turbulent/changing environment, faced with a variety of professional challenges, the difficulty of career success is increasing. The knowledge workers need timely by means of a kind of artificial intelligence required in the employment market trends, and through the positive behavior change and the way of lifelong learning to constantly improve their own competence, keep you up with time can be hired through Labor market ability and the ability to better career life. In this background, how to build a model to implement intelligent person-post matching system, make vast amounts of knowledge workers need for their jobs to multiply semantic resource sharing and reuse, and accurately measure the similarity is the key to intelligent person-post matching. Ontology and case-based Reasoning (CBR) technology can solve this problem, the case-based Reasoning technology Based on Ontology can provide technological support for semantic matching. The previous traditional resume screening, online recruitment techniques, such as man of the match, case- based reasoning ontology technology can enhance person-post matching intelligent integration of knowledge, simplify and speed up the person-post matching process, to enable the organization(enterprise) to match specific job situations to employee attributes, characteristics and competencies more accurately and on this basis predict and match the best employee to a specific task, this will enable the organization to keep track of staff information and to keep the best data on staff. At the same time, it also enables knowledge workers to timely grasp the job competency requirements in the labor market, and continuously improve their own competency, and continue to be able to be employed in the job market, this research has important theoretical significance and application value. V Based on ontology theory and case-based reasoning theory, this dissertation focuses on the practical problem of person-post matching for knowledge workers in enterprises, and studies the key problems in decision-making such as knowledge sharing, feature recognition, prediction and matching between person and posts. The main research contribution and achievements are summarized as follows: (1) Ontology construction of job competency characteristics of enterprise knowledge workers. For the characterstics of person-post matching knowledge as polyphyly, heterogeneous, uncertainty, and the semantic inconsistency, this research focus on the internet IT knowledge workers, using the Python web crawler tools and Jieba participle technology to collect and analyze the mass recruitment information of enterprise, set up post demand competence characteristics of knowledge ontology model, implements the employers need to post demand knowledge of unified structured expression competence characteristics, solved the position requirements competencies characteristics storage, organization and reuse of knowledge problem, it provides a semantic foundation for further research of ontology-based knowledge sharing, case representation, case retrieval and case-based reasoning. (2) Ontology-based person-post matching case representation of enterprise knowledge workers. Based on domain ontology building post competence characteristics, has been clear about the enterprise knowledge workers hillock matching case representation of the problem description, a description, solution three components, defines the post match case o