由于具有强大的知识共享和重用的能力，ontology在智能信息检索、语义web 等众多领域具有重要作用，但是目前很多ontology都是通过人工方式构建完成的，采用人工构建ontology的方式需要花费漫长的时间和巨大的成本，因此ontology的自动或半自动构建成了一个亟待解决的问题。当前ontology自动构建技术仍然很不成熟，构建高质量的本体还是需要专家花费大量时间来开发与维护。本文在对国内外本体自动构建相关研究进行全面分析和总结的基础上，提出了分布协作式的ontology构建方法。首先通过学习领域文本来自动建立领域初始本体，然后为用户提供一个分布式的平台通过协作来完成本体构建。这种构建过程的理念和web2.0的理念相同，都是充分挖掘个人的积极性参与到整个体系的建设中来。通过这种方式，广大用户所贡献的智慧能够弥补机器自动构建的不足。自动构建和用户协作构建两者结合能够兼顾本体构建的成本与质量。因此，本文的核心工作主要围绕以下两个方面展开：1．初始本体的自动构建，主要围绕概念获取和关系获取展开。本文对面向中文文本的本体概念和概念间语义关系自动获取的方法、工具、过程等关键技术进行了分析和总结，特别介绍了基于统计和词法规则的本体概念获取方法以及基于关系句法模式的关系获取方法，在此基础上进行了实例分析验证。2．本体进化模型分析设计。本体进化过程一方面要确保用户的需求能够有效的表达，另一方面要保证用户的进化请求执行后本体仍然保持一致性。因此，本文提出了一个三阶段的本体进化模型，并详细介绍了进化表示、进化分析、进化实施这三个阶段的工作。在本体进化模型的基础上，本文设计了本体协作构建原型系统。

Since ontology has the good ability of data sharing and reusing, it will be helpful in intelligent information retrieval, semantic web and many other domains. Now lots of ontology is mainly built manually, which takes lots of time and its cost is huge. As a result, automatic ontology building is an urgent problem to be resolved.Now the automatic ontology building technology is still immature.To ensure quality, domain experts still need to spend a lot of time to develop and maintain the ontology. Based on comprehensively analyzed theories about automatic ontology building, this paper proposes a novel method for distributed cooperative ontology building. Firstly, we could build an initial ontology for a certain domain. Secondly, place the ontology in a distributed cooperative building environment so that mutiple users could edit the ontology. Its idea is same to the web 2.0, which is arousing users' enthusiasm of content contribution. In this way, the wisdom of users has contributed to the task which automatic ontology buiding can't complete. Automatic ontology buiding combining with collaborative ontology buiding will balance the cost and quality better. Key work of this paper is mainly about two aspects as follows:1. Automatic building an initial ontology, mainly about domain specific concepts extraction and semantic relation learning. This paper analyzes the method, tool, process and other key techniques of domain concept extraction and semantic relations learning from chinese domain text, especially introduces domain concept extraction method based on statistic and rule matching and semantic relations learning based on relation syntactic patterns. In the end, some methods are certified through case analysis.2. Ontology evolution model analysis and design. Users' requirement should be captured and executed correctly in the process of ontology evolution. This paper proposed a three stages ontology evolution model including requirement capturing, sematic of change, evolution implementation. Based on the ontology evolution model, this paper designs a collaborative ontology engineering system in the end.