【作者】 滕广青；

【导师】 毕强；

【作者基本信息】 吉林大学 ， 情报学， 2012， 博士

【摘要】 20世纪90年代以来，随着计算机和网络技术的发展，图书馆的概念逐渐超越了人们传统意识当中那幢钢筋水泥的建筑物。以数字化存储与呈现、网络化检索与获取为特征的数字图书馆，成为网络数字时代集知识存储、获取、传播、交流等多功能为一体的知识集散中心。数字图书馆的相关研究日渐成为现代图书情报学的重要分支，并在多学科理论与技术的支撑下获得了相对独立的发展空间与地位。这一期间，学术界对数字图书馆的理论探索和实践开展掀起了研究的热潮，特别是在针对数字化馆藏资源的建设方面取得了不菲的成绩。然而，随着近年来Web2.0的兴起及语义网络的发展，开放式语义网络环境下的数字图书馆知识管理面临着许多新的问题与挑战，其中数字图书馆的知识组织是这些问题当中最为基础而且突出的核心焦点。如何把握数字图书馆知识组织的基本脉络和发展方向，融合多学科的理论与技术，通过准确分析和深入研究探寻语义网络环境下的数字图书馆知识组织的内在机理与演变规律，构建开放式语义网络环境下数字图书馆知识组织的理论框架与发展蓝图，促进中国数字化知识产业的快速成长与发展，已成为一项亟待解决的重大课题。有鉴于此，论文以国际数据分析领域中在概念化知识处理方面最新的研究成果——形式概念分析（FCA）与概念格（CL）理论为基础,从知识的概念化、语义化、形式化的视角，对数字图书馆知识组织展开研究。致力于基于形式概念分析的概念格理论与技术构建数字图书馆知识组织的模型框架，创新数字图书馆知识组织的技术策略，推进数字图书馆在开放式语义网络环境下的实践进程。具体研究包括：（1）从国内外形式概念分析与概念格理论在概念化知识处理领域的应用和数字图书馆知识组织相关理论研究进展的文献调研入手，基于调研资料进行细致的梳理和分析。重点探讨了基于形式概念分析的概念格理论与技术在数字图书馆各类知识组织与服务中的应用，总结出适合本项目的核心理论、必备方法和关键技术。并通过对知识组织体系演进路径的分析，探讨了当前数字图书馆知识组织的困境与发展趋势，明确界定了论文的研究思路与逻辑起点。（2）通过对基于概念格的数字图书馆用户知识组织的研究，从数字图书馆用户内隐知识挖掘与萃取的层面，探讨了将形式概念分析与概念格理论应用于数字图书馆知识组织中用户知识需求识别、概念认知分析、行为偏好挖掘等方面的功能与优势。并进一步将这种优势延伸到数字图书馆新兴的社群分类法（Folksonomy）和开放存取（Open Access）领域，阐述了形式概念分析与概念格相关理论与技术对开放式、分布式语义网络环境下的数字图书馆知识组织活动的理论支撑与技术保障，论证了基于形式概念分析的概念格理论与技术应用于数字图书馆知识组织的科学性和有效性。（3）构建了相关领域知识的概念格。针对数字图书馆特定领域内的相关知识（包括内隐知识与外显知识），进行知识的语义化、概念化研究。依据知识概念的对象与属性创建形式背景，采用形式概念分析的技术构建领域知识概念格，实现领域知识的概念化、形式化描述。并基于领域知识概念格对特定领域的相关知识结构进行分析和呈现，揭示了知识之间的层级结构与关联关系，挖掘发现潜在的隐含规则与模式。（4）提出了“本体与概念格互补融合”的数字图书馆知识组织的技术策略。从哲学到图书情报学、从认识论到本体论、从内隐知识到外显知识，结合开放式语义网络发展的实际，对数字图书馆知识组织的理论与基于形式概念分析的概念格理论进行了全面、系统、深入的分析与研究。并基于以上研究提出了“本体概念格”互补融合的数字图书馆知识组织的技术策略。（5）基于概念格实现了领域知识本体的构建。针对数字图书馆中主题词表与文本两大主要类型的知识资源进行分析，借助概念格的并叠置运算构建数字图书馆异构资源概念格。通过领域知识概念格与领域本体之间的映射规则，在领域知识概念格的基础上构建领域本体。基于概念格的本体构建，提高了本体构建的形式化、自动化程度，大大降低了人为的干扰因素，实现了“概念格本体”的正向促进。（6）建立了基于概念格的跨本体映射。对相关领域本体进行基于概念格的逆向解析，将标准词典中的相关同义词、上位词等关系嵌入本体中的知识概念及层级结构，将不同的异构本体转换为相应的概念格，并提取有效的知识概念。进而基于概念的属性与对象，创造性地提出了基于概念格的“对象-属性相似度（OAS）”法。利用该方法对通过逆向解析获得的概念进行相似度计算，并根据设定的阈值提取满足阈值要求的异构本体间的相似“概念对”，基于相似“概念对”建立了跨异构本体的映射关系，完成了“本体概念格”的逆向解析。（7）构建了基于概念格的多本体协同知识地图。从哲学与情报学角度对人类知识的应然状态与实然表现进行系统的分析与总结，对当前知识的本体化进程与作为知识组织终极表现的知识地图进行系统的分析与论述，为构建基于概念格的多本体协同系统奠定理论基础。选取现实中特定领域范围内典型的、具有代表性的成熟领域本体，基于概念格建立跨本体映射，并据此构建多本体协同知识地图，以“拼图”形式实现了更大范围的知识组织。论文基于形式概念分析与概念格相关理论与技术，以解决开放式、分布式语义网络环境下数字图书馆知识组织相关问题为逻辑起点，在梳理、总结形式概念分析与概念格在相关领域中的应用的基础上，对知识组织及其体系结构的演进进行了分析与归纳，综合运用知识组织理论、本体理论、概念格理论、语义学理论，全面、深入、系统地研究探索数字图书馆知识组织的技术策略和框架模型。构建了相关领域知识概念格，提出了本体与概念格互补融合的数字图书馆知识组织技术策略，并据此实现了数字图书馆异构资源领域本体构建和跨本体映射，并在此基础上构建了多本体协同知识地图。论文的理论价值在于，丰富和完善了数字图书馆知识组织的理论体系与方法体系，促进现代语义网络环境中数字图书馆知识组织理论与方法的变革。对用户内隐知识的挖掘与发现是数字图书馆知识组织理论新的生长点，本体与概念格互补融合为知识描述、知识组织、知识导航、知识构建提供了新的理论支撑和解决方案。论文的现实意义在于，柔性化的数字图书馆知识组织体系是当前语义网络环境下数字图书馆实践的现实需求，基于概念格的数字图书馆知识组织，通过构建多本体知识地图，概念化、语义化、形式化地呈现和揭示知识结构与关联，提高数字图书馆用户知识获取与利用的效率，更好地释放和发挥数字图书馆在现代语义网络环境中的潜能和价值。更多还原

【Abstract】 Since the 1990s, with the development of computer and network technology, theconcept of the library is gradually moving beyond that big steel and concretebuildings in which people’s sense of tradition. Digital library with the features,storage and presentation of digital, network-based retrieval and access, is amulti-function distribution and integration of knowledge center which is set ofknowledge storage, access, dissemination, exchange in networked digital era. Thestudy of digital library is becoming an important branch of modern library andinformation science, and acquires relatively independent development space andposition in support of multi-disciplinary theory and technology. During this period,the academic community set off a craze of the study on the theoretical explorationand practice of digital library, in particular, many outstanding achievements haveobtained in resources building of digital library.However, with the rise of of Web 2.0 and the development of the semantic webin recent years, the knowledge management of digital library is facing many newproblems and challenges in open semantic web environment. Digital libraryknowledge organization is the most basic and salient core focus in these issues. Howto grasp the basic context and direction of development of the digital libraryknowledge organization, with the integration of multidisciplinary theories andtechniques, through accurate analysis and in-depth studies to explore the internalmechanism and evolution law of the digital library knowledge organization in thesemantic network environment, to build a theoretical framework and blueprint for thedevelopment of digital library knowledge organization in the open semantics webenvironment, to promote rapid growth and development of the digitalknowledge-based industries in China, has become a major issue to be solved.In view of this, with the base of formal concept analysis (FCA) and conceptlattice (CL), which are the latest research in the conceptualize knowledge processingin the field of data analysis of international academic community, from theconceptualize, semantic, formal perspective of knowledge, the digital libraryknowledge organization is studied. This dissertation commits to construct the modelframework of digital library knowledge organization with concept lattice theory andtechnology based on formal concept analysis, to innovate the technology strategy of the digital library knowledge organization, to promote the process of practice ofdigital library in the open semantic web environment. Specific studies include:(1) Starting from the literature survey of the progress of application of formalconcept analysis and concept lattice in the conceptualize knowledge processing athome and abroad, and theoretical studies of digital library knowledge organization,detailed outlines and analysis are executed based on survey data. It focus on theconcept lattice theory and technique based on formal concept analysis for varioustypes of knowledge organizations and services in the digital library, sums up the coretheory, the necessary methods and key technologies for this project. And through theanalysis of the evolutionary path of knowledge organization system to explore theplight and development trend of the current digital library knowledge organization,the research route and logical starting point of this study are defined clearly.(2) By the research of users’knowledge organization of digital library based onconcept lattice, from the mining and extraction on users’implicit knowledge ofdigital library, the functions and advantages on identification of users knowledgeneeds, analysis of concept cognition, mining of behavioral preferences and otherfeatures are explored, in which formal concept analysis and concept lattice theoriesare applied to the digital library knowledge organization. And these advantages areextended to emerging field of digital library such as community taxonomy(Folksonomy) and open access (OA), theoretical support and technical guarantee onthe knowledge organization of digital library based on formal concept analysis andconcept lattice theory and technology are elaborated in open distributed semanticnetwork environment, the scientific nature and effectiveness of the digital libraryknowledge organization based on formal concept analysis and concept lattice theoryand techniques are demonstrated.(3) The concept lattices of the knowledge in related fields are constructed. Theresearch on semantic and conceptualize knowledge within specific areas of digitallibrary (including the implicit knowledge and explicit knowledge) is executed. Basedon the objects and attributes of knowledge concept, the formal context is created, theconcept lattice of domain knowledge is constructed with formal concept analysistechniques, the formal conceptualization description of domain knowledge isachieved. The structures of knowledge in specific areas are analyzed and presentedbased on the concept lattice of domain knowledge, the hierarchy and the relationshipbetween knowledge are displayed, the potential implicit rules and patterns are minedand discovered.(4)‘Complementary Fusion between Ontology and Concept Lattice’as a noveltechnology strategy of digital library knowledge organization is proposed. From thephilosophy to the library and Information science, from epistemology to ontology,from implicit knowledge to explicit knowledge, combined with the actual development of open semantic network, the digital library knowledge organizationtheory and the theory of concept lattice based on formal concept analysis areanalyzed and researched comprehensively, systematically and deeply. Based on theabove study, complementary integration on‘ontology concept lattice’astechnology strategy of digital library knowledge organization is proposed.(5) The construction of domain knowledge ontology is realized based on conceptlattice. Two main types of knowledge resources in digital library, thesaurus and text,are analyzed, and the concept lattice of heterogeneous resources is constructed byapposition and overlap operations. Through the mapping rules between the domainknowledge concept lattice and domain ontology, domain ontology is built on the basisof the domain knowledge concept lattice. Building ontology by concept latticeimproves the degree of formal and automation of ontology construction, greatlyreduces the man-made interference factors of ontology construction, and achieves thepositive promoting of‘concept lattice ontology’.(6) The mappings crossing heterogeneous ontologies are established based onconcept lattice. The reverse resolving of related domain ontology is executed basedon concept lattice, the relationships of synonyms, hyponyms and other relations instandard dictionary are embedded in concepts of knowledge and hierarchy ofontology, different heterogeneous ontologies are converted into the correspondingconcept lattices, and the effective concepts of knowledge are extracted. Then basedon the properties and objects of concept, the‘Object-Attribute Similarity (OAS)’method based on concept lattice is presented creatively. Using this method tocalculate the similarity of the concepts which are acquired by reverse resolving, and‘pair of concepts’across heterogeneous ontologies are extracted under therequirements to meet the threshold, the mapping relationships across heterogeneousontologies are established based on similar‘pair of concepts’, the reverse resolutionof‘ontology concept lattice’is achieved.(7) The knowledge-map of multi-ontologies collaboration is constructed basedon concept lattice. From philosophy and information science point of view, itanalyzes and summarizes the state of ought-to-be and performance of to-be of humanknowledge systematically, analyzes and discourses the current ontology process ofknowledge and knowledge-map as an ultimate performance of knowledgeorganization, builds the theoretical foundation for construction of multi-ontologiescollaboration system based on concept lattice. It selects the typical representativemature domain ontologies in specific areas, across ontologies mappings areestablished based on concept lattice, and accordingly to build a knowledge-map ofmulti-ontologies collaboration, in order to achieve a wider range knowledgeorganization with the form of‘Jigsaw Puzzle’. Based on formal concept analysis and concept lattice theory and technology,solving related issues of digital library knowledge organization in open distributedsemantic network environment as a logical starting point, on the basis of combing andsummarizing the applications of formal concept analysis and concept lattice in relateddomains, this dissertation sums up and analyzes the evolution of knowledgeorganization and its architecture, studies and explores the technology strategy andframework model of digital library knowledge organization with integrated usage ofknowledge organization theory, body theory, concept lattice theory, semantic theory,comprehensively, deeply and systematically. It builds the concept lattice ofknowledge in related domains, complementary fusion between ontology and conceptlattice as a novel technology strategy of digital library knowledge organization isproposed. Accordingly, the construction of domain ontology on heterogeneousresources in digital library and across ontologies mappings are achieved, andknowledge-map of multi-ontologies collaboration is established on this basis.The theoretical value of this dissertation is to enrich and improve the theoreticalsystem and methodology of digital library knowledge organization, to promote thetransformation and revolution on theories and methods of digital library knowledgeorganization in modern semantic network environment. Mining and discovering theusers’implicit knowledge are new growth points of the digital library knowledgeorganization theory, complementary fusion between ontology and concept latticeprovides a new theoretical support and solutions for knowledge description,knowledge organization, knowledge navigation, knowledge construction. Thepractical significance of this dissertation is to enhance flexible nature of digitallibrary knowledge organization system for meeting the practical needs of the digitallibrary practice in the current semantic web environment. By establishing theknowledge-map of multi-ontologies collaboration, the digital library knowledgeorganization based on concept lattice, presents and discovers the structure andassociation of knowledge conceptually, semantically and formally, increases theefficiency of knowledge access and utilization of digital library users, better releasesand produces the potential and value of digital libraries in the modern semanticnetwork environment.更多还原