CASIA OpenIR  > 毕业生  > 博士学位论文
基于WWW的综合集成研讨厅体系的模型研究
其他题名Modeling the Hall for Workshop of Metasynthetic Engineering Based on WWW
崔霞
学位类型工学博士
导师戴汝为
2003-06-01
学位授予单位中国科学院研究生院
学位授予地点中国科学院自动化研究所
学位专业模式识别与智能系统
关键词开放复杂巨系统 综合集成法 综合集成研讨厅体系 有向属性图 交互 复杂网络 复杂性 系统思考 学习型组织 对话 链接结构分析方法 决策支持 知识管理 人工智能 Ocgs Meta-synthetic Engineering Metasynthesis The Hall For Workshop Of metasynThetic Engineering Attributed Directed Graph
摘要在综合集成研讨厅体系理论框架的指导下,本文致力于研究如何应用信息 技术具体化处理与开放复杂巨系统有关问题的综合集成研讨厅体系。本文研究 了综合集成研讨厅体系建模的关键因素,分析了处理与开放复杂巨系统有关问 题的综合集成研讨厅体系与WWW的结构相似性,建立了包含人的因素、基于 WWW的、人机结合的处理与OCGS有关问题的HWME有向属性图模型,开发 了HWME的理解工具,并进一步剖析了综合集成法是开放复杂巨系统的唯一方 法。这为理解和描述开放复杂巨系统开辟了思路,为开放复杂巨系统和综合集 成法的研究从理论到具体化开辟了途径,更进一步完善和发展了从定性到定量 的综合集成方法,提高了集成专家意见的智能化水平。主要工作概括为以下几 个方面: 1.提出了面向问题的特殊专家-WWW和广义专家群体的概念。信息情报资 源的高效化,是综合集成研讨厅体系实现的关键因素之一。围绕着将要研讨的 问题,通过WWW,获得针对问题的“权威”网页和“中心”网页,这些网页 的集合及作为人的群体,构成广义专家群体。研讨厅是广义专家群体充分利用 信息技术进行交互作用的场所。这为有效激活海量网络信息资源,将其拿来为 我所用,提供了新的有效方法。 2.在综合集成研讨厅体系中,广义专家群体关于开放复杂问题的动态互过 程,导致了整体层次上充满相互响应的综合集成研讨厅体系网络链接结构。考 虑专家的发言,类似于WWW的一个网页,而专家是网页的创建者,这样一来 广义专家群体的研讨过程,或者进一步说,广义专家群体互动形成的网络链接 结构和WWW的网络链接结构具有结构相似性。 3.在HWME与WWW具有结构相似性的基础上,借鉴模式识别领域描述 复杂模式的方法,建立了处理与开放复杂巨系统有关问题的基于WWW的综合 集成研讨厅体系的有向属性图模型,为综合集成研讨厅体系的研究开辟了建模 和实现途径。其中,以专家个体为节点的有向属性图模型为研究专家系列属性 的动态实现和交互环境提供了思路。而以专家发言为节点的有向属性图模型, 节点属性和边连接属性综合集成了所有参与者的理性与非理性知识,也集成了 亿万web用户关于问题的群体见解,在整体层次上呈现集体优势。以专家发言为节点的HWME有向属性图模型,为理解综合集成研讨厅体系的综合优势提供 了准确有效的途径。 4.建立了综合集成研讨厅体系链接结构分析方法,为综合集成研讨厅体系 的整体优势:群体智慧提供了思路和理解工具。
其他摘要With the theoretical framework of metasynthetic engineering, i.e. the hall for workshop of metasynthetic engineering, this paper is devoted to study the implementation of the hall for workshop of metasynthetic engineering based on WWW. With understanding the core of ME, an attributed directed graph model of HWME is established to solve the problem of OCGS after analyzing the structure analogy between WWW and I-1WME. Building on this, an algorithm for understanding and inferring the HWME is developed. And, we further understand the unique methodology of metasynthesis for OCGS. In a word, this paper breaks a new path for modeling and understanding OCGS and HWME. More details are explained in the following. 1) Tremendous information resources in internet are exploited fully in HWME. Internet is not only an information resource but also a carrier of collective intelligence of the thousands of millions of web users, so it can be regarded as a special expert orienting on some certain topic in HWME, i.e. the counterpart of the human expert. Moreover, the especial expert-internet and the group of experts constitute the generalized group of experts in HWME. In addition, intranet provides the room for interactions among the generalized experts. 2) The process of effective interactions among the generalized experts leads to the network structure of inter-responded environment of HWME. Comparing the network of HWME to that of WWW, the discourse of one expert each time is similar to a Web page, and the inter-response in the content of discourse is similar to a hyper-linkage in the content of web page, moreover, the growth of HWME and WWW are driven by the inter-response or hyper-linkage resulted from human's behavior, so there is an structure analogy between them. 3) Inspired by the pattern description as attributed graph in pattern recognition, the network structure of HWME is characterized by an attributed directed graph based on the structure analogy between HWME and WWW. In one graph, each vertex represents each the generalized expert with basic attributes, discourse attributes and social attributes, and edge represents kinds of relationships among the experts. This graph is applied to study the interaction environment and the dynamic implement of expert's attributes. In the other graph, each vertex with discourse attributes represents a discourse of one expert each time, and each directed edge with relation attributes describes the response embedded in the discourse content. The attributes in the graph have embodied the experience and intelligence of experts and millions of thousands of web users. And the global organization (or network) of I-IWME Meta-synthesizes the rational and non-rational knowledge of all participants, which evolves over the interactions among the generalized experts. 4) The concept of the tools understanding and inferring WWW is applied to understand HWME, and a set of algori
馆藏号XWLW742
其他标识符742
语种中文
文献类型学位论文
条目标识符http://ir.ia.ac.cn/handle/173211/5772
专题毕业生_博士学位论文
推荐引用方式
GB/T 7714
崔霞. 基于WWW的综合集成研讨厅体系的模型研究[D]. 中国科学院自动化研究所. 中国科学院研究生院,2003.
条目包含的文件
条目无相关文件。
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[崔霞]的文章
百度学术
百度学术中相似的文章
[崔霞]的文章
必应学术
必应学术中相似的文章
[崔霞]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。