【作者】 安海岗；

【导师】 雷涯邻；

【作者基本信息】 中国地质大学（北京） ， 管理科学与工程， 2010， 博士

【摘要】 随着网络技术的飞速发展,互联网已成为广大群众学习、生活和娱乐不可缺少的一部分,当人们登陆互联网时,各种新闻事件到处可见,其中大多的新闻信息都是以文本形式随机地“堆”在用户面前,使得用户无法快速地从大量随机信息中获取自己所需信息。本文在对信息可视化和复杂网络相关理论进行研究的基础之上,结合新闻要素本身的特点和用户的信息需求,提出了专题新闻文本集信息可视化理论模型,然后以新浪网四川汶川大地震灾后重建专题新闻文本集为实例对其进行实证研究。本文完成了如下研究工作:(1)在对新闻要素、信息可视化、复杂网络、文本挖掘、广义合作网络等理论进行综合分析研究的基础之上,结合用户的信息需求,以card提出的信息可视化理论模型和时间墙模型为依据,构建了专题新闻文本集信息可视化理论模型,并对专题新闻文本集信息可视化模型的要素进行了形式化表示。(2)界定了新闻文本关键词粒度的内涵,阐述了新闻文本关键词粒度的划分原则;叙述了新闻文本关键词的提取原则与方法;把新闻文本集中的每个关键词看作是一个项目,运用广义合作网络理论界定了关键词的关联规则。(3)提取了新浪网“汶川地震灾后重建”专题新闻文本集的关键词,然后根据各类关键词的粒度,应用概念分层理论对关键词进行了处理,构造了关键词库,最后根据关键词关联规则,利用EXCEL构造了关联矩阵。(4)根据关联矩阵,运用Ucinet,对汶川地震灾后重建专题新闻文本集的关联进行了可视化,对新闻文本总体情况和各类关键词交叉关联情况进行了分析。(5)分析了汶川地震灾后重建专题新闻文本集文本数量变化情况;提出了新闻事件“关注度”的概念,对各事件关键词关注度逐月变化情况和主要事件关注度演化情况进行了分析。更多还原

【Abstract】 With the fast development of network technology, Internet network has become an important part of people’s learning, life and entertainment. Accessing to the internet network, people will find many kinds of news, which is listed in the form of texts mostly, and couldn’t find the useful information to them quickly from the random information. After studying the theory of information visualization and complex network,the paper proposes a information visualization model of special news texts set, associating the news elements’characteristic and users’information needs. Then, the paper visualizes the Wenchuan earthquake reconstruction news texts set on Sina as an example. The paper achieves the research results listed as below: First, after studying the theory of news’elements, information visualization, complex network, texts’digging, corporative broad network and etc., the paper proposes a information visualization model of special news texts set, associating users’information needs and basing on the theory of Stuart K. Card and time wall model. Then, the paper formalizes the model’s elements.Second, the paper explains the concept of the granularity of news texts’keywords, and narrates the principles and methods of news texts’keywords’selection. Regarding every keyword of the news’texts as a project, the paper illuminates the association rules of keywords by using the corporative broad network.Third, the paper distills the keywords of Wenchuan earthquake reconstruction news texts set on Sina. Then, it processes them by using concept hierarchy theory on the basis of the keywords’granularity and forms the keywords’database. Finally, it constructs the association matrix by using Excel on the basis of association rules.Fourth, the paper visualizes the association of Wenchuan earthquake reconstruction news texts set by using Ucinet on the basis of association matrix and analyses these results.Fifth, the paper analyses the change of Wenchuan earthquake reconstruction news texts set. Then, it proposes the concept of event attention-getting degree, analyses the change of all events keywords’attention-getting degree and of primary events keywords’attention-getting degree.更多还原