Research On Knowledge Discovery Model Of Regional Government Weibo From The Perspective Of Archival Science

With the development of technology in information production and dissemination,Chinese local government and users have a more utilitarian and efficient interaction platform called Sina Weibo by which local government could improve their ability in government information service while users could learn some important events rapidly and get lots of knowledges from these events.At the present time,users' reading habit has a conspicuous transformation from full text reading to fragmentation reading,thus,the knowledge map constructed by information element or knowledge element have a significant role in husbanding users' time and enhancing users' efficiency of knowledge absorption.In order to uphold the provenance among information elements or knowledge elements,and to ensure the quality of knowledge organization and knowledge discovery,this thesis proposes a regional government Weibo knowledge discovery model from the perspective of Archival Science which aims to catalyze government information resources with records and archive attributes becoming knowledge resources,facilitates user's knowledge assimilation and utilization,and improves the level of government information knowledge service of local state organs.Firstly,this thesis explains the meaning of some intrinsic concepts like regional government Weibo,knowledge discovery of archives as well as social media records,makes the literature review and summary of research related to this model and relevant concepts,mentions the crucial theories and technologies adopted by this thesis.Secondly,this thesis tries to make a fusion between the theories of LIS and Archival Science,proposes an Integration Theory in Archival Science based on the Principle of Provenance and Records Continuum Theory and described by six tuple to adapting the situation of knowledge discovery.Thirdly,building the regional government Weibo knowledge discovery model from the perspective of Archival Science based on the natural language processing technology and knowledge organization and discovery technology like text cluster by LDA model and applied ontology of regional government Weibo knowledge discovery as well as establishing the procedure of data mining and knowledge discovery.Ultimately,this thesis selects Xi'an in Shaan Xi province as an individual of the concept “region”,collects the text from Sina Weibo pages by web crawler and classifies the text into different projects and events.Then through the extraction of Information elements,this thesis builds the knowledge maps as the result of knowledge discovery.After the evaluation of the result,this thesis draws a conclusion that this model is more feasible.There is a space for improvements of the regional government Weibo knowledge discovery model this thesis built,such as the rationality of the integrated theory mentioned above,the ability of knowledge organization and discovery as well as the reliability and validity of the empirical research for this model.Enhancing the dynamic representation and reasoning of knowledge and the intelligent of the whole model based on deep learning is the main research direction for the next step of this thesis.