Research Achievement Orientied Literature Management And Mining

Scientific literature plays an important role in the scientific field. Researchers read scientific literatures to catch up the latest research progress made by others, and they write scientific literatures to report their original empirical and theoretical work. Therefore, researches on literature have attracted much attention.Most of those works focus on the content of literatures and on the relation between literatures and their authors, but few of them consider literatures as the achievements of research work. This thesis organizes and analyses literatures as the research achievement made by some researcher or some research group, and get three major results as follows:First, this thesis proposes a framework to automaticlly get attributes of literatures, such as title, authors, and publication. This framework extracts information from the PDF file of a piece of literature by machine learning algorithms, and extracts information from web sites by human roles.Second, this thesis proposes a framework to automaticlly get and update the evaluating indicators of literatures from web sites. For a piece of literature, the framework can get and update the number of times being cited, the publications the citing literature published, and the famous researchers citing the literature.Third, this thesis designs and implements a literature management system called PubMS, which implements the first two frameworks into a real system. Futhermore, PubMS provides batch formatting of literatures for writing reports.