Font Size: a A A

Research On Knowledge Mining For Journal Papers

Posted on:2015-04-13Degree:MasterType:Thesis
Country:ChinaCandidate:Q BianFull Text:PDF
GTID:2298330422969996Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the development of science and technology, the amount of scientific literature in thenetwork grew rapidly. Lots of scientific literature gives us new methodologies and newknowledge. However they also bring about new issues, such as there is a contradictionbetween time spent in readings and the amount of scientific literature, so it is unable to makefull use of the scientific literature resource. Researching on knowledge mining for journalpapers will help the science and technology managers to grasp the overall structure and thedevelopment condition of journal, so that they can make reasonable decisions. Finding out thehot topics and the evolvement of topics will help researchers to carry out sorting and analysisand summary the scientific literature and select the research targets scientifically andreasonably. It has a certain guiding significance for researchers.The researchers, geographic distribution, foundation and topics of the journal articles areevaluated in this paper to reveal the research trends and features. The main contributions ofthe dissertation are as follows:1. This paper studies the relevant basic theories and methods and uses these theories tomine knowledge.2. The article uses the scientometrics as the method for researching and finding thecharacteristics and regularities of the journal. This result provides quantitative reference to thejournal evaluation.3. The topic models to the knowledge mining for journal papers are introduced and thispaper proposes co-topic analysis. This method can show the relation between different topics,it is better than the co-word analysis. The result shows the fact that topic model applied toknowledge mining for journal has achieved good results.
Keywords/Search Tags:Scientometrics, Visualization, Text mining, Topic model, LDA
PDF Full Text Request
Related items