面向期刊论文的知识挖掘研究

发布时间：2018-05-07 03:27

本文选题：科学计量学 + 可视化　；参考：《河北大学》2014年硕士论文

【摘要】：随着科技的发展，人们在各领域的研究越发深入，科技文献的数量也随之增多，在为我们提供了更多新方法和新知识的同时也带来了新的问题：海量的文献资源与人们的阅读时间和利用能力形成了矛盾，阻碍了人们对科技文献的利用。本文面向期刊论文进行知识挖掘研究：其一有助于科技管理人员把握期刊整体结构及发展现状，从而帮助相关人员针对期刊建设做出合理的决策；其二从大量的科技文献中发现主流主题及其演化情况，可以帮助科研人员减少查阅科技文献的时间，便于对科技文献进行整理和总结，把握学科前沿方向，对科研人员的研究具有一定的指导意义。本文通过对2000-2011年间某期刊论文进行统计和分析，呈现了该期刊在研究队伍、地域分布、基金资助及其主题内容等情况，以揭示该期刊研究的趋势和特征。本文的主要工作和贡献如下： 1.研究分析了科学计量学、主题模型的基本理论和方法，并应用科学计量学方法和主题模型对期刊论文进行了知识挖掘。 2.应用科学计量学方法对期刊论文的作者、机构、基金资助等内容进行了挖掘，实验结果可以帮助相关人员针对期刊建设做出合理的决策，为期刊的评价提供了量化依据。 3.把主题模型方法引入期刊论文挖掘，挖掘出了主题分布、主题变化等知识。本文给出“主题共现”方法来描述主题情况，相对于传统的关键字共现的方式，具有描述清晰简洁、语义明确的特点。实验表明，主题模型方法应用于中文期刊论文知识挖掘，具有较好的效果。可以帮助科研人员减少查阅时间，便于对科技文献进行整理和总结，正确把握学科前沿方向。
[Abstract]:With the development of science and technology, people's research in various fields has become more and more in-depth, and the number of scientific and technological documents has also increased. It not only provides us with more new methods and new knowledge, but also brings new problems: the huge amount of literature resources and people's reading time and ability to use the contradiction, hindered the use of scientific and technological documents. This paper aims at the research of knowledge mining for journal papers: firstly, it helps the scientific and technological managers to grasp the whole structure and the current development of periodicals, thus helping the relevant personnel to make reasonable decisions on the construction of periodicals; Second, finding the mainstream themes and their evolution from a large number of scientific and technological documents can help researchers to reduce the time for consulting scientific and technological documents, facilitate the sorting and summing up of scientific and technological documents, and grasp the frontier direction of science and technology. It has certain guiding significance to the research of scientific research personnel. Based on the statistics and analysis of a periodical paper from 2000 to 2011, this paper presents the research team, regional distribution, fund support and its subject content in order to reveal the trend and characteristics of the periodical research. The main work and contributions of this paper are as follows: 1. This paper studies and analyzes the basic theories and methods of science metrology and subject model, and applies the scientific metrology method and subject model to mine the knowledge of journal papers. 2. The scientific metrology method is applied to excavate the contents of the journal papers such as authors, institutions, funds, etc. The experimental results can help the relevant personnel to make reasonable decisions on the periodical construction and provide the quantitative basis for the evaluation of the periodicals. 3. The method of topic model is introduced into the mining of journal papers, and the knowledge of topic distribution and theme change is excavated. This paper presents a method of "theme co-occurrence" to describe the topic situation. Compared with the traditional keyword co-occurrence method, it has the characteristics of clear description and clear semantics. The experimental results show that the method of topic model is effective in the knowledge mining of Chinese periodicals. It can help the scientific research personnel to reduce the time of consulting, facilitate the collation and summary of the scientific and technological documents, and correctly grasp the frontier direction of the subject.
【学位授予单位】：河北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP391.1

【共引文献】