网络安全领域研究主题师承紧密度计算与分析

网络安全领域研究主题师承紧密度计算与分析(开题报告,论文14000字)
摘 要
本文计算与分析网络安全领域的“研究主题师承紧密度”可以反应导师与学生之间研究方向的紧密程度，师承紧密度的高低也反映着这个领域的发展状况。本文从统计分析和内容分析两个方面研究网络安全领域内2010年至2018年间的研究方向和研究趋势以及师承紧密度与高校论文发表数量、时间的关系，并对网络安全领域做学术生态探索。
本研究采集了2010年至2018年来网络安全领域内博士导师以及博士生的基本信息，包括他们的姓名以及导师与学生之间的对应关系；还采集了博士导师与博士生在这9年间所发表的论文信息，包括论文题目、关键词、摘要以及发表时间、发表机构等。由上述信息构成研究所需要的原始文本语料库。随后，经过严格的文本数据清洗得到适合模型训练的本文语料并统计每位学者的论文信息中的前20个高频词，由此构成每位学者的研究主题词。使用自然语言处理中获取词向量工具Word2vec并以所有学者论文信息作为其输入信息训练模型，由模型得到每个主题词的向量。学者向量来源于该学者的研究主题词向量。最后，由不同学者向量之间的距离代表他们之间的研究主题师承紧密度，由此计算出每一对博士导师与博士生之间的研究主题师承紧密度。
由本研究对师承紧密度的计算与分析可以得到结论：从统计分析的结果来看，2010年到2012年师承紧密度集中在大于0.6的阶段，而2014年往后开始逐渐出现学生与导师研究方向较疏远的情况；从内容分析的基于领域内容视角来看这9年间网络安全领域在2014年转向更多新主题，集中关注系统、网络协议、无线以及物联网安全，并结合机器学习以及算法等技术发展网络安全技术；从高校角度来看，几所在网络安全领域大有发展的高校，不仅在论文发表数量上位居前列，在师承紧密度上也都低于0.3，而在所有发表论文数量位居前10的高校中师承紧密度也都低于0.4，处于一个较低的水平，这说明师承紧密度与高校该学科发展关系密切。
关键词：网络安全；师承关系；word2vec；数据可视化
 
The calculation and analysis of tightness of mentors' and PhD students' research topics in the field of network security
Abstract
In this paper, the calculation and analysis of the "research topic teacher closely degree" in the field of network security can reflect the research direction between the tutor and the student closely degree, the degree of teacher closely degree also reflects the development of this field. This paper studies the research direction and trend between 2010 and 2018 in the field of network security from the perspective of statistical analysis and content analysis, as well as the relationship between the closeness of teachers and the number and time of university paper publication.
This study collected basic information of doctoral supervisors and doctoral students in the field of network security from 2010 to 2018, including their names and the corresponding relationship between tutors and students. In addition, information about papers published by doctoral supervisors and doctoral students during the 9 years was collected, including title, key words, abstract, publication time, publication institution and so on. The above information constitutes the original text corpus required by the research. Then, after strict text data cleaning, the corpus suitable for model training was obtained, and the top 20 high-frequency words in each scholar's paper information were counted, so as to constitute the research subject words of each scholar. The word vector tool Word2vec in natural language processing is used to get the vector of each subject word. The scholar vector is derived from the research subject word vector of the scholar. Finally, the distance between different scholar vectors represents the research subject affinity between them, and then the research subject affinity between each pair of doctoral supervisors and doctoral students is calculated.
Conclusion can be drawn from the calculation and analysis of the teacher engagement degree in this study: according to the results of statistical analysis, the teacher engagement degree from 2010 to 2012 was concentrated at a stage greater than 0.6, while from 2014 onwards, students gradually became estranged from the research direction of their tutors. From the perspective of content analysis based on domain content, the network security field focused on system, network protocol, wireless and Internet of things security in the past 9 years, and developed network security technology by combining machine learning, algorithm and other technologies. From the perspective of colleges and universities, colleges and universities, a few in the great development in the field of network security is not only the number on the top in the paper, on the tightness are below 0.3 in shicheng, and published in all publications in the top 10 universities inherited the tightness is below 0.4, is in a lower level, suggesting that inherited the tightness and closely related to the subject development of colleges and universities.
Keywords: network security, mentoring relationship, word2vec, data visualization
 
目录
摘要    I
Abstract    Ⅲ
第一章  绪论    1
1.1  研究对象    1
1.2  研究目的与意义    1
1.3  研究现状    2
1.4  研究方法    3
第二章  构建网络安全领域博士师生论文语料库    4
2.1  确定师生特征信息    4
2.2  获取论文信息    5
2.2.1  人工获取基本信息    5
2.2.2  使用网络爬虫自动化获取信息    5
2.3  原始语料文本预处理    8
2.4  本章小结    9
第三章 师承紧密度计算    10
3.1  获取研究主题    10
3.2  文本特征提取    11
3.2.1  TfidfVectorizer与Word2Vec方法比较    11
3.2.2  从Word2vec到Scholar2vec    12
3.3  余弦相似度计算    13
3.4  本章小结    14
第四章 师承紧密度分析    15
4.1  统计分析    15
4.2  内容分析    17
4.2.1  基于领域研究内容视角    17
4.2.2  基于时间视角    18
4.2.3  基于高校视角    19
4.3  本章小结    21
第五章 总结    22
5.1  网络安全领域研究主题师承紧密度分析结论    22
5.2  本文研究的展望    22
参考文献    24
致 谢    26