高低程度网络依赖者的线上语言特性差异分析

发布时间：2018-05-10 08:29

本文选题：网络依赖 + 文本分析　；参考：《华中师范大学》2017年硕士论文

【摘要】：随着互联网的普及与飞速发展,对网络的过度使用现象得到了学界的普遍关注。许多研究者从认知功能缺陷、情绪功能障碍等不同角度对网络成瘾行为进行了研究,普遍认为网络成瘾个体具有对现实生活适应不良的表现特征。目前对网络成瘾行为的研究,主要是通过观察个体线下的行为表现来分析并甄别其网络使用程度是否异常,缺乏对网络成瘾者线上行为特征的实证研究。本研究通过诱发个体的特定情绪状态、选取不同类型的网络线索,结合文本分析技术来考察高、低程度网络依赖个体的在线语言用词特征,为今后实现在线动态地甄别网络成瘾行为提供参考依据。研究一通过诱发个体的特定情绪诱发,分别考察高、低程度的网络依赖者在积极-消极情绪状态下、面对积极-消极阅读文本时其在线语言用词的差异。结果表明,个体的情绪状态对高程度网络依赖个体的线上语言用词的影响甚微,阅读文本的情绪属性对其影响显著。同时,网络活动可以改善个体的负面情绪体验,特别是对高程度网络依赖的使用者而言,他们在进行网络活动可以得到更大程度的情绪调节。研究二通过模拟社交媒体平台常见的图文阅读模式,分别进行积极和消极文本评论实验,考察了高、低程度网络依赖者面对积极、消极阅读文本与不同的配图类型时的线上语言用词线索。结果表明,配图类型能够影响高程度网络依赖者的线上语言用词;尤其是卡通表情这一与互联网高度关联的线索,高程度网络依赖的个体对卡通表情配图更加熟悉,若阅读文本结合了卡通表情配图,则会阻碍他们对积极文本信息的再次评估和唤起对消极文本信息的重新评价。同时卡通表情配图的存在,会降低高程度网络依赖的个体对文本信息的负面感受,加重低程度网络依赖的个体对消极文本信息的负面情绪体验。综上所述,高程度网络依赖个体的线上行为表现与现实生活中有所差异。今后研究可利用更高级的数据挖掘技术,通过先分析主题文本的情绪属性与配图类型、后探索基于主题的评论文本语言特征,来实现在线动态地甄别网络成瘾行为。
[Abstract]:With the popularity and rapid development of the Internet, the phenomenon of excessive use of the network has received widespread attention. Many researchers have studied the behavior of Internet addiction from different angles such as cognitive impairment emotional dysfunction and so on. It is generally believed that the individual of Internet addiction has the characteristics of poor adaptation to real life. At present, the research on Internet addiction behavior is mainly to analyze and identify whether the degree of internet use is abnormal or not by observing the behavior of individuals below the line, and lack of empirical research on the online behavior characteristics of Internet addicts. In this study, the specific emotional state of individuals was induced, different types of network cues were selected, and text analysis techniques were used to investigate the online language features of individuals with high and low degree of network dependence. To provide a reference for the realization of online dynamic discrimination of Internet addiction behavior in the future. In the first study, we investigated the differences of online language words of high and low degree network dependent people in the positive and negative emotional state in the face of the positive and negative reading text by inducing the specific emotion of the individual. The results show that the emotional state of the individual has little effect on the online language words of the individual with high degree of network dependence, while the emotional attribute of the reading text has a significant effect on it. At the same time, Internet activities can improve the negative emotional experience of individuals, especially for users with high degree of network dependence, they can get a greater degree of emotional adjustment in the process of online activities. In the second study, by simulating the common pattern of text reading on social media platform, the positive and negative text review experiments were carried out respectively, and the high and low degree of network dependence were investigated to face the positive. Negative reading text with different map types when online language use cues. The results show that the type of matching images can affect the online language words of the high degree network dependent people, especially cartoon expression, which is a highly related clue to the Internet, and the individuals with high degree of network dependence are more familiar with the cartoon facial expression matching. If the reading text combines cartoon facial expression matching, it will prevent them from re-evaluating the positive text information and evoking the re-evaluation of the negative text information. At the same time, the existence of cartoon facial expression matching images will reduce the negative feelings of individuals with high degree of network dependence on text information, and aggravate the negative emotional experience of individuals with low degree of network dependence on negative text information. To sum up, the online behavior of individuals with high degree of network dependence is different from that of real life. In the future, more advanced data mining techniques can be used to dynamically identify online addictive behaviors by analyzing the emotional attributes and matching graph types of topic texts, and then exploring the language features of subject based comments.
【学位授予单位】：华中师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：B845

【参考文献】