基于用戶行為的情感分析技術(shù)的研究
[Abstract]:With the application of web2.0 and the rapid popularization of mobile network and mobile terminal, social network has become an important part of people's daily life. The rise of social networks brings opportunities for emotional analysis. The emotional analysis of social networks is helpful for service providers to optimize services, accurate marketing; help users to improve their experience, efficient consumption; help government regulators to monitor and guide public opinion, and so on. At present, many scholars begin with the processing and analysis of short texts, or use emotional dictionaries or extract semantic features for emotional analysis. In the exploration and study of human behavior, it can be found that behavior and emotion affect each other and reflect each other. Therefore, this paper is different from the approach from the perspective of text semantics, from the point of view of the behavior of network users to explore and analyze the relationship between user behavior and their emotional tendencies and laws. And construct the classifier of emotion tendency based on user behavior. In the social network, the user scale is huge, but its behavior is normative, easy to divide and obtain, which provides convenience for the study of network user behavior. Taking Sina Weibo as an example, this paper studies the influence and function of user behavior in the affective analysis of social network from the characteristics of user behavior. Firstly, the current research status of emotional analysis and user behavior is studied, and the theoretical basis and research situation of social network are systematically introduced. Then, according to the general process of feature-based affective analysis, data capture, preprocessing and affective tagging are carried out in the data preparation stage, and in the feature extraction stage, the user behavior features are extracted. Through statistical analysis and association rule mining, the relationship between user's behavior and emotional tendency is studied. Finally, the classification technology and Bayesian and decision tree classifier are studied deeply and systematically. Combined with the actual situation, the naive Bayesian algorithm and C4.5 algorithm are used to construct the emotion classification model based on user behavior characteristics. It is verified by experiments. Through a series of studies and experiments, this paper proves that there is a certain relationship and law between the user's behavior and their emotional tendency in social network, thus laying a foundation for further research.
【學(xué)位授予單位】:北京郵電大學(xué)
【學(xué)位級別】:碩士
【學(xué)位授予年份】:2016
【分類號】:TP391.1
【參考文獻(xiàn)】
相關(guān)期刊論文 前10條
1 楊善林;王佳佳;代寶;李旭軍;姜元春;劉業(yè)政;;在線社交網(wǎng)絡(luò)用戶行為研究現(xiàn)狀與展望[J];中國科學(xué)院院刊;2015年02期
2 陽愛民;周詠梅;周劍峰;;中文微博語料情感類別自動標(biāo)注方法[J];計(jì)算機(jī)應(yīng)用;2014年08期
3 賴清楠;馬皓;宋維佳;李婷婷;蔣廣學(xué);張蓓;;高校BBS與微博的用戶社交行為特征分析[J];通信學(xué)報;2013年S2期
4 何躍;王迪;張麗麗;;基于關(guān)聯(lián)規(guī)則的微博主題搜索策略研究[J];情報雜志;2013年06期
5 周勝臣;瞿文婷;石英子;施詢之;孫韻辰;;中文微博情感分析研究綜述[J];計(jì)算機(jī)應(yīng)用與軟件;2013年03期
6 謝麗星;周明;孫茂松;;基于層次結(jié)構(gòu)的多策略中文微博情感分析和特征抽取[J];中文信息學(xué)報;2012年01期
7 何黎;何躍;霍葉青;;微博用戶特征分析和核心用戶挖掘[J];情報理論與實(shí)踐;2011年11期
8 楊成明;;微博客用戶行為特征實(shí)證分析[J];圖書情報工作;2011年12期
9 趙妍妍;秦兵;劉挺;;文本情感分析[J];軟件學(xué)報;2010年08期
10 王曉光;;微博客用戶行為特征與關(guān)系特征實(shí)證分析——以“新浪微博”為例[J];圖書情報工作;2010年14期
相關(guān)會議論文 前1條
1 姚天f ;彭思崴;;漢語主客觀文本分類方法的研究[A];第三屆全國信息檢索與內(nèi)容安全學(xué)術(shù)會議論文集[C];2007年
相關(guān)博士學(xué)位論文 前2條
1 肖云鵬;在線社會網(wǎng)絡(luò)用戶行為模型與應(yīng)用算法研究[D];北京郵電大學(xué);2013年
2 易蘭麗;基于人類動力學(xué)的微博用戶行為統(tǒng)計(jì)特征分析與建模研究[D];北京郵電大學(xué);2012年
相關(guān)碩士學(xué)位論文 前2條
1 高巖;微博情感分析的相關(guān)技術(shù)研究[D];華北電力大學(xué);2014年
2 楊琳;基于社交網(wǎng)絡(luò)的用戶行為分析及預(yù)測[D];西安郵電大學(xué);2013年
,本文編號:2159188
本文鏈接:http://sikaile.net/kejilunwen/ruanjiangongchenglunwen/2159188.html