An Analysis of Sentiment Orientation in 1984 Based on Sentiment Dictionary 基于情感词典的《一九八四》情感倾向分析文献综述

 2023-05-05 09:05

文献综述

1. Introduction1.1 Research backgroundNineteen Eighty-Four (1984) is the last masterpiece of the British novelist George Orwell during his lifetime, and it is also his most famous work. It enjoys a pivotal position in the English literary world. In this work Orwell portrays a suffocating world of horror. In an imaginary future society, dictators take power as their ultimate goal. Humanity is completely stifled by power, freedom is completely deprived, and ideas are severely suppressed. The peoples lives have fallen into extreme poverty, and the lives of the lower classes have become a tedious cycle. What Orwell wants to tell the reader is that although peoples living environment is absurd, people have no choice but to mock self-existence in a pessimistic and desperate situation, and strongly deny absurd existence. The novel is very attractive. In addition to the imaginary world of totalitarianism, the authors protagonist constantly tries to understand the past, finds love, and participates in fraternal parties in the process of political oppression. In 1984, Orwell illustrates the negative consequences of the main characters positive struggle against totalitarianism. There is a lot of emotional vocabulary in the novel. Can we carry out a quantitative analysis on it instead of a literary or a pragmatic one. Based on sentiment dictionary, this paper will explore the sentiment orientation in 1984.1.2 Research purposeThis paper attempts to start with the emotional attributes of positive words and negative words, using word frequency and part-of-speech analysis software to analyze the emotional orientation of the protagonists at different stages of the story, and further explore the spirit of resistance and the flash of humanity under political oppression in the novel. 1.3 Organization of the studyThe first chapter is the introduction, expounding the research background, research purpose and the structure of this paper. The second chapter is literature review of related concepts and theoretical foundations, and previous researches on 1984. The third chapter describes the process of sorting and analyzing the 1984 text data, including the preprocessing and importing of text data, the description of the part of speech and word frequency analysis process of the text data, and paving the way for the next step of data analysis. The research results are shown in Chapter 4. The last chapter is the conclusion and outlook, which provides the main findings, limitations and suggestions for further research.2. Literature Review2.1 Previous studies on 1984Researchers have studied 1984 from different perspectives, such as analysis of the political background or the main characters, or the use of ambiguous language. Chen (2019) believes that based on Orwells questioning and reflection on the workings of power in the superstructure, the depiction of the workings of power and its political tragedies in 1984 is strongly metaphorical in its reality. Liu (2014) believes that in the novel, it is through Winston a minor character like Winston in his quest for affection love and truth, and in his reflection on history and reflections on history and reality, the author expresses his thoughts on human nature and the dignity of human beings. The novel is a reflection on human nature and human dignity.Nineteen Eighty-Four depicts a future world under totalitarian control. Li (2016) argues that by extinguishing human emotions and struggle through totalitarian means, the government transforms people into a paranoid uniformity, ultimately creating a horrible world in which humanity is extinguished, emotions are absent and freedom and democracy are gone under the control of an ideological tyranny. A world in which the ideological tyranny of the mind has destroyed humanity, emotions are absent, and freedom and democracy are non-existent.George Orwell achieved a great deal of ambiguous language in 1984. Wu (2013) argues that Winston breaks the taboo of Big Brothers language and that the ambiguous language used in everyday conversation reflects his sense of rebellion against political oppression.2.2 Previous studies on sentiment analysis2.2.1 Sentiment dictionariesForeign sentiment analysis research first started in the 1990s, earlier than domestic ones. Stone et al. of Harvard University constructed a sentiment dictionary GI (General Inquirer) in 2006. They collected 1915 masculine words and 2293 feminine words and labelled each word according to the attributes of affective polarity, affective intensity and discourse. A set of semantic disambiguation rules was defined for each word by giving each word three labels and listing the different meanings contained in each word to distinguish between positive and negative polarity of words with the same meaning and under the same part of speech. Hu et al. provided and constructed a Sentiment Lexicon in 2004, which has been continuously updated to a sentiment lexicon of 6800 words. The Italian Information Science Institute (ISTI) developed SentiWordNet in 2009, based on the existing WordNet, which gives different sentiment scores for each synonym under different parts of speech. The latest version of SentiWordNet 3.0 contains a total of 117,658 records, each consisting of six attributes such as part of speech, record number, positive score, negative score, synonym entry name and comment. Wilson et al. A MPQA dictionary containing 8000 sentiment words was constructed in 2012. The above sentiment dictionaries are the most widely used English sentiment dictionaries.2.2.2 Sentiment analysisWord sentiment analysis is an important basis for studying text sentiment analysis. At present, most of the research on sentiment analysis at the word level only starts from one dimension and judges the polarity of sentiment. In order to quantitatively analyze the sentiment polarity of words, a real number in the numerical interval [-X, X] is usually used as the sentiment polarity of words, and if it is less than 0, it means derogatory meaning, greater than 0 means positive meaning, and its absolute value represents the strength of the degree of polarity. Words with emotional polarity are mainly nouns, verbs, adjectives and adverbs. Among them, except that the sentiment polarity of some words can be obtained by looking up the dictionary, the sentiment polarity of the remaining words not included in the sentiment dictionary cannot be directly obtained. There are two main methods to obtain sentiment polarity of words, one is based on sentiment dictionary and one is based on annotated corpus.The acquisition of emotional polarity information of English words is mainly carried out on the basis of WordNet and General Inquirer (GI), and the acquisition of emotional polarity information of Chinese words is mainly carried out on the basis of HowNet and Synonyms.Turney et al (2003) used the Pointwise Mutual Information (PMI) of candidate sentiment words and benchmark sentiment words to calculate the semantic correlation between words 1 and 2. In this application paper, the author discriminates the sentiment polarity of words by calculating the PMI values of the extracted phrases and the seed sentiment words. The advantage of this method is that it can use smaller-scale seed sentiment words to determine the sentiment polarity of other words, but the disadvantage is that it is very dependent on seed sentiment words to a certain extent.Based on annotated corpus, Wiebe (2003) obtained adjective sentiment words through the word clustering method of similarity distribution. However, in this application paper, this method is limited to adjectives and ignores sentiment words of other parts of speech. In order to solve the problem of part-of-speech limitation, Riloff (2013) wrote in their report: Based on the artificially selected seed words of a template, an iterative method was used to obtain sentiment words of noun part-of-speech. Lou (2005) used part-of-speech tagging on the text to find the dependencies between words according to the part-of-speech, and calculate the emotional polarity of words. The biggest advantage of the corpus-based method is that it is simple and convenient to use statistical methods, but the disadvantage is that the available emotional annotation corpus is limited. The corpus-based method for obtaining sentiment polarity of words is overly dependent on the distribution of sentiment words in the corpus, while the method for obtaining sentiment word polarity based on sentiment lexicon also relies heavily on the perfection of sentiment lexicon. Both options are deficient in each other. Therefore, it is particularly important whether to choose an emotional dictionary or a corpus. After comprehensive consideration, in this paper, the author chooses the sentiment dictionary as the research tool.2.3 Research gap The analysis of 1984 focuses on the pragmatic perspective of the novel, which uses a large number of emotional words and tends to change the protagonists emotions in a way that is worth exploring using an emotive dictionary. The authors sentiment analysis on 1984 is based on ready-made sentiment analysis tools. Since sentiment dictionaries are rarely used for novel analysis, and are mostly used for phrase-level or sentence-level sentiment analysis, this research is only in the preliminary exploratory stage.ReferencesChen, X., Shang, J., Zhao, Y., 346.刘德喜（2016），情感词扩展对微博情感分类性能影响的实验分析，《小型微型计算机系统》，37（05）：957-965。

王科、夏睿（2016），情感词典自动构建方法综述，《自动化学报》，42（04）：495-511。

武斌（2013），浅析《一九八四》中模糊语言的语用功能，《新余学院学报》，18(5):107-109刘陕（2014），集权主义之下人性的微光解读乔治奥威尔《1984》，《安徽文学》，377：63-64。

陈驰、张少清（2019），《1984》悲观政治图景的存在论意义，《湘南学院学报》，40(1)：62-65李超（2016），无知即力量论《1984》中的思想暴政，《河南工业大学学报(社会科学版)》，12(1):94-96徐坤、毕强（2019），次高频关键词的选择及在共词分析中的应用，《情报理论与实践》，42(5):148-152。

张华平（2019），NLPIR-Parser：大数据语义智能分析平台，《语料库语言学》，87-104。

张颖怡、章成志、陈果（2019），基于关键词的学术文本聚类集成研究，《情报学报》，38(8):860-871。

郑海鹏（2020），Spark 集群实现统计文档单词频次实例，《电脑知识与技术》，16（23）：33-34。

资料编号：[579226]

剩余内容已隐藏，您需要先支付 10元 才能查看该篇文章全部内容！立即支付

以上是毕业论文文献综述，课题毕业论文、任务书、外文翻译、程序设计、图纸设计等资料可联系客服协助查找。

您可能感兴趣的文章