|
|
|
| The Image Changes of Chinese Characters in European and American Films: A Word-frequency Study Based on Text Mining Technology |
| Lu Ruixi1,2, Mo Chou3 |
1.College of English, Beijing Language and Culture University, Beijing 100083, China
2.College of International Studies and Education, Tongren University, Tongren 554300, China
3.Utrecht Institute of Linguistics OTS, Utrecht University, 3512 JK Utrecht, The Netherlands |
|
|
|
|
Abstract The image of film role is an important tool for shaping ideology and spreading the values of the mainstream culture of a nation. As China’s international status has undergone a series of changes in the past century, the images of Chinese characters in European and American films have also experienced multiple changes. This paper attempts to explore the attitudes of the European and American film communities towards Chinese culture in different historical periods by analyzing the image changes of Chinese characters in European and American films from 1919 to 2020, and sheds light on the topics that Chinese film industry needs to work on in the process of getting globalized. Different from the traditional theory-driven film studies, the paper adopts an empirical research method. Altogether 58 films produced by European and American countries were selected and the summary texts about the film plots were collected. Word frequencies in the collected texts were analyzed through the NLPIR big data search and mining platform. The results of the text mining investigation were analyzed and explained through the ranking of frequent words and through close reading of the summary texts. It is found out that the images of Chinese characters in different historical periods are not constant, but the changes in the images of Chinese characters show continuity across temporal periods. The attitude of European and American film producers has changed from imposing subjective assumptions on Chinese culture to gradually accepting facts of China, from fearing or disdaining Chinese people to showing understanding and affirmation. However, in the current mainstream European and American films, generally speaking Chinese people and Chinese culture are still abstract symbols, lacking solid cultural connotations. The paper also makes methodological innovations by combining qualitative and quantitative research approaches. The rapid development of the Internet technology and natural language processing has enabled digital text mining technology to be widely used in the studies of the traditional humanities. On the one hand, the massive digital texts available in the era of big data provide raw data for quantitative researches. On the other hand, natural language processing technology based on computer algorithms can quickly explore a variety of indicators that reflect cultural features. Quantitative research based on digital text mining technology is an important supplement to qualitative research relying on the manual work of experts. Quantitative research makes it possible to observe and analyze various cultural behaviors in human society more extensively and thoroughly. By collecting a large number of digital texts from the Internet and using natural language processing platforms such as NLPIR for text mining, the authors are able to conduct word frequency analysis within a short period of time so as to outline and summarize the cultural features of many films in different historical periods within the past century. As an objective indicator of linguistic patterns reflected by the texts, word frequency makes up for the limitation of subjectivity of the expert-oriented qualitative research dominant in the studies of the humanities. It should be noted that although the method of word frequency analysis has the advantage of being objective and efficient, it is found that the advantages are highly dependent on the quality of the original texts. If the selection of the original texts is biased, not sufficiently representative, or has too little information, then the high-frequency words that are extracted from texts can hardly reflect the cultural contours of historical periods. At the micro-level, the paper has reflected on topics such as part-of-speech selection, text segmentation, and representativeness of high-frequency words, and thus has pointed out the necessity of combining computer-based quantitative research and expert-oriented qualitative research. The paper has also summarized detailed empirical experience of conducting algorithm-based word frequency research, which contributes to future cultural studies based on text mining technology.
|
|
Received: 11 April 2021
|
|
|
|
| 1 |
Graham S.， Milligan I. & Weingart S.， Exploring Big Historical Data： The Historian’s Macroscope， London： Imperial College Press， 2016.2 张华平、商建云： 《NLPIR-Parser：大数据语义智能分析平台》，《语料库语言学》2019年第1期，第87-104页。3 杨林： 《浅谈中国形象在美国电影中的变迁》，《新闻世界》2015年第1期，第126-127页。4 高兴梅： 《好莱坞电影里中国形象的变迁》，《南京政治学院学报》2015年第6期，第74-81页。5 美］爱德华·W.萨义德： 《东方学》，王宇根译，北京：生活·读书·新知三联书店，1999年。6 马来西亚］马修·D.约翰逊、张华： 《战地之旅：早期外国电影中的中国形象构建》，《当代电影》2018年第1期，第71-79页。7 陈晨： 《1980年美国电影中的中国形象和华人形象研究》，《视界观》2019年第14期，第98，100页。8 雷茜： 《好莱坞电影中的中国形象和华人形象的嬗变》，《电影文学》2010年第10期，第41-42页。9 罗维： 《论电影跨文化传播中的“中国形象”建构》，《中南大学学报（社会科学版）》2016年第4期，第174-179页。
|
|
|
|
2021, Vol. 51 
