토픽 모델링과 N-gram을 활용한 한국 현대시 연구 동향 분석 -2010년~2023년 WOS 및 SCOPUS DB를 중심으로- (A Trend Analysis of Korean Contemporary Poetry Research Using Topic Modeling and N-gram Analysis : Focusing on WOS and SCOPUS DB from 2010 to 2023)

서지정보

· 발행기관 : 고려대학교 국제한국언어문화연구소
· 수록지 정보 : Journal of korean Culture / 61권 / 169 ~ 201페이지
· 저자명 : 강은진, 이용재

초록

이 연구는 2010년부터 2023년까지 SCI급 저널에 게재된 한국 현대시 분야의 학술논문의 연구 동향을 토픽 모델링과 N-gram 방식으로 분석하였다. 토픽 모델링을 통해 한국 현대시 연구 분야의 핵심 주제군을 도출하고, 이 결과를 N-gram 분석을 통해 키워드들이 문서 내에서 연속적으로 사용되는 패턴 각 키워드들의 연결망을 파악하고자 하였다. 이와 같은 분석 과정을 통해 총3개의 최적 토픽 그룹을 추출하였으며 각 토픽 그룹들에 속한 키워드들의 사용 패턴과 관계망을 도출할 수 있었다. 토픽1은 전쟁, 분단, 지역성, 현대성에 관련된 주제어들이 도출되었으며 토픽2에는 일본, 식민, 탈식민, 번역과 관련된 주제어들이 포함되었다. 마지막으로 토픽3은 문화, 정치, 민중, 증언과 같은 다양한 주제어들이 도출되었다. 이러한 토픽들을 N-gram으로 분석하여 각 키워드들이 맺는 관계망과 패턴을 시각적으로 도식화하여 한국 현대시 연구의 국제 동향을 파악할 수 있는 기초 자료를 축적할 수 있었다. 이를 통해 가장 최신의 한국 현대시에 관한 국제 연구가 한국의 역사를 기반으로 한 식민지배, 전쟁, 분단 등을 중요 인식 토대를 형성하고 있음을 알 수 있었다. 또한 이 연구를 통해 한국 작가 및 시인에 대한 관심도가 높은 반면 구체적으로 연구되는 한국 시인은 제한적이라는 사실도 드러났다.

영어초록

The purpose of this study was to analyze and identify the latest trends in international research on modern Korean poetry in a quantitative and scientific way. To this end, we analyzed and selected meaningful data from academic papers on modern Korean poetry published in SCI-level academic journals. We mainly focused on topic modeling and N-gram analysis based on the title, abstract, author, and publisher of the document. We obtained quite significant results, but there were also limitations due to the mechanical analysis method. During the preprocessing process, we mechanically refined the data, but it was difficult to completely exclude the possibility that some of the keywords or tokens extracted from the N-gram analysis would hinder the consistency of the analysis.
In addition, at the stage of selecting the analysis target, there were quite a few cases where the search result data deviated from the subject, so we had to examine and remove them one by one. Since the analysis target was limited to SCI-level academic journal papers published after 2010, the size of the confirmed data was reduced as expected after this selection process. However, the purpose of this study was to identify the latest research trends, so we did not expand the search target period and performed the analysis. In the future, if we expand the analysis target and collect more abundant data, we will be able to identify research trends by country and language, and we will be able to obtain detailed analysis results based on the connection between specific universities and researchers and specific keywords.
Through topic modeling, three research topics on modern Korean poetry were extracted: topic 1 can be compressed into war and division, regionality, and aesthetics; topic 2 can be categorized into colonialism, decolonization, historicity, and connectivity; and topic 3 is about culture, politics, testimony, and resistance. These results show that research on modern Korean poetry is closely related to Korean history and pays attention to the relationship with neighboring countries and Western countries such as the United States. It also reflects the high interest in Korean poets. However, it was found that the actual active poets are limited to a limited range, such as Yun Dong-ju, Kim So-wol, and Kim Su-young. This can be said to be a result that provides a reference point for the direction of international academic activities related to the study of modern Korean poetry.
After topic modeling analysis, we used N-gram analysis to see how the main keywords are related to other words in the research paper. The important results extracted from the network of keywords belonging to topic 1 show that research on modern Korean poetry is being dealt with in the context of the connection with East Asian literature, the relationship with other cultures based on regionality, and the problem of diaspora or immigrants derived from the experience of war. In addition, it was confirmed that the keyword group of Korean modern literature writers or poets is mentioned together with feminism or feminist.
The results of N-gram analysis for the keywords of topic 2 show that research on modern Korean poetry is still being dealt with in the category of Japan-Korea relations and colonial literature. This perspective shows that modern Korean poetry is being highlighted in the context of comparing Japanese literature and Korean literature, clarifying colonial and postcolonial characteristics, and defining modernity and modernity.
In addition, the keywords of topic 3 are mentioned in a series of contexts such as culture, politics, testimony, and resistance. It was confirmed that they are forming a network of words related to people's resistance and testimony, and language, such as the situation of division between North and South Korea, minjung literature, generational characteristics, environment, and memory. Some of these keywords form groups with Kim So-wol and Kim Su-young, indicating that they find the popular archetype of modern Korean poetry in these two poets. The results of the analysis of topic 3 show that the research topics on modern Korean poetry are becoming broader, but at the same time, it shows the limitation that international research on modern Korean poetry has not yet been expanded to Korean poetry after 2000.

참고자료

· 없음