디지털 기법을 활용한 시조 분석 ― 코로나-19 제재 창작 시조를 중심으로

초록

The world is changing everyday. The phrase "the 4th Industrial Revolution" is no longer unfamiliar. The same is true of the term artificial intelligence, used in several fields of our society. In fact, the world has always been changing; however, it seems that the changing world has never given serious consideration to classical researchers. These days, the strong wind of change does not distinguish between countries, generations, or regions. The same holds for literary research, and classical science is no exception. This does not necessarily mean that everything has to change. In the case of classical science, traditional research still exists and needs to continue that way, as the classics were not created in a day. Nonetheless, it is necessary to determine a new methodology for classical studies. This work attempts to address this need by analyzing literary works using digital technology, including artificial intelligence, as a tool. For the study, first, college students were required to create a predecessor by focusing on COVID-19. Therefore, 105 analysis targets could be secured, although certainly, 105 works are insufficient for analysis. Considering that it takes a substantial amount of time for individual researchers to analyze all works one at a time, this paper is meaningful in that it explores a methodology to analyze many works in a short duration. The above works were analyzed using digital techniques. First, morpheme was analyzed using KoNLPy's Mecab. Next, word embedding was performed using the gensim word2vec function, to derive words with high cosine similarity to 'Corona', and the results were analyzed. In addition, considering the fact that the meaning of addition and subtraction between word vectors can be calculated, the perceptions of college students were examined with respect to the end of COVID-19through the calculation of a meaning in the form of 'I-Corona'. Throughword2vec, 100 words that students frequently used in the process of creating the sijo were extracted, visualized, and expressed in two-dimensional space. The analysis conducted on the above works provided results similar to those conducted by myself on several of the works, although I admit that the target works are insufficient for generalization. Therefore, for this work to be recognized for its significance as a new methodology in research, a sufficient number of target works should be secured to analyze and develop the capabilities necessary to comprehensively understand the digital analysis results.

키워드