초록 데이터를 활용한 한국무역연구분야 키워드, 문서간 관계성 모델링

초록

Purpose : This study analyzes the research trends through English keyword and abstracts of 11,774 papers listed in Korean academic journals, which were selected from 16 different academic journals in trade research. The 16 selected journals were searched using large-class category as social science, and middle-class category as trade. Research design, data, methodology : The 11,774 articles were analyzed by Keyword Frequency Analysis, Word Embedding (Word2vec), t-SNE and Scattertext using Python. Results : The analysis results are as follows : First, as a result of analyzing the frequency of 49,293 keyword of 11,774 articles, “FTA (420 times)”, “China (280 times)”, and “e-Trade (237 times)” were top 3 keyword. Second, with Word Embeddig, we presented the words that showed a high correlation (cosine similarity) with the Top 5 keyword, and visually illustrated the research connection of top 60 keyword among the nouns of the whole abstracts. Finally, Scattertext was used to visually indicate which keyword were frequently used in studies from 2002 to 2010, and from 2011 to 2019. Conclusions : This study is the first study to derive implications for academic development through keyword analysis in English abstract by applying the big data approach to the field of trade research, focusing on domestic journal articles. Through these research results, it can be concluded that more in-depth approaches - such as analyzing the development of trade research from diverse aspects by turning text materials into vectorized data or researching trade research of various countries - are required in the future.

키워드