신문 빅 데이터 기반의 단어 사용과 트렌드 분석: 신문의 명사 빈도 사용 패턴을 중심으로Language Use and Trend Analysis basedon Big Data Newspaper
- Other Titles
- Language Use and Trend Analysis basedon Big Data Newspaper
- Authors
- 김일환; 이도길
- Issue Date
- 2016
- Publisher
- 고려대학교 언어정보연구소
- Keywords
- Trends21 Corpus; Big Data; Frequency of Word Use; Patterns of Language Use; Co-occurrence Word; Trend
- Citation
- 언어정보, no.22, pp.41 - 62
- Indexed
- KCI
- Journal Title
- 언어정보
- Number
- 22
- Start Page
- 41
- End Page
- 62
- URI
- https://scholar.korea.ac.kr/handle/2021.sw.korea/132593
- ISSN
- 1226-8011
- Abstract
- This study aims at analyzing the macroscopic trends of language, social, and cultural changes by tracing aspects of language through the 'Trends 21' corpus, which is constitute of four major Korean newspaper articles (The Chosun Ilbo, The Dong-A Ilbo, The JoongAng Ilbo, The Hankyoreh) for the past 14 years from 2000 to 2013. For this purpose, we examined the frequency patterns of 6,000 common nouns that appeared more than 7,000 times in the corpus. As a result, we found out that nouns related to food, leisure, fashion, and consumers are consistently increasing. In contrast, we found the characteristics that many nouns related to 'performance' and 'art' have a decreasing frequency pattern. Also, 'construction' and 'book' related nouns show increasing trennd over a period of time and then have decreasing trend. On the other hand, for the social trends analysis we examined some nouns related to 'emotion', the prominent changes of co-occurrence words about 'happiness' and 'communication' have been observed. This study has a limitation in that we handle only common nouns with high frequency, but has its significance in that tries to capture social and cultural trends through changes in the specific language aspects.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - Associate Research Center > Research Institute of Korean Studies > 1. Journal Articles
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.