신문 기사 속의 직업명사와 관련어

Abstract

In this paper, we have examined the aspects of occupation nouns and related words in news articles. Our investigation is based on the [Trends 21] Corpus (Kim et al., 2010; Kang & Kim, 2011), which is a morphologically annotated 400 million word corpus of Koreans newspaper texts. It is a collection of Korean newspaper texts covering most of the topics in print. The corpus is composed of four major Korean national newspapers issued from the year 2000 to 2009. We adopt a network analysis using co-occurring words with target words, i.e., occupation nouns. The extraction of co-occurring words is based on the statistical information (t-score) about the co-occurrences of words. The co-occurring words are regarded as not only semantically related but also mentally related to target words; thus we call the co-occurring words in a given context related words. A network is a graph-based structure made up of nodes and links. We adopt co-occurrence network analysis, which views lexical relationships in terms of nodes (representing individual words within the network) and connections or links (which represent relationships between the words). Pajek (Nooy, Mrvar & Vladimir, 2005) is used to visualize the relationship between the occupation nouns and their related terms. In addition, we have identified peoples‘ interest in social, cultural and economic issues that are portrayed in Korean newspaper articles.