확장된 어휘 단위에 대한 연구 동향과 한국어의 기술

남길임(Nam Kilim) 한국사전학회 2011 한국사전학 Vol.- No.18

The purpose of this study is to survey the universal research trend of ELU(extended lexical unit) and to describe Korean ELU according to a corpus-driven approach. In chapter 2, the characteristics of the corpus-driven approach are described, as a new methodology of corpus linguistics; plain text policy, frequency & distribution, anti-categorization and register-based analysis. This paper shows there is no strict boundary between the corpus-driven approach and the corpus-based approach, and sometimes corpus-driven studies incorporate some corpus-based methods. It is possible to use a hybrids in actual practice. In chaper 3, using the methodology as mentioned earlier chapter, we discuss proper methodology for extracting n-gram based bundles in Korean, an agglutinative language. we focus on Korean 5-gram morphemic bundles including kes/ke, meaning "thing". As a result we extract Korean morphemic bundles across three registers and analyze the functions and characteristics of them.

『개방형 한국어 지식 대사전』의 생활용어 추출과 기술

남길임(Nam, Kilim),최준(Choi, Jun) 한국사전학회 2012 한국사전학 Vol.- No.19

The purpose of this study is to introduce the procedure and issues on the extraction of everyday vocabulary on Gaebanghyeong-Hangugeo-Jisik-Daesajeon(The Open Korean Knowledge Dictionary). We have worked for Compiling the dictionary since 2011 and are currently at the stage of the second year of this project. We have created a list of 10 thousand headwords and are going to add 15 thousand words which will provide for daily life vocabulary for Korean. We classify Korean everyday vocabulary into 4 types including 3 main types and 1 additional type according to the characteristic of the words. 3 main types are composed of frequent everyday words which have not been included in Pyojun-Gugeo-Daesajeon(The Standard Korean Dictionary), frequent derivatives combined with productive affix and frequent phrases composed of more than 2 words. The additional type consists of cultural words which have been used in Korean traditional life. These words are also valuable enough to preserve even though the frequency is relatively low and some words are rarely used nowadays. This study discusses a few procedures for the creation of Korean everyday vocabulary list according to the characteristic of each vocabulary type.

‘의미 단위’의 사전 기술을 위한 사전 사용자 인식 조사

남길임 ( Nam Kilim ) 연세대학교 언어정보연구원 2023 언어사실과 관점 Vol.59 No.-

본 연구의 목적은 사전에 누락된 미등재 의미 단위의 사전 등재에 대한 사용자의 인식를 조사하고, 의미 단위의 유형에 따른 사용자 인식의 차이를 분석하는 것이다. 이러한 연구의 전제는 이상적인 사전의 방향은 실제 언어가 어떻게 사용되고 있는지를 반영하는 것과 더불어 사용자들이 그들의 언어를 어떻게 인지하고 있는지도 함께 반영해야 한다는 것이다. 이를 위해 본 연구에서는 기존 사전 등재 여부, 고빈도 여부, 비속어/신어 여부 등의 변수를 포함한 총 6개 유형을 26개 구 단위를 선별하였고 이에 대해 사용자 1,000여 명의 설문 조사를 수행하였다. 이를 통해 알 수 있는 사실은 다음과 같다. 첫째, 사용자들이 비교적 높은 점수를 부여한 유형은 기존 등재어와 일반 구 단위, 상투 표현이며, 낮은 점수를 부여한 유형은 엔그램, 비속어, 신어이다. 이는 사용자들의 사전 등재 판단에 영향을 미치는 요인이 빈도보다는 형태적 완결성, 비속어나 신어 등 표제어의 사회적 가치 등 전통적인 사전의 관습이 더 지배적임을 나타낸다. 둘쨰, 설문조사의 두 집단 중, 의미단위 등재의 필요성에 대한 사전학적 원리에 대한 부가 설명은 사용자들의 사전 등재 가능성 판단 여부에 긍정적인 요인으로 작용하였다. 이러한 결과는 사전 자체 또는 사전학의 관련 정보가 표제어 등재 가치에 대한 사용자의 인식에 일정한 영향을 미친다는 것으로, 향후 의미 단위 사전의 기술과 사용자의 사전 활용 교육 방향성 등에 일정한 시사점을 가진다. This study aims to investigate users' perceptions of dictionarization of unlisted semantic units that are missing from the dictionary and to analyze differences in user perceptions based on the type of semantic unit. The premise of this study is that the ultimate dictionary should reflect how users perceive their language. For this study, we selected 26 phrase units of six types, applying variables such as [± already registered], [± high frequency], and [± slang or neologisms] to the classification, and surveyed more than 1,000 users. First, users gave relatively low scores to frequently used N-grams, profanity phrases, and phrase neologisms. This indicates that traditional dictionary conventions such as morphological completeness and the social value of headwords strongly influence users' judgments about dictionarization. Second, among the two groups in the survey, the explanation for Group 2, which was provided with an additional explanation of the lexicographic need for semantic units inclusion to dictionaries, positively influenced users' judgment. These results have a few implications for developing thoroughgoing semantic unit dictionaries and the direction of users' dictionary utilization education.

기사문과 댓글에 나타난 코로나 신어의 사용 양상과 사전학적 기술

남길임(Nam, Kilim),안진산(An, Jinsan),강현아(Kang, Hyeonah) 한국사전학회 2021 한국사전학 Vol.- No.38

본 연구는 다양한 웹 장르를 고려한 신어의 추출과 사전 기술의 가능성을 모색하기 위해, 기사문과 댓글을 중심으로 코로나19 말뭉치를 시기별로 구축하고, 코로나19 신어의 출현 빈도 및 사용 양상을 살펴본다. 연구 방법으로 기사문과 댓글 두 장르의 빈도와 더불어 고빈도 코로나 신어의 1차 공기어를 분석하고, 사례 연구로서 고빈도 코로나 신어 ‘K-방역’을 대상으로 2차 공기어 및 엔그램 분석을 수행한다. 이를 통해 일반적인 신어 추출과 기술의 주요 대상인 기사문과 댓글은 확연한 차이를 보이며, 대중의 언어로서 댓글 말뭉치가 가지는 가치와 한계를 확인할 수 있다. 특히 사전학적 관점에서 비전문가인 대중의 언어를 대표하는 댓글 자원이 거시구조와 미시구조 전반에서 사용자 언어 자원으로서 중요한 가치를 가진다는 것을 논의하고, 주요 사례로 ‘우한 폐렴’과 ‘K-방역’에 대한 사전 모형을 제시한다. This paper examines the occurrence frequencies and usage trends of COVID-19 neologisms in news articles and comments related to the pandemic in order to demonstrate the differences in neologism usage and evaluation across genre. In addition to frequency and primary collocate analyses, we carry out secondary collocate and n-gram analyses which provide a better understanding of the use context for neologisms, in a case study of the neologism ‘K-quarantine’. Results show that comment data prove invaluable for lexicographic description of neologisms, although experts need to decide to what extent the politically incorrectness of commenters’ language may be used in dictionaries. Finally, we propose a microstructural model for COVID-19 neologisms which integrates the findings of the study, taking the neologisms ‘Wuhan pneumonia’ and ‘K-quarantine’ as examples.

현대 한국어 [+사람] 신어의 사회 · 문화적 의미

남길임(Nam, Kilim),송현주(Song, Hyunju),최준(Choi, Jun) 한국사전학회 2015 한국사전학 Vol.- No.25

The purpose of this paper is to show that the study of neologisms is a key to understanding social changes and the new attitudes of a speech community. We will first introduce the ongoing investigation into Korean neologisms initiated by the Korean government and then analyze the neologisms found online in 2012 and 2013 through the Detector of Korean Neologisms, a web-based extracting software. Our research is twofold. Through a quantitative analysis of the typology of the origin of words composing neologisms as well as the subject areas in which they appear, we will see that the coinage of new words is more likely to involve foreign words rather than pure Korean or Sino-Korean words, but also to happen within certain domains, including those of society, telecommunications, economics and gaming. Through the qualitative analysis of the ‘[+person] neologisms’ in particular that constitute 258 out of the 1,000 new words detected, we will examine how these neologisms reflect the changing Korean society as well as the concerns of contemporary Koreans. This study represents a major step forward in the research into neologisms as, dismissing the researcher’s subjectivity, it examined objectively and thoroughly a great wealth of sources consisting of 139 online mass media in order to detect any word that came into use within the past two years. Above all, this paper intends to overcome the shortcomings of previous studies on neologisms that mainly focused on their grammatical characteristics by examining in depth their sociocultural aspects.

대규모 웹크롤링 말뭉치를 활용한 신어 사용 추이 조사의 현황과 쟁점

남길임(Nam, Kilim),이수진(Lee, Soojin),최준(Choi, Jun) 한국사전학회 2017 한국사전학 Vol.- No.29

This paper aims to examine diachronic aspects of neologisms that have been used consistently up until now by analyzing their frequency in a dynamic corpus. Section 2 presents the methodology adopted for investigating the changes in the use of neologisms using a largeᐨscale webᐨcrawled corpus. Section 3 explores the cases of a few neologisms that have been used consistently and discusses various aspects regarding these. The case studies allow for the detailed examination and discussion of the life cycle of neologisms, from their first appearance and evolution to their obsolescence, their disappearance, and even their revival. Then, a number of issues regarding the analysis of neologism use are looked into. The neologisms that are brought into focus in this study are those that have survived to date from their first appearance in 2005, 2006, and 2007 as investigated by The National Institute of the Korean Language. News articles from Naver News have been analyzed to study their development.

"-ㄹ 수 없-"의 의미와 담화 기능에 대한 말뭉치언어학적 분석

남길임 ( Kilim Nam ) 한국텍스트언어학회 2015 텍스트언어학 Vol.38 No.-

This study amis to investigate the meanings and the discourse functions of ‘-l su eobs-’ through an examination of the sequences in 21th Sejong Corpus. There have been few research studies of Korean ‘-l su eobs-’ even though the sequence have been considered to have similar function as ‘mos’ negation and play an important role in Korean negation. This study examines ‘-l su eobs-’ and its collocations, using n-gram and phrase-frame analysis. The results are summarized as follows: 1) Overall frequency of ‘-l su eobs-’ is higher than ‘mos’ negation in Korean written corpus. 2) ‘-l su eobs-’ are much more frequently used in double negation than ‘mos’ negation and the antonym ‘-l su iss-’, rather play an strong affirmation function. 3) There seems to be strong correlation between the meanings of ‘-l su eobs-’ and those of ‘-l su iss-’. Examination of 230 examples show that ‘-l su eobs-’ has 2 main meanings [possibility negation], [ability negation].

한국어 정형화된 표현의 분석 단위에 대한 연구

남길임(Nam, Kilim) 담화·인지언어학회 2013 담화와 인지 Vol.20 No.1

This study aims to investigate proper methodology for extracting formulaic expressions in Korean, an agglutinative language. For this purpose, two principles to extract formulaic expressions in an agglutinative language are discussed and applied. One is morpheme-based approach and the other is word-based approach. The former, 5-gram morphemic analysis proves that Korean speakers do use formulaic expressions in abundance. The latter, 3-gram word analysis shows some useful units Korean native speakers use in free conversations. The former approach result in much more types of Korean formulaic expressions than the latter, even though they have similar amount of information. Since Korean, an agglutinative language, has abundant particles and endings as one of its characteristic linguistic features, it was necessary to extract formulaic expressions of 5-gram from the full-morpheme. This study suggests that studies on Korean formulaic expressions need to consider typological characteristics of agglutinative languages and an appropriate methodology should be applied depending on the purpose of the study.

담화 유형에 따른 한국어 비유창성 현상 연구

남길임(Nam, Kilim) 담화·인지언어학회 2011 담화와 인지 Vol.18 No.3

This study aims to classify the types of disfluencies in spoken Korean and analyze frequency and distribution in different discourse types. Although disfluencies appearing in natural speech have been regarded as errors or unavoidable mistakes, it will be worth investigating from the linguistic point of view since disfluencies are believed to occur spontaneously and make speech more natural. This research intends to provide a basic resource for spoken Korean spoken processing, by looking into types of disfluencies, since disfluency analysis is one of the essential procedures of preprocessing in parsing spoken Korean corpora. This study, at first, reviews the characteristic features of disfluencies considering typological features of the Korean language, and attempts to classify the types of disfluencies. Then frequency and distribution of disfluencies collected from three typical discourse types will be observed aiming to find out generic regularities. Finally, possible implications in natural language processing in spoken Korean will be discussed.

상품평 텍스트에 나타난 감성표현 연구

남길임(Nam, Kilim) 언어과학회 2016 언어과학연구 Vol.0 No.78

This study aims to define sentiment expressions from the perspective of Korean linguistics and present the research process for extracting sentiment expressions in Korean product reviews. For the purpose of this study, an 8 million word clothing corpus was built; this was composed of product reviews, online magazines and professional fashion blogs, which have been morphologically annotated. Chapter 2 presents the corpus-driven methodology adopted for the research process, which focuses on n-gram and POS-gram analysis. Chapter 3 addresses issues regarding the pre-processing procedure and the establishment of the list of candidate sentiment expressions at word and phrase level. Chapter 4 deals with issues concerning phrase level analysis and discusses the correlation between sentiment analysis in domain-specific genres and the notion of ‘semantic unit’ as defined in corpus linguistics.