담화 표지 화제 결속 기능 연구-화제 시작 기능을 중심으로-

심란희 한국언어문화교육학회 2020 언어와 문화 Vol.16 No.1

This study aims to categorize ‘topic coherence function’ of Korean discourse markers and analyze ‘starting a topic function’. First, the study analyzes the functions of discourse markers from the perspective of ‘discourse coherence’ and analyzes their functions to classify them into seven types of coherence layers. The topic coherence function was further divided into ‘macroscopic coherence of topic’ including starting a topic, changing a topic, and returning to a topic and ‘microscopic coherence of topic’ including ‘filling the gap,’ ‘modifying the articulation,’ and ‘listing.’ Then discourse markers provided according to the functions and the usage of each function have been analyzed. A list of discourse markers by function is provided. The significance of the present study comes from systematically classifying the topic coherence functions of discourse markers that have been sporadically suggested and providing a list of discourse markers for Starting a topic function from the perspective of teaching Korean as a foreign language. The results of the present study can be utilized as fundamental data in selecting discourse markers as a lesson and in composing a syllabus. .

버토픽과 텍스트랭크의 융합을 통한 토픽모델링의 개선 및 사례 분석

김근형,강재정 한국정보시스템학회 2024 정보시스템연구 Vol.33 No.3

Purpose The purpose of this paper is to develop a method to improve topic representation by incorporating the TextRank technique in Bertopic-based topic modeling and additional indicators for determining the optimal number of topics. Design/methodology/approach In this paper, we propose a method to extract important documents from documents assigned to each topic of a topic model using the TextRank technique, and to calculate secondary diversity and generate topic representations based on the results. First, we integrate the TextRank algorithm into the Bertopic-based topic modeling process to set local secondary labels for each topic. The secondary labels of each topic are derived through extractive summarization based on the TextRank algorithm. Second, we improve the accuracy of selecting the optimal number of topics by calculating the secondary diversity index based on the extractive summary results of each topic. Third, we improve the efficiency by utilizing ChatGPT when deriving the labels of each topic. Findings As a result of performing case analysis and analysis evaluation using the proposed method, it was confirmed that topic representation based on TextRank results generated more accurate topic labels and that the secondary diversity index was a more effective index for determining the optimal number of topics. Purpose The purpose of this paper is to develop a method to improve topic representation by incorporating the TextRank technique in Bertopic-based topic modeling and additional indicators for determining the optimal number of topics. Design/methodology/approach In this paper, we propose a method to extract important documents from documents assigned to each topic of a topic model using the TextRank technique, and to calculate secondary diversity and generate topic representations based on the results. First, we integrate the TextRank algorithm into the Bertopic-based topic modeling process to set local secondary labels for each topic. The secondary labels of each topic are derived through extractive summarization based on the TextRank algorithm. Second, we improve the accuracy of selecting the optimal number of topics by calculating the secondary diversity index based on the extractive summary results of each topic. Third, we improve the efficiency by utilizing ChatGPT when deriving the labels of each topic. Findings As a result of performing case analysis and analysis evaluation using the proposed method, it was confirmed that topic representation based on TextRank results generated more accurate topic labels and that the secondary diversity index was a more effective index for determining the optimal number of topics.

주제 진행에 따른 there 존재구문의 담화 기능

정효진 韓國敎員大學校 敎育硏究院 2009 敎員敎育 Vol.25 No.3

이 연구의 목적은 there 존재구문의 논리 주어인 후치 명사구(post verbal NP)의 담화 기능을 주제 진행의 측면에서 제시하는 것이다. there 구문의 의미상의 주어는 정보성 위주의 설명에서는 관찰되지 않았던 담화 기능을 수행하는데,이는 논리 주어의 정보성이나 단순한 의미 해석에 의존하는 것이 아니라 담화 흐름과 주제 응집성에 따라 그 기능을 달리한다는 것을 의미한다. 이 접근 방식은 학교 문법에서 there 구문의 의미 해석보다는 논리 주어가 후행 담화에서는 어떤 명사구로 연결되어 나타나는지를 명확히 설명하는 계기를 제공해 줄 것이다. 따라서 이 논문에서는 담화의 기본이 되는 문장과 나아가서는 담화를 구성하는 요소의 정보성,그 요소와 선, 후행 발화와의 관련성,그리고 화자와 청자의 상호 관계 등 담화에 관여한 모든 요소들을 고려하여 결정해야 한다는 담화 전개의 관점을 통해 there 존재구문에 나타난 후치 명사구의 담화 기능이 ‘주제 도입’임을 제시한다. This study aims to reanalysis discourse functions of English existential there constructions, which are called marked sentence types, with the focus on topic development.' It is argued that the discourse function of the postverbal NP (henceforth PVNP) of there constructions is 'topic introduction.' 1 have examined the discourse functions considering not only the common grammaticality of PVNPs but also the more detailed aspects such as the informational status, location of discourse, discourse processing and connections of the PVNPs in there constructions. In 'topic development,' the topic is introduced by types of definite NPs and indefinite NPs. It candinues as formed co-referential NP, deletion and anaphoric expressions, and when the different topical elements are introduced in discourse, then topic re-identification occurs. Also, the topics are realized as forming a 'topic coherence chain.' Based on these assumptions, I have proposed the discourse functions of PVNPs for there existential constructions The findings of this study suggest that discourse functions in English marked constructions apply not only to the word expressions of natural language in general but also to their discourse coherence, topic-chain and discourse topic management. Furthermore, topic-chain, it is helpful for students (readers) to identify pronouns, referring expressions in the discourse and to give logical thinking for the effective writings.

학문 목적 한국어 학습자의 토론 수행에서 주제 친숙도가 담화 응집성에 미치는 영향

이선영 국제한국어교육학회 2015 한국어 교육 Vol.26 No.4

The purpose of this study is to find out how topic familiarity of those learning Korean for academic purpose affects discourse coherence during discussion. During discussion, an important ability is to precisely understand the context of discourse and unfold insistence and proof fit for the context. Discussion with insistence and proof fit for the context has strong coherence. This study focused on the concept of ‘topic familiarity’ in a try to find out ways for those learning Korean to produce more coherent discussion discourse. So-far second language education has researched correlation between topic familiarity and learner’s performance mostly in the comprehension realm such as reading and listening. Researches in that realm have reported much higher comprehension when learners read or listen to texts with topics familiar to them. This study’s experiment, however, showed different result in the expression realm such as speaking. This study had 26 learners of Korean for academic purpose have discussion both on topics familiar and unfamiliar to them, which result showed that they produced more coherent discourse when discussing unfamiliar topics. This result may come from their enough preparation for insistence and proof before discussion is made during high-level Korean speaking class. It can be seen that learners make thorough preparation for proof and logic when they have to discuss unfamiliar topics and such preparation lead to more coherent discourse.

토픽 모델링에 따른 고등학생 논설문의 응결성과 응집성의 상관분석

이슬기 ( Lee Seulki ) 국어교육학회 2017 국어교육학연구 Vol.52 No.3

This study investigates the validity of the method used to measure coherence by analyzing the correlation between cohesion scores and the coherence of student texts calculated using topic modeling. For this purpose, cohesion through topic similarity was categorized into local, global, and overall text coherence. Next, the results were compared with teacher-generated coherence scores. The results showed that local cohesion was negatively correlated with the coherence scores. Global cohesion was not significantly correlated with teachers` grading results, and overall text cohesion showed a positive correlation. This study shows that coherence is related to the consistency of the whole text and that topic modeling can be a useful tool for providing objective information about the text.

단문의 주제 분석을 위한 LDA와 BTM 토픽모형 평가

문길성(Gil-Seong Mun) 한국자료분석학회 2021 Journal of the Korean Data Analysis Society Vol.23 No.3

토픽 모형은 대량의 문서 집합에서 잠재된 주제를 발견하기 위한 비지도학습의 하나이며, 가장 많이 이용되고 있는 모형은 LDA(Latent Dirichlet Allocation)인 것으로 알려져 있다. LDA는 문서 수준에서 단어의 동시 출현 패턴에 기반을 둔 모형이어서 길이가 긴 문서에서는 효과적으로 이용될 수 있으나, 트윗이나 인스턴트 메시지와 같은 길이가 짧은 문서에는 단어의 동시 출현이 희박하고 자주 사용되지 않는 단어의 출현으로 인하여 양질의 주제 추론이 어렵다는 한계가 있다. 이 문제를 개선하기 위한 하나의 대안으로 BTM(Bi-term Topic Model)이 있다. BTM은 bi-term을 사용하여 단어의 동시 발생을 모델링하고, 주제를 찾기 위하여 이러한 단어의 동시 발생 패턴을 bi-term의 말뭉치에서 집계한다. 본 연구의 목적은 LDA와 BTM의 성능을 비교하고 한국어 기반의 텍스트 데이터에서 BTM의 활용 가능성을 검토하는 데 있다. 이를 위하여 토픽의 질을 양적으로 측정할 수 있는 측도들을 검토하고 실제 자료에 적용하여 두 모형을 평가하였다. 평가결과, 주제 일관성과 문서 분류의 정확성 모두 BTM이 LDA 보다 우수한 것으로 나타났다. 이러한 결과는 단문에서 양질의 토픽을 추출하는 방법으로 BTM의 활용 가능성이 있음을 시사한다. Topic modeling is an unsupervised technique for discovering the underlying topics that occur in a collection of documents, and one popular model is known as LDA (Latent Dirichlet Allocation). It can be used effectively on a variety of long document types because the model implicitly captures the document-level word co-occurrence patterns to reveal topics. The model, however, may have difficulty extracting topics due to the severe data sparsity in short texts such as tweets and instant messages. An alternative to improve this problem is to BTM (Bi-term Topic Model). BTM models the word co-occurrence by using the explicit bi-term, and aggregate these word co-occurrence patterns in the corpus for topic discovering. This study is to compare the performance of LDA and BTM and explore the availability of BTM in Korean-based text data. We reviewed the measures to quantitatively represent the quality of the topics, and evaluated the two models based on real data. As a result, BTM outperformed LDA on short texts, using topic coherence and document classification accuracy as performance measures. The results suggest the availability of BTM to extract topics from short texts.

화제 중심의 텍스트 응집성 고찰

이신형 ( Sin Hyeong Lee ) 청람어문교육학회 2012 청람어문교육 Vol.45 No.-

The micro-structure and macro-structure of a text are connected each other by the medium of Topic.It performs a leading semantic function in forming text coherence. First, Topic functions as the central figure in the discourse development. ``Discoursetopic`` exists as ``abstract schema`` and it can be actualized as a shape of sentence topic word, combining with specific postpositions. Second,The ``semantic relation`` which is a decisive factor in forming text coherence is acted on by the knowledge structure. The knowledge structure is made up being centered on topic. The topic presents ``scope of predication`` which is originated from ``cognitive domain``. Third, The actualization of text coherence centered on topic can be examined in some cases. A case of the formal cohesion of textual surface, another case of the background knowledge of text acceptor, the other case of the situational context that the text is used.

Locative Inversion and Information Coherence

김윤희,김정석 한국영어학회 2016 영어학 Vol.16 No.2

Yunhui Kim and Jeong-Seok Kim. 2016. Locative Inversion and Information Coherence. Korean Journal of English Language and Linguistics 16-2, 245-264. This paper explores Locative Inversion (LI) in English. Following Kim (1998, 2000), we argue that the inverted locative PP is in fact an NP with a null head N. Consequently, we confirm Chomsky’s (1995) suggestion that it is only the nominal feature that enters into a checking relation with the EPP requirement of T. LI involves the special alignment of semantic properties with syntactic positions driven by information structure. It allows the presentation of relatively familiar information before comparatively unfamiliar information. In a minimalist framework, we propose that syntax may manipulate discourse-related features like [+topic] and [+focus] via uninterpretable versions of these on functional heads, Top and Foc. The proposal will be extended to the analysis of the cleft construction involving LI.

영어 모국어 화자와 중국어 모국어 화자의 한국어 이야기 주제어 학습: 언어 유형별 생략 현상 대조와 응집성

김지은(Jieun Kim) 한국중원언어학회 2021 언어학연구 Vol.- No.59

This study investigates how Chinese- and English-speaking learners of Korean as a foreign language ellipt a topic and maintain cohesion of a narrative text accordingly. It is known to have characteristics that Chinese and English are typologically different in terms of prominence of a topic vs. a subject (Li & Thompson, 1976). It has been reported that a topic is ellipted in a topic-prominent language in referring to old information, unlike in subject-oriented languages which typically disallow a null component in a sentence. The present study tests whether it is a typological difference or a topic-comment universality that is related to foreign language acquisition. A total of 36 subjects in their twenties in Seoul, Korea, including the two experimental groups and Korean native speakers as a control group, participated in this study. Spoken narratives were collected in a task of describing The Pear Story (Chafe, 1976). The results showed that Chinese and Korean native speakers used ellipsis more frequently than English native speakers. It is concluded that typology-specific first language properties play a role in foreign language learning.

법조문의 텍스트 분석

고성환 ( Song Hwan Ko ) 한국텍스트언어학회 2010 텍스트언어학 Vol.29 No.-

Each act is a text in itself. It is comprised of several chapters which are another types of texts. Each chapter, in turn, is composed of some articles, which lead to lower texts. Except for some exceptional cases, most articles form lower texts. An article has a completely standardized pattern, and so does its description structure. Since these characteristics are directly reflected on expression structures, some predominant expressions emerge in an article: the frequent use of topics, the fronting of transitional adverbs ``daman`` and ``geureona``, and negatives expressions in the predicates. In addition, it is noted that adnominal endings are used unusually from the standpoint of tense. These characteristic expression methods are a good way to secure cohesion. In particular, identical expressions in articles of the same act serve this function. Also, the unity in the expression method of titles not only enhances cohesion, but intensifies coherence.