• Proceedings of the Korean Information Science Society Conference
• /
• 2010.06c
• /
• pp.385-389
• /
• 2010

논문에 부여되는 저자 주제어(author keyword)는 논문을 분류 및 검색하는데 활용될 수 있다. 이렇게 주제어를 부여할 때 자동으로 저자 주제어를 추천한다면 사용자에게 편리성을 제공하고 저자가 직접 부여한 저자 주제어 이외에 추가적으로 주제어가 있는지도 확인할 수 있어 유용하다. 본 연구에서는 논문에 달려있는 다수의 주제어 중 하나의 주제어를 선별하여 Labeled LDA를 이용해 주제어와 초록(abstract)의 관계를 학습했다. 이후 초록이 주어지면 자동으로 저자 주제어를 부여할 수 있도록 추천하는 기법을 제안하고 그에 따른 실험을 진행했다. 본 논문에서는 실험을 통하여 기계학습을 이용한 저자 주제어의 추천이 어느 정도의 성능을 보이는지 평가하고 향후 연구의 방향을 제시한다.

• Journal of the Korean Society for information Management
• /
• v.25 no.1
• /
• pp.191-210
• /
• 2008

We would generally use subject terms such as subject indexing for searching and accessing documents. So then, there must be any relationship between document's full-text and its subject terms. This study is started in this question. Master's theses in field of science and technology are worked with because full-text is relatively formatted. This study is to study locations of subject term on Thesis, distribution patterns of subject terms on content of full-text; 'Contents', 'Introduction', 'Theory', 'Main subject', 'Conclusion' and 'References'. Thesis were averagely composed of 1226.3 terms. And Subject terms were averagely compose of $12{\sim}13$ terms. As a result, 'Contents' and 'Introduction' have had the most frequency of subject.

• Proceedings of the Korean Society for Information Management Conference
• /
• 2013.08a
• /
• pp.43-46
• /
• 2013

본 연구에서는 국내 문헌정보학 분야의 연구 논문 중 이용자 관련 연구 논문 125편을 대상으로 논문에 부여된 주제어간의 연결 관계를 분석 하였다. 사전 작업을 통하여 정리된 226개의 주제어에 대한 연결 관계를 네트워크 분석을 통하여 분석하고 시각화 하였다. 그래프를 통하여 주제어간 연결 강도를 확인하였고, 다른 주제어와 연결성이 높은 상위 20개의 주제어를 제시하였다. 주제어간 근접성이 높은 주제어를 군집화한 결과 14개의 군집으로 정리되었다. 다른 주제어와 연결이 없이 고립된 군집이 8개, 연결된 군집이 6개였다.

• Annual Conference on Human and Language Technology
• /
• 2015.10a
• /
• pp.139-141
• /
• 2015

본 논문은 커뮤니티 기반의 질의-응답 서비스에서 사용자 질의의 주제를 분류하는 시스템을 소개한다. 커뮤니티 기반의 질의-응답 서비스는 분야에 따라 다양한 주제를 가질 수 있으며 오늘 날 사용자 질의의 주제 분류에는 통계 기반의 분류 방법이 많이 이용되고 있다. 통계 기반의 분류 방법으로 사용자 질의를 분류하기 위해서는 주제에 적합한 대량의 학습 말뭉치가 필요하다. 주제에 적합한 대량의 학습 말뭉치를 사람이 직접 구축하는 것은 많은 시간과 비용이 든다. 따라서 본 논문에서는 이러한 문제를 해결하기 위해 위키피디아 문서를 Supervised K-means Clustering 기법으로 주제별로 분류함으로써 학습 말뭉치를 반자동으로 구축하는 방법을 제안한다. 그 다음, 생성된 학습 말뭉치로 지지 벡터 기계를 학습하여 사용자 질의의 주제를 분류하게 된다. 위키피디아 문서와 사용자 질의는 다른 도메인의 문서임에도 불구하고 본 논문의 시스템으로 사용자 질의의 주제를 분류한 결과 77.33%의 정확도를 보였다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2015.01a
• /
• pp.19-22
• /
• 2015

본 논문에서는 문장 및 문단에서 키워드의 역할에 따른 가중치에 근거하여 도서 본문에서 주제어를 추출하는 방법을 제안한다. 기존의 주제어 추출 방식은 도서 본문이 아닌 신문이나 논문에 대한 방식이므로 도서 본문에서의 주제어 추출에 그대로 적용하기에는 어려움이 있다. 따라서 본 논문에서는 빈도수뿐만 아니라 문장 내 중요 요소에 대한 가중치와 중요 문장에 대한 가중치를 후보 키워드에 부여하는 방식을 제안하였다. 제안한 계산 방식을 비문학 도서에 대하여 실험한 결과, 빈도수만으로 주제어를 추출한 기존 방식보다 본 논문에서 제안한 방식의 주제어 추출 결과의 정확도가 향상되는 것을 확인하였다.

• The Journal of the Korea Contents Association
• /
• v.18 no.8
• /
• pp.178-186
• /
• 2018

Subject classification of thesis units is essential to serve scholarly information deliverables. However, to date, there is a journal-based topic classification, and there are not many article-level subject classification services. In the case of academic papers among domestic works, subject classification can be a more important information because it can cover a larger area of service and can provide service by setting a range. However, the problem of classifying themes by field requires the hands of experts in various fields, and various methods of verification are needed to increase accuracy. In this paper, we try to classify topics using the unsupervised learning algorithm to find the correct answer in the unknown state and compare the results of the subject classification algorithms using the coherence and perplexity. The unsupervised learning algorithms are a well-known Hierarchical Dirichlet Process (HDP), Latent Dirichlet Allocation (LDA) and Latent Semantic Indexing (LSI) algorithm.

• Journal of the Korean Society for information Management
• /
• v.30 no.1
• /
• pp.237-258
• /
• 2013

As topics of researchers become diverse horizontally or vertically, academic libraries have difficulties to identify the dynamic change of researchers' needs for academic publications. This research aims to illustrate the topic areas of researchers in a department of university by analyzing bibliographies of their publications. First, researchers' publications were used to discover the topic areas where the researchers had published. Second, the cited publications in those papers were analysed to identify the expanded topic areas of these researchers. Finally, highly cited journals were analyzed by network analysis method. The major finding is that the importance of topic areas by the number of journals was not necessarily proportional to that by the number of papers. Researchers have a tendency to use many papers in a small number of journals in a certain topic area. Furthermore, the importance of topic areas discovered by researchers' publications was not the same as that discovered by researchers' citations.

• Journal of the Korea Society of Computer and Information
• /
• v.13 no.7
• /
• pp.59-66
• /
• 2008

This paper describes a novel method for identifying the main topic and detecting topic changes in a text-based dialogue as in Instant Messaging (IM). Compared to other forms of text, dialogues are uniquely characterized with the short length of text with small number of words, two or more participants, and existence of a history that affects the current utterance. Noting the characteristics, our method detects the main topic of a dialogue by considering the keywords not only the utterances of the user but also the dialogue system's responses. Dialogue histories are also considered in the detection process to increase accuracy. For topic change detection, the similarity between the former utterance's topic and the current utterance's topic is calculated. If the similarity is smaller than a certain threshold, our system judges that the topic has been changed from the current utterance. We obtained 88.2% and 87.4% accuracy in topic detection and topic change detection, respectively.

• The Journal of the Korea Contents Association
• /
• v.13 no.12
• /
• pp.136-148
• /
• 2013

I would like to trace back periodic social trend by searching if regular performance might have reflected on social trend on its theme by analyzing theme trend of The National Dance Company of Korea, Dance Company of Seoul city, Dance Company of KyungKi-Do and to provide the meaningful results for further study by checking if the theme of dancing performance is in relation with social structure. To perform this research, I had studied on previous thesis and reference books. For example, I selected three groups, of The National Dance Company of Korea, Dance Company of Seoul city, Dance Company of KyungKi-Do, to research their theme of regular performance through checking previous thesis related to, performance material, news articles, pamphlets from beginning to present. How to analyze is being proceeded from foundation of dancing company to present according to Kim Byungseok's classification method, which was consistently used for searching theme trend from previous study as below; 1) Theme based on traditional conscious,2) Theme based on Literature, 3) Theme based on Historic issues, 4) Theme based on abstract, 5) Theme based on reality, 6)Theme based on social issues.

• Journal of the Korean BIBLIA Society for library and Information Science
• /
• v.22 no.3
• /
• pp.353-371
• /
• 2011

The purpose of this paper is to analyze the interrelation among research areas based on domestic journal papers, achievements of korea researchers. Generally, the content of papers is appeared through abstracts, subjects, full-text and so on. This paper is focused on subject terms of Domestic journal papers. The experimental data are 80 domestic journals, 7,616 papers and 58,143 subject terms and papers published in 2009. As the result, it was different to use subject terms on each research area: Engineering, Agriculture & Oceanography, Interdisciplinary Science, Social Science, Arts & Physical Education, Medicine & Pharmacology, Humanities and Natural Science. Subject terms of Engineering have used the most in the other research areas in aspect of term co-occurrence. The 8 research areas were grouped in 3 clusters: C1(Engineering, Natural Science, Social Science, Interdisciplinary Science, Humanities), C2(Medicine & Pharmacology, Arts & Physical Education), and C3(Agriculture & Oceanography).