• Annual Conference on Human and Language Technology
• /
• 2019.10a
• /
• pp.276-279
• /
• 2019

의존 구문 분석은 문장을 구성하는 성분들 간의 의존 관계를 분석하고 문장의 구조적 정보를 얻기 위한 기술이다. 의미역 결정은 문장에서 서술어에 해당하는 어절을 찾고 해당 서술어의 논항들을 찾는 자연어 처리의 한 분야이다. 두 기술은 서로 밀접한 상관관계가 존재하며 기존 연구들은 이 상관관계를 이용하기 위해 의존 구문 분석의 결과를 의미역 결정의 자질로써 사용한다. 그러나 이런 방법은 의미역 결정 모델의 오류가 의존 구문 분석에 역전파 되지 않으므로 두 기술의 상관관계를 효과적으로 사용한다고 보기 어렵다. 본 논문은 포인터 네트워크 기반의 의존 구문 분석 모델과 병렬화 순환 신경망 기반의 의미역 결정 모델을 멀티 태스크 방식으로 학습시키는 통합 모델을 제안한다. 제안 모델은 의존 구문 분석 및 의미역 결정 말뭉치인 UProbBank를 실험에 사용하여 의존 구문 분석에서 UAS 0.9327, 의미역 결정에서 PIC F1 0.9952, AIC F1 0.7312의 성능 보였다.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.1
• /
• pp.110-114
• /
• 2010

We describe a new method for text categorization using TextRank algorithm. Text categorization is a problem that over one pre-defined categories are assigned to a text document. TextRank algorithm is a graph-based ranking algorithm. If we consider that each word is a vertex, and co-occurrence of two adjacent words is a edge, we can get a graph from a document. After that, we find important words using TextRank algorithm from the graph and make feature which are pairs of words which are each important word and a word adjacent to the important word. We use classifiers: SVM, Na$\ddot{i}$ve Bayesian classifier, Maximum Entropy Model, and k-NN classifier. We use non-cross-posted version of 20 Newsgroups data set. In consequence, we had an improved performance in whole classifiers, and the result tells that is a possibility of TextRank algorithm in text categorization.

• Journal of the Korean Society for Library and Information Science
• /
• v.50 no.3
• /
• pp.53-69
• /
• 2016

The purpose of this study is to measure self images of public librarians who should increasingly take charge of services for children and young people through the semantic differential method and analyze its' general characteristics. According to the results of analysis public librarians have been satisfied with their works and believed that they are kind, fair and gentle. One the other hand, they have had negative self images such as no influence, neglect, poor and stuffiness. Therefore, the result show that public librarians consider their ability, quality and activity as positive images, but they have felt lack of the social consensus and affirmative evaluation of their role. So, we can see building a good physical environment and running great programs to create access of children and young people to the public libraries, demonstrating leadership among senior librarians are so important in order to beat the stereotype and improve their positive self images of their evaluation and role. It is also necessary to promote the new value and role of public libraries and librarianship and to enable retraining for cultivating their qualities as a teacher.

• Proceedings of the Korean Society for Language and Information Conference
• /
• 2002.06a
• /
• pp.10-38
• /
• 2002

본고는 한국어 영형 대명사의 적절한 해석을 위해 개념적으로 전혀 새로운 이론을 제안한다. 일련의 다양한 제약들이 서로 연관되어 있음을 보인 후, 그러한 규칙의 다양성을 적절히 포착하기 위해 적절성 이론 (Optimality Theory)을 도입할 것을 제안하고, 그 토대 위에 다양한 제약들을 형식화한 후, 그 규칙들의 위계관계를 설정한다. 가장 우선순위를 갖는 제약으로 인접 요소간 어휘의미자질들이 일치해야 한다는 어휘의미제약(*Feature Mismatch)과 통사적 결속규칙을 의미론적으로 재해석한 결속원리 B(Principle B)를 선정한다. 그 다음 순위를 갖는 제약으로, 가능한 한 선행명사를 지칭하도록 요구하는 대용존중제약(DOAP: Don't Overlook Anaphoric Possibilities)과, 센터링 이론의 전이방식 개념을 도입하여 정의한 계속선호제약 (CONTINUE)을 제안한다

• Journal of the Korean Society for information Management
• /
• v.21 no.4 s.54
• /
• pp.187-208
• /
• 2004

In this study, aesthetic impressions were used for a high-level feature of image retrieval. The term, 'aesthetic' has been studied in psychology, art, and literature. It means unconscious, instantaneous parts of visual perception and emotion. The literatures related to aesthetic impressions were reviewed and four kinds of aesthetic impressions were defined operationally : strong impression, soft impression, courteous impression, and refined impression. 66 image files of paintings were sampled randomly from 1100 paintings and low-level color features were extracted from them by a using perceptual color model(Lai, & Tait, 1998). The high-level features of an image, that is, four kinds of aesthetic impressions of each painting were measured by 4 subjects and averaged. In CBIR, 2 subjects performed image retrievals using example queries. They were asked to retrieve images by using the aesthetic impressions or the keywords. In evaluations, subjects showed that they were satisfied with the aesthetic impression-based image retrieval system on the average. And R-precision of the image retrieval with both color features and aesthetic impressions was higher than that of the image retrieval with color features only. But further studies with larger test collections and query sets should be followed for generalization of the result of this study.

• Annual Conference on Human and Language Technology
• /
• 2000.10d
• /
• pp.411-418
• /
• 2000

Hausser (1999)와 이기용 (1999a, 1999c)에서는 데이터베이스 관리 시스템(DBMS)을 이용하여 자연언어의 의미를 다루는 데이터베이스 의미론을 제안하였다. 특히 이기용 (1999c)에서는 수형도(tree), 논리 형태(logical fomulas), 자질 구조(feature structure)와 같은 다양한 언어 표상 형식들을 관계형 데이터베이스 관리 시스템(RDBMS)의 표상 형식인 테이블 형식으로 전환 가능함을 보임으로써 데이터베이스 의미론에 관계형 데이터 베이스 관리 시스템을 도입할 수 있음을 제시하였다. 한편, Lee (2000)에서 제시한 데이터베이스 의미론 모형에서는 데이터베이스 관리 시스템과 사용자(end-user)를 연결하는 언어 정보 처리 시스템(LIPS; Linguistic Information Processing System)을 제안하였다. 이 언어정보 처리 시스템은 사용자에 의해 입력된 언어 자료를 처리하여 그 분석 결과를 데이터베이스 관리 시스템에 전달하고, 이를 통해 구축된 데이터베이스에서 추출한 정보를 다시 사용자에게 전달하는 시스템이다. 이 논문은 한국어 '이, 히, 리, 기' 피동형을 전산처리를 할 수 있도록, 데이터베이스 의미론에서 핵심 요소인 언어정보 처리 시스템과 데이터베이스 관리 시스템을 구현하는 것 이 목적이다.

• Journal of KIISE
• /
• v.43 no.7
• /
• pp.773-780
• /
• 2016

Semantic information and features are very important for Semantic Role Labeling(SRL) though many SRL systems based on machine learning mainly adopt lexical and syntactic features. Previous SRL research based on semantic information is very few because using semantic information is very restricted. We proposed the SRL system which adopts semantic information, such as named entity, word sense disambiguation, filtering adjunct role based on sense, synonym cluster, frame extension based on synonym dictionary and joint rule of syntactic-semantic information, and modified verb-specific numbered roles, etc. According to our experimentations, the proposed present method outperforms those of lexical-syntactic based research works by about 3.77 (Korean Propbank) to 8.05 (Exobrain Corpus) F1-scores.

• The KIPS Transactions:PartB
• /
• v.9B no.1
• /
• pp.23-28
• /
• 2002

A fundamental problem in building an intelligent agent is that an agent does not understand the meaning of its perception or its action. One reason that an agent cannot understand the world is partially caused by a syntactic approach that converts a semantic feature into a simple string. To solve this problem, Cohen introduces a semantic approach that an agent autonomously learns a meaningful representation of physical schemas, on which some advanced conceptual structures are built, from physically interacting with environment using its own sensors and effectors. However, Cohen does not deal with a meta level of conceptual primitive that makes recognizing a schema possible. We propose that negation is a meta schema that enables an agent to recognize a physical schema. We prove some semantic aspects of negation.

• Journal of KIISE
• /
• v.42 no.4
• /
• pp.475-482
• /
• 2015

Developing a high-performance Semantic Role Labeling (SRL) system for a domain requires manually annotated training data of large size in the same domain. However, such SRL training data of sufficient size is available only for a few domains. Performances of Korean SRL are degraded by almost 15% or more, when it is directly applied to another domain with relatively small training data. This paper proposes two techniques to minimize performance degradation in the domain transfer. First, a domain adaptation algorithm for Korean SRL is proposed which is based on the prior model that is one of domain adaptation paradigms. Secondly, we proposed to use simplified features related to morphological and syntactic tags, when using small-sized target domain data to suppress the problem of data sparseness. Other domain adaptation techniques were experimentally compared to our techniques in this paper, where news and Wikipedia were used as the sources and target domains, respectively. It was observed that the highest performance is achieved when our two techniques were applied together. In our system's performance, F1 score of 64.3% was considered to be 2.4~3.1% higher than the methods from other research.

• The KIPS Transactions:PartB
• /
• v.17B no.1
• /
• pp.85-92
• /
• 2010

A shallow semantic parsing system analyzes the relationship that a syntactic constituent of the sentence has with a predicate. It identifies semantic arguments representing agent, patient, instrument, etc. of the predicate. In this study, we propose a two-phase shallow semantic parsing model which consists of the identification phase and the classification phase. We first find the boundary of semantic arguments from partial syntactic parsing results, and then assign appropriate semantic roles to the identified semantic arguments. By taking the sequential two-phase approach, we can alleviate the unbalanced class distribution problem, and select the features appropriate for each task. Experiments show the relative contribution of each phase on the test data.