• Annual Conference on Human and Language Technology
• /
• 2021.10a
• /
• pp.203-208
• /
• 2021

본 논문에서는 맥락에 따라 개체명의 범주가 달라지는 어휘를 중심으로 교차 태깅된 개체명의 성능을 레이블과 스팬 정답률, 문장 성분과 문장 위치에 따른 정답률로 나누어 살펴 보았다. 레이블의 정확도는 KoGPT2, mBERT, KLUE-RoBERTa 순으로 정답률이 높아지는 양상을 보였다. 스팬 정답률에서는 mBERT가 KLUE-RoBERTa보다 근소하게 성능이 높았고 KoGPT2는 매우 낮은 정확도를 보였다. 다만, KoGPT2는 개체명이 문장의 끝에 위치할 때는 다른 모델과 비슷한 정도로 성능이 개선되는 결과를 보였다. 문장 종결 위치에서 인식기의 성능이 좋은 것은 실험에 사용된 말뭉치의 문장 성분이 서술어일 때 명사의 중첩이 적고 구문이 패턴화되어 있다는 특징과 KoGPT2가 decoder기반의 모델이기 때문으로 여겨지나 이에 대해서는 후속 연구가 필요하다.

• Annual Conference on Human and Language Technology
• /
• 2012.10a
• /
• pp.154-156
• /
• 2012

영한 기계번역 시스템의 개발을 위해서는 여러 가지 사전이 필요하고, 다앙한 모호성 해소를 위한 연구를 위한 데이터가 필요하며, 번역 시스템의 테스트를 위해 많은 영어 문장이 필요하다. 따라서 영어 말뭉치를 구축하여 이로부터 사전에 필요한 정보, 모호성 해소 연구에 필요한 데이터, 번역 테스트를 위한 문장 등을 추출할 필요가 있다. 본 논문에서는 영어 말뭉치를 구축하기 위해 인터넷 영어 신문 사이트로부터 영어로 작성된 신문기사를 추출하는 도구를 개발하였다. 이를 통해 자동적으로 영어 신문기사를 추출하여 말뭉치를 구축할 수 있으며, 이를 통해 영한 기계번역 시스템의 성능 향상을 지원할 수 있다.

• Journal of the Korea Society of Computer and Information
• /
• v.18 no.12
• /
• pp.91-101
• /
• 2013

One issue with keyword search in general is its ambiguity which can ultimately impact the effectiveness of the search in terms of the quality of the search results. This ambiguity is primarily due to the ambiguity of the contextual meaning of each term in the query. In addition to the query ambiguity itself, the relationships between the keywords in the search results are crucial for the proper interpretation of the search results by the user and should be clearly presented in the search results. We address the keyword search ambiguity issue by adapting some of the existing approaches for keyword mapping from the query terms to the schema terms/instances. The approaches we have adapted for term mapping capture both the syntactic similarity between the query keywords and the schema terms as well as the semantic similarity of the two and give better mappings and ultimately 50% raised accurate results. Finally, to address the last issue of lacking clear relationships among the terms appearing in the search results, our system has leveraged semantic web technologies in order to enrich the knowledgebase and to discover the relationships between the keywords.

• Journal of KIISE:Software and Applications
• /
• v.30 no.9
• /
• pp.850-861
• /
• 2003

When creating annotation information in a free~form environment, ambiguity arises during the analysis stage between geometric information and the annotations. This needs to be resolved so that the accurate creation of annotation information in a free-form annotation environment is possible. This paper identifies and analyzes the ambiguities, specifying methods that are tailored to each of the various contexts that can cause conflicts with free-form marking in a XML-based annotation environment. The proposed general method is based on context which includes various textual and structure information between free-form marking and the annotations themselves. The context information used is expressed in XML based DTD, within the paper. The results are printed and shared through a system specifically implemented for this study. The results from the implementation of the Proposed method show that the annotated areas included in the free-form marking information are more accurate, achieving more accurate exchange results amongst multiple users in a heterogeneous document environment.

• Proceedings of the Korea Society of Information Technology Applications Conference
• /
• 2001.11a
• /
• pp.65-65
• /
• 2001

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.9 no.4
• /
• pp.958-963
• /
• 2008

Korean language has different structural properties which are controlled by semantic constraints of verbs. Also, most of Korean sentences are complex sentences which consisted of main clause and embedded clause. Therefore it is difficult to describe appropriate syntactic grammar or constraint for the Korean language and the Korean parsing causes various syntactic ambiguities. In this paper, we suggest how to describe CFG-based grammar using sentence patterns as syntactic constraint and solve syntactic ambiguities. To solve this, we classified 44 sentence patterns including complex sentences which have subordinate clause in Korean sentences and used it to reduce syntactic ambiguity. However, it is difficult to solve every syntactic ambiguity using the information of sentence patterns. So, we used semantic markers with semantic constraint. Semantic markers can be used to solve ambiguity by auxiliary particle or comitative case particle.

• Proceedings of the Korean Society for Cognitive Science Conference
• /
• 2000.06a
• /
• pp.116-123
• /
• 2000

한국어 구문 분석은 체언구 부착이나 부사구 부착의 문제를 가진다. 이런 부착의 문제는 많은 구문 모호성을 만들어 내어 올바른 의미를 가지는 파스 트리의 선택을 어렵게 한다. 한국어에서 이런 부착의 문제는 대부분 한국어 문장이 내포문을 포함하는 복문의 형태로 구성되어 있기 때문이다. 단문에서는 부착의 문제가 발생하지 않지만 복문에서는 체언구나 부사구가 어떤 용언에 부착하느냐에 따라 체언구 부착이나 부사구 부착의 문제가 발생한다. 따라서 용언이 가지는 정보를 이용하여 내포문의 범위를 결정해서 하나의 구문범주의 기능을 가지도록 분할한다. 이를 단문 분할이라 하며 문형이 가지는 필수격들을 최대로 부착하여 이루어진다. 단문분할을 하면 복문의 구조가 단문으로 바뀌므로 이런 부착의 문제가 자연스럽게 해결된다. 본 논문에서는 문형과 단문 분할을 이용하여 많은 구문 모호성을 해결할 수 있음을 제안한다.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.3
• /
• pp.239-248
• /
• 2006

Current context-aware applications In ubiquitous computing environments make the assumption that the context they are dealing with is correct. However, in reality, both sensed and interpreted context informations are often uncertain or imperfect. In this paper, we propose a probability extension model to ontology-based model for rep resenting uncertain contexts and use Bayesian networks to resolve about uncertainty of context informations. The proposed model can support the development and operation of various context-aware services, which are required in the ubiquitous computing environment.

• The Journal of Korean Institute of Next Generation Computing
• /
• v.13 no.2
• /
• pp.102-110
• /
• 2017

Current word sense disambiguation techniques employ various machine learning-based methods. Various approaches have been proposed to address this problem, including the knowledge base approach. This approach defines the sense of an ambiguous word in accordance with knowledge base information with no training corpus. In unsupervised learning techniques that use a knowledge base approach, graph-based and similarity-based methods have been the main research areas. The graph-based method has the advantage of constructing a semantic graph that delineates all paths between different senses that an ambiguous word may have. However, unnecessary semantic paths may be introduced, thereby increasing the risk of errors. To solve this problem and construct a fine-grained graph, in this paper, we propose a model that iteratively constructs the graph while eliminating unnecessary nodes and edges, i.e., senses and semantic paths. The hybrid similarity estimation model was applied to estimate a more accurate sense in the constructed semantic graph. Because the proposed model uses BabelNet, a multilingual lexical knowledge base, the model is not limited to a specific language.

• The KIPS Transactions:PartB
• /
• v.16B no.3
• /
• pp.225-232
• /
• 2009

Author name disambiguation is essential for improving performance of document indexing, retrieval, and web search. Author name disambiguation resolves the conflict when multiple authors share the same name label. This paper introduces a novel approach which exploits ontologies and WordNet-based category utility for author name disambiguation. Our method utilizes author knowledge in the form of populated ontology that uses various types of properties: titles, abstracts and co-authors of papers and authors' affiliation. Author ontology has been constructed in the artificial intelligence and semantic web areas semi-automatically using OWL API and heuristics. Author name disambiguation determines the correct author from various candidate authors in the populated author ontology. Candidate authors are evaluated using proposed WordNet-based category utility to resolve disambiguation. Category utility is a tradeoff between intra-class similarity and inter-class dissimilarity of author instances, where author instances are described in terms of attribute-value pairs. WordNet-based category utility has been proposed to exploit concept information in WordNet for semantic analysis for disambiguation. Experiments using the WordNet-based category utility increase the number of disambiguation by about 10% compared with that of category utility, and increase the overall amount of accuracy by around 98%.