• The KIPS Transactions:PartB
• /
• v.10B no.2
• /
• pp.229-236
• /
• 2003

Named entity recognition is the process indispensable to question answering and information extraction systems. This paper presents an HMM based named entity (m) recognition method using the construction principles of compound words. In Korean, many named entities can be decomposed into more than one word. Moreover, there are contextual relationships among nouns in an NE, and among an NE and its surrounding words. In this paper, we classify words into a word as an NE in itself, a word in an NE, and/or a word adjacent to an n, and train an HMM based on NE-related word types and parts of speech. Proposed named entity recognition (NER) system uses trigram model of HMM for considering variable length of NEs. However, the trigram model of HMM has a serious data sparseness problem. In order to solve the problem, we use multi-level back-offs. Experimental results show that our NER system can achieve an F-measure of 87.6% in the economic articles.

• Journal of Internet Computing and Services
• /
• v.13 no.2
• /
• pp.99-108
• /
• 2012

Many methods had been developed to improve the accuracy of extracting information from a vast amount of data. This paper combined a number of natural language processing methods such as NER (named entity recognition), sentence extraction, and part of speech tagging to carry out text analysis. The data source is comprised of texts obtained from the web using a domain-specific data extraction agent. A framework for the extraction of information from unstructured data was developed using the aforementioned natural language processing methods. We simulated the performance of our work in the extraction and analysis of texts for the detection of organizational structures. Simulation shows that our study outperformed other NER classifiers such as MUC and CoNLL on information extraction.

• Journal of Creative Information Culture
• /
• v.5 no.2
• /
• pp.145-156
• /
• 2019

In this study, we have developed and evaluated an information extracting module based on the named entity recognition technique. For the given purpose in this paper, the module was designed to apply to the problem dealing with extraction of postal address information from arbitrary documents without any prior knowledge on the document layout. From the perspective of information technique practice, our approach can be said as a probabilistic n-gram (bi- or tri-gram) method which is a generalized technique compared with a uni-gram based keyword matching. It is the main difference between our approach and the conventional methods adopted in natural language processing that applying sentence detection, tokenization, and POS tagging recursively rather than applying the models sequentially. The test results with approximately two thousands documents are presented at this paper.

• Journal of KIISE:Software and Applications
• /
• v.36 no.8
• /
• pp.642-652
• /
• 2009

In order to improve the performance of the existing relation extraction approaches, we propose a method for combining two pivotal concepts which play an important role in classifying semantic relationships between entities in text. Having built a composite kernel-based relation extraction system, which incorporates both entity features and syntactic structured information of relation instances, we define nine classes of lexical features and synthetically apply them to the system. Evaluation on the ACE RDC corpus shows that our approach boosts the effectiveness of the existing composite kernels in relation extraction. It also confirms that by integrating the three important features (entity features, syntactic structures and contextual lexical features), we can improve the performance of a relation extraction process.

• Korean Journal of Cognitive Science
• /
• v.21 no.4
• /
• pp.655-667
• /
• 2010

The named entity recognition task is one of the most important subtasks in Information Extraction. In this paper, we describe a Korean named entity recognition using structural Support Vector Machines (structural SVMs) and modified Pegasos algorithm. Using the proposed approach, we could achieve an 85.43% F1 and an 86.79% F1 for 15 named entity types on TV domain and sports domain, respectively. Moreover, we reduced the training time to 4% without loss of performance compared to Conditional Random Fields (CRFs).

• Annual Conference on Human and Language Technology
• /
• 2021.10a
• /
• pp.482-485
• /
• 2021

최근 지식 그래프를 확장하기 위해 많은 연구가 진행되고 있다. 지식 그래프를 확장하기 위해서는 relation을 기준으로 entity의 방향성을 고려하는 것이 매우 중요하다. 지식 그래프를 확장하기 위한 대표적인 연구인 관계 추출은 문장과 2개의 entity가 주어졌을 때 relation을 예측한다. 최근 사전학습 언어모델을 적용하여 관계 추출에서 높은 성능을 보이고 있지만, entity에 대한 방향성을 고려하여 relation을 예측하는지 알 수 없다. 본 논문에서는 관계 추출에서 entity의 방향성을 고려하여 relation을 예측하는지 실험하기 위해 문장 수준의 Adversarial Attack과 단어 수준의 Sequence Labeling을 적용하였다. 또한 관계 추출에서 문장에 대한 이해를 높이기 위해 BERT모델을 적용하여 실험을 진행하였다. 실험 결과 관계 추출에서 entity에 대한 방향성을 고려하지 않음을 확인하였다.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.8
• /
• pp.371-380
• /
• 2023

Relation extraction is to extract relationships between named entities from text. Traditionally, relation extraction methods only extract relations between predetermined subject and object entities. However, in end-to-end relation extraction, all possible relations must be extracted by considering the positions of the subject and object for each pair of entities, and so this method uses time and resources inefficiently. To alleviate this problem, this paper proposes a method that sets directions based on the positions of the subject and object, and extracts relations according to the directions. The proposed method utilizes existing relation extraction data to generate direction labels indicating the direction in which the subject points to the object in the sentence, adds entity position tokens and entity type to sentences to predict the directions using a pre-trained language model (KLUE-RoBERTa-base, RoBERTa-base), and generates representations of subject and object entities through probabilistic crossover operation. Then, we make use of these representations to extract relations. Experimental results show that the proposed model performs about 3 ~ 4%p better than a method for predicting integrated labels. In addition, when learning Korean and English data using the proposed model, the performance was 1.7%p higher in English than in Korean due to the number of data and language disorder and the values of the parameters that produce the best performance were different. By excluding the number of directional cases, the proposed model can reduce the waste of resources in end-to-end relation extraction.

• Journal of Information Science Theory and Practice
• /
• v.6 no.4
• /
• pp.28-38
• /
• 2018

Wikipedia is composed of millions of articles, each of which explains a particular entity with various languages in the real world. Since the articles are contributed and edited by a large population of diverse experts with no specific authority, Wikipedia can be seen as a naturally occurring body of human knowledge. In this paper, we propose a method to automatically identify key entities and relations in Wikipedia articles, which can be used for automatic ontology construction. Compared to previous approaches to entity and relation extraction and/or identification from text, our goal is to capture naturally occurring entities and relations from Wikipedia while minimizing artificiality often introduced at the stages of constructing training and testing data. The titles of the articles and anchored phrases in their text are regarded as entities, and their types are automatically classified with minimal training. We attempt to automatically detect and identify possible relations among the entities based on clustering without training data, as opposed to the relation extraction approach that focuses on improvement of accuracy in selecting one of the several target relations for a given pair of entities. While the relation extraction approach with supervised learning requires a significant amount of annotation efforts for a predefined set of relations, our approach attempts to discover relations as they occur naturally. Unlike other unsupervised relation identification work where evaluation of automatically identified relations is done with the correct relations determined a priori by human judges, we attempted to evaluate appropriateness of the naturally occurring clusters of relations involving person-artifact and person-organization entities and their relation names.

• Interdisciplinary Bio Central
• /
• v.3 no.3
• /
• pp.9.1-9.5
• /
• 2011

Entity Attribute Value (EAV) is the widely used solution to represent high dimensional and sparse data, but EAV is not search efficient for knowledge extraction. In this paper, we have proposed a search efficient data model: Optimized Entity Attribute Value (OEAV) for physical representation of high dimensional and sparse data as an alternative of widely used EAV. We have implemented both EAV and OEAV models in a data warehousing en-vironment and performed different relational and warehouse queries on both the models. The experimental results show that OEAV is dramatically search efficient and occupy less storage space compared to EAV.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.10d
• /
• pp.676-678
• /
• 2002

질의 응답 시스템(Question Answering: QA)에서 정답 유형 부류(Answer Type Taxonomy: ATT)란 사용자 질문 분석을 위한 미 부류 체계를 의미하는 것으로, ATT의 크기가 클수록 시스템의 성능은 높아진다. ATT를 확장하기 위해서는, 개체(Named Entity)에 의미 범주를 결정하는 개체 분류기(Named Entity Tagger의 분류 체계가 세분되어야 하는데, 기존의 개체 분류기는 한문서 내에서 그 개체의 분류를 시도하기 때문에, 분류를 위한 문맥 정보의 양이 부족하여, 정확하고 상세한 분류를 기대하기 힘들다. 본 논문에서는 동일 개체에 대한 문맥 정보를 수집하기 위해, 그 개체가 나타나는 다른 문서들을 검색하는 개체 피드백 Named Enti쇼 Feedback)이라는 기법을 사용한다. 개체가 상세히 분류됨에 따라 ATT도 확장될 수 있었으며, 이렇게 확장된 ATT상에서의 정답 추출은 baseline보다 약 7%정도의 성능 향상을 보여, 개체 피드백의 효과를 확인할 수 있었다.