• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2014년도 춘계학술발표대회
• /
• pp.761-764
• /
• 2014

본 연구는 지능형 QA시스템과 관련한 연구에서, 자동 패러프레이즈 생성 시스템을 평가하는 새로운 방법을 제시한다. 기존의 패러프레이즈 생성 시스템의 자동 평가 방법은 참조할 수 있는 패러프레이즈 정보의 양이 크게 제한되어 있었으며, 원 문장의 콘텍스트(context)와 이에 의존하는 통사적 구조(syntactic structure) 및 의미적 구조의 유사성을 고려하지 않고, 단순 구/단어 수준의 의미 유사성을 기반으로 생성된 패러프레이즈를 평가하였다. 이러한 문제를 해결하기 위해 본 연구는 시맨틱 프레임(semantic frame)을 이용한 패러프레이즈 문장 평가 방법을 제시한다. 본 연구에서 제시하는 방법론은 문장의 콘텍스트를 표현하는 프레임과 이러한 프레임이 발생시키는 통사적, 의미적 구조의 유사성을 바탕으로 원 문장과 패러프레이즈 문장의 '의미 유사성', '어휘 형태 비 유사성'을 평가하는 방식이다.

• International Journal of Computer Science & Network Security
• /
• 제21권12spc호
• /
• pp.712-718
• /
• 2021

Virtual communities are studied to analyze their characteristic features, types, and tole to modern society. The article is aimed at creating a classification of virtual communities according to specific characteristics, which can be used to model the interaction, and necessity of components that are important for the community. The classification of virtual communities will contribute to their better performance and satisfy the users' needs in information. The study reveals the value structure of virtual communities, educational and communicative influence, and the possible threats these communities may bring to society and security.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2016년도 추계학술대회
• /
• pp.375-378
• /
• 2016

The difficulty of "information sharing", "information reusing" issues happening in Wireless Sensor Network is due to the heterogeneity of the application environment, data processing, communication protocol etc. Based on the introduction of the Ontology theory, though analyzing the sensor characteristic a general type of sensor ontology contains the definition of concept, frame structure and OWL design was proposed from the standpoint of sensor observation. The paper expounded a system framework of the domain ontology through the expansion of knowledge base on the general sensor could achieve the information sharing and reuse by semantic communication between the general sensor ontology and user. The research of this method would bring new idea to the semantic sensor network construction.

• 중국학논총
• /
• 제71호
• /
• pp.41-59
• /
• 2021

With the progress of social economy and science and technology, network media technology has developed rapidly, China has ushered in the network information age, and the network buzzwords emerged to reflect the interaction and influence between language and society. The network buzzwords of "X+ ren "indirectly show the social psychology and value orientation of modern people with their unique structural characteristics, semantic connotation and cultural deposits, and so on. Based on this, we have conducted a multi-angle investigation on the network buzzwords "X+ ren". This paper first analyzes the structure types and syntactic functions of the lexical model of "X+ ren ", then makes a semantic analysis of the lexical model of "X+ Ren ", and finally investigates the causes and influences of the popularity of "X+ ren ". Through the investigation, we believe that "X+ ren "will continue to grow, and "X+ ren" will continue to attract the attention of the academic community.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권1호
• /
• pp.60-79
• /
• 2022

The U-Net architecture-based segmentation models attained remarkable performance in numerous medical image segmentation missions like skin lesion segmentation. Nevertheless, the resolution gradually decreases and the loss of spatial information increases with deeper network. The fusion of adjacent layers is not enough to make up for the lost spatial information, thus resulting in errors of segmentation boundary so as to decline the accuracy of segmentation. To tackle the issue, we propose a new deep learning-based segmentation model. In the decoding stage, the feature channels of each decoding unit are concatenated with all the feature channels of the upper coding unit. Which is done in order to ensure the segmentation effect by integrating spatial and semantic information, and promotes the robustness and generalization of our model by combining the atrous spatial pyramid pooling (ASPP) module and channel attention module (CAM). Extensive experiments on ISIC2016 and ISIC2017 common datasets proved that our model implements well and outperforms compared segmentation models for skin lesion segmentation.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2008년도 춘계학술발표대회
• /
• pp.427-430
• /
• 2008

본 논문은 비음수 행렬 인수분해(NMF, non-negative matrix factorization)와 NMF 군집방법을 이용하여 다중문서를 요약하는 새로운 방법을 제안하였다. 본 논문에서 NMF에 의해 계산된 의미 특징(semantic feature)은 문서의 고유 구조(inherent structure)를 반영하여 문장을 추출함으로써 요약의 질을 높일 수 있고, 의미 변수(semantic variable)를 이용한 문장의 군집은 문장 간의 유사성과 다양성 고려하여서 쉽게 과잉정보를 제거하여 문장을 요약할 수 있는 장점을 갖는다.

• 한국언어정보학회:학술대회논문집
• /
• 한국언어정보학회 2001년도 학술대회 논문집
• /
• pp.25-58
• /
• 2001

This study explores the lexical-semantic structure of derived causative verbs in Korean based on Pustejovsky(1995)'s Generative Lexicon Theory (GL), Mor-phological causative verbs are derived from their root stems by affixing ‘-i, -hi, -li, -gi’ in Korean and the meanings of derived predicates are closely related to the meanings of their root verbs. In particular, the change of the ARGUMENT STRUCTURE by morphological derivation leads to the change of the EVENT STRUCTURE. In this study, causation is defined as the cause-effect relation having a causer. The ARGUMENT STRUCTURES of derived causative verbs includes a causer argument, which is added to the ARGUMENT STRUCTURE of their root verbs. Their EVENT STRUCTURE has a headed process related to a causer and their result is the event which their root verbs represent. This approach can also suggest that the (in)directness of causative is determined by which verb is its root and explain the difference between the morphological causativization and the syntactic causativization in Korean.

• International Journal of Internet, Broadcasting and Communication
• /
• 제5권1호
• /
• pp.14-17
• /
• 2013

RDF data is widely used for exchanging data nowadays to enable semantic web era. This leads to the need for storing and retrieving these data efficiently and effectively. Recently, the structure index in graph-based perspective is considered as a promising approach to deal with issues of complex query graphs. However, even though there are many researches based on structure indexing, there can be a better architectural approach instead of addressing the issue as a part. In this research, we propose architecture for storing, query processing and retrieving RDF data in efficient manner using structure indexing. Our research utilizes research results from iStore and 2 relation-based approaches and we focus on improving query processing to reduce the time of loading data and I/O cost.

• 말소리와 음성과학
• /
• 제9권2호
• /
• pp.61-68
• /
• 2017

The aim of this study is to investigate the prosodic aspects of ambiguous sentences in Korean produced by Chinese Korean Learners (L1: Chinese, L2: Korean). In Korean, sentence ambiguity can be caused by homonym or syntactically ambiguous structure. In spoken language however all ambiguity can be resolved by different prosodic features according to the meaning that they transmit. In this study we examined whether Chinese Korean Leaners also distinguish, in production, ambiguous sentences on the basis of prosodic characteristics. For this study 4 Korean natives speakers and 10 advanced Chinese Korean learners participated in the production test. The material analysed constituted 10 Korean sentences in which 6 sentences are lexically ambiguous and 4 sentences contain structural ambiguity. The results show that Korean native speakers produced ambiguous sentences by different prosodic structure depending on their semantic and syntactic structure. Chinese speakers also show distinct prosodic structure for different ambiguous sentences in most cases. But in the phonetic realization, the internal pitch range was greater for Korean native speakers than Chinese learners.

• 한국과학교육학회지
• /
• 제35권6호
• /
• pp.1031-1038
• /
• 2015

사람의 머릿속에 어떤 단어들이 저장되어 있으며, 이들이 어떤 구조로 연결되어 있는지에 대해 심리학, 정신분석학, 언어학, 교육학 등 다양한 분야에서 관심을 가지고 연구하고 있다. 과학 교육 분야에서는 구성주의적 관점에서 학생들의 머릿속에 과학용어로 표현되는 개념들이 어떻게 자리 잡고 있는지가 관심 대상 가운데 하나이다. 본 연구에서는 과학용어들에 대한 학생들의 머릿속 의미 구조를 알아보기 위한 하나의 방법인 단어연상검사를 실시함에 있어 검사 결과에 영향을 미치는 검사 환경 변인을 알아보았다. 학생들이 제시된 자극어를 과학적 상황으로 제한하도록 유도할 가능성이 있다고 여겨지는 변인을 4가지(자극어의 과학용어 여부 명시, 일상어와 섞어 제시, 검사자의 과학 교사 여부, 해당 개념의 학습 여부)로 정하고, 실험 집단과 통제 집단의 검사 결과 비교를 통해 이들 4가지 변인이 과학용어에 대한 단어연상검사 결과에 영향을 미치는지 확인해 보았다. 일상적 의미와 과학적 의미를 모두 포함하는 과학용어 9개를 정하고 중학교 학생 282명을 통제집단과 실험집단으로 나누어 단어연상검사를 실시하였다. 그 결과 검사지에 제시된 자극어들이 모두 과학용어임을 직접적으로 명시하는 경우에 학생들은 과학적 의미로 인식하는 비율이 높았으며, 나머지 변인에 대해서는 통제집단과 실험집단 사이에 차이가 없는 것으로 나타났다.