• Proceeding of Spring/Autumn Annual Conference of KHA
• /
• 2006.11a
• /
• pp.190-195
• /
• 2006

The purpose of this study is to suggest the digitalizing model of Belonging relationship information. for this study, the psychological demands of people are analysed based on the premise that the formation of various space is made by mental needs rather than functional needs. The Digitalized Belonging relationship information model consisted of such structuring factors, the visual relationship information, the accessible relationship information and the opening ratios of spaces As a result, the suggested digitalizing model will be considered as an efficient tool for the objective analysis of Belonging relationship between space. In addition, the model will contribute to the expansion of terminology in the field of digital space design

• Journal of the Korean Geographical Society
• /
• v.32 no.4
• /
• pp.543-548
• /
• 1997

New technologie of information and interaction create ambiguous, multifaceted relationships between concepts and realities that once seemed clear and simple. The acvent of cyberspace implies the existence of cyberplaces and relationships among cyberspace and places that could be as or more complex then those that exist among the spaces and places of the non-digital world. Virtual reality-the current uses of infomation technologies to create replicas of places and processes in the world-offers a point of exploring such relationships. But more penetrating insights become evident in the examination of real virtuality-the uses of information technologies to enhance the experience of places in the world. Real virtuality offers geographers a variety of innovative and powerful tools for augmenting the effectiveness and domain of their craft. Widespread use of real virtuality by geograpgers will alter their use of data, methods, and theory in research, teaching, and practice.

• Journal of KIISE:Computing Practices and Letters
• /
• v.5 no.5
• /
• pp.553-563
• /
• 1999

The Spatial Database Engine(SDETM) developed by Environmental Systems Research Institute, Inc. is a spatial database that employs a client-server architecture incorporated with a set of software services to perform efficient spatial operations and to manage large, shared and geographic data sets. It currently supports a wide variety of spatial search methods and spatial relationships determined dynamically. Spatial objects in the space world can be changed by either non-spatial operations or spatial operations. Conventional geographical information systems(GISs) did not manage their historical information, however, because they handle the snapshot images of spatial objects in the world. In this paper we propose a spatio-temporal data model and an algorithm for temporal relationship macro which is able to manage and retrieve the historical information of spatial objects. The proposed spatio-temporal data model and its operations can be used as a software tool for history management of time-varying objects in database without any change.

• Journal of Cadastre & Land InformatiX
• /
• v.50 no.1
• /
• pp.47-62
• /
• 2020

As demands and applications of spatial information increase, different aspects of quality management have been raised as an important issue. This study suggests a quality management method for consumer-oriented spatial information, providing the consumer satisfaction. As opposed to the demands mentioned, less attention has been paid on the quality for the spatial information. So far, most of spatial information producers have kept their own independent quality management system and standard. Because of this unconscious response, it has been difficult to reflect the consumer's various and flexible demands from the silo structure of the quality management systems and regulations. The explosive increase of spatial information products creates quality-related problems such as limited time and budget for the consumer-oriented quality management. To solve these problems this study suggests to use a minimum number of the basic spatial information products which can guarantee better qualities when other products are combined. Because each of the spatial information product can include several other sub-products inside, it can have intrinsic characteristics such as geographic accuracy relationships between the products when combined and a hierarchical structure in each product in terms of the quality management mentioned. To prove the usability of the model, a case of the National Spatial Data Infrastructure Center is used because the Center collects and distributes an enormous amount of spatial information to the public and private sector.

• The KIPS Transactions:PartD
• /
• v.8D no.5
• /
• pp.516-572
• /
• 2001

Geographic Information System(CIS) deal with data which can potentially be useful for a wide range of applications. The information needed by each application can be vary, specially in resolution, detail level, application view, and representation style, as defined in the modeling phase of the geographic database design. To be able to deal with such diverse needs, GIS must offer features that allow multiple representation for each geographic entity of phenomenon. This paper addresses on the problem of formal definition of the objects and their relationships on the geographical information systems. The geographical data is divided into two main classes : geo-objects and geo-fields, which describe discrete and continuous representations of spatial reality. I studied the attributes and the relationship roles over geo-object and nongeo-object. Therefore, this paper contributed on the efficient design of geographical class hierarchy schema by means of formalizing attribute-domains of classes.

• KIPS Transactions on Software and Data Engineering
• /
• v.5 no.11
• /
• pp.569-584
• /
• 2016

In this paper, we present a spatial knowledge extractor implemented in Hadoop MapReduce parallel, distributed computing environment. From a large spatial dataset, this knowledge extractor automatically derives a qualitative spatial knowledge base, which consists of both topological and directional relations on pairs of two spatial objects. By using R-tree index and range queries over a distributed spatial data file on HDFS, the MapReduce-enabled spatial knowledge extractor, MRSPAKE, can produce a web-scale spatial knowledge base in highly efficient way. In experiments with the well-known open spatial dataset, Open Street Map (OSM), the proposed web-scale spatial knowledge extractor, MRSPAKE, showed high performance and scalability.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2005.11a
• /
• pp.221-224
• /
• 2005

비디오 데이터가 가지는 움직임 정보에 대한 의미적 표현을 위해 저차원의 특징뿐만 아니라 시 공간적 관계 표현, 이벤트 인식에 이르기까지 의미적 인식을 위한 많은 연구가 되고 있다. 하지만, 이러한 노력은 메타 데이터 형태로 미디어객체에 단순히 키워드를 부여하는 방법으로 내용을 표현하는 정도이고 진정한 의미적 내용을 표현하는 것은 현재로서는 불가능하다. 본 논문은 비디오 데이터의 객체에 대한 움직임 클래스를 정의하여 의미적 검색을 위한 시 공간 관계 표현 규칙을 제안한다. 특히 축구 비디오 데이터에 대한 의미적 검색을 위해 객체간의 시 공간적 관계 모델을 제시한다. 이벤트에 대한 계층적 표현으로 축구경기의 "Goal" 장면을 사용한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2001.04a
• /
• pp.93-96
• /
• 2001

공간 데이터베이스에서는 공간 객체에 대해 객체마다 서로 다른 보안 등급을 설정함으로써 공간 객체에 대한 보안을 유지하고 있다. 다단계 보안을 사용하여 사용자의 보안 등급에 따라 공간 객체에 대한 접근을 제어할 경우 인접 객체와의 위상관계를 통해서 공간 정보뿐만 아니라 비공간 정보의 노출(information flow)이라는 문제가 발생하게 된다. 이에 대한 해결책으로 데이터베이스에 대한 부분집합만을 추출하여 사용자의 권한에 맞는 질의 결과를 제공하는 것이다. 즉 공간 뷰를 정의함으로써 사용자 등급에 맞는 결과 집합(result set)을 제공함으로써 보다 효율적인 연산을 수행할 수 있게 된다. 따라서 본 논문에서는 실세계의 지리 객체들에 대한 다양한 사용자 관점과 다단계 보안 등급을 가진 뷰를 지원하기 위해 관계형 데이터베이스의 뷰 개념을 확장한 공간 데이터베이스의 공간 뷰를 제안하고, 공간 연산 시 발생하는 정보 흐름 제어를 위한 방법을 제시한다.

• 한국게임학회지
• /
• v.6 no.1
• /
• pp.52-54
• /
• 2009

프로그램된 정보를 사용하는 자율 가상 캐릭터는 항상 반복된 패턴 행동을 하기 때문에 사용자가 흥미를 잃는 경우가 많고 현실성도 떨어진다. 본 논문에서는 이러한 문제점을 해결하기 위해, 자율 가상캐릭터가 자율적으로 인지한 정보를 저장하고 저장한 정보를 활용하여 상황에 맞는 행동을 수행할 수 있는 메모리 체계를 제안한다. 본 논문은 자율 가상 캐릭터가 시각 주목으로 인지한 정보를 저장하고 관리하는 메모리 체계의 모델을 제시한다. 메모리 용량을 효율적으로 사용할 수 있도록 게임 환경에 적합한 빠른 시각 주목 알고리즘을 연구하여 중요하고 눈에 띄는 정보만 저장한다. 자율 가상 캐릭터의 인지 메모리를 크게 시각 기억와 공간 관계 기억 구조로 구성한다. 시각 기억은 쿼드그래프로 구현된 저장 구조에 인지한 정보를 저장한다. 공간 관계 기억은 공간 관계 그래프 이론을 기반으로 객체들간의 방향과 거리 정보를 저장한다. 본 논문의 제안 방법을 가상 환경에서 실험한 결과, 자율 가상 캐릭터는 시각 주목 기능으로 3차원 가상 환경의 동적 객체까지 감지하여 자율적으로 정보를 주목하여 저장하고 있음을 확인했다. 자율 가상캐릭터는 메모리 정보를 활용하여 목표 객체를 빠르게 탐색하며 길찾기에 필요한 경로 계획을 수립한다. 성능면에서는 주목맵만들기 위한 특징맵으로 가장 주목할 수 있는 특징들로 구성하여 처리속도가 1.6배 이상 향상됨을 확인했다.

• The Journal of the Korea Contents Association
• /
• v.19 no.11
• /
• pp.278-287
• /
• 2019

Spatial information extraction is to retrieve static and dynamic aspects in natural language text by explicitly marking spatial elements and their relational words. This paper proposes a deep learning approach for spatial information extraction for Korean language using a two-step bidirectional LSTM-CRF ensemble model. The integrated model of spatial element extraction and spatial relation attribute extraction is proposed too. An experiment with the Korean SpaceBank demonstrates the better efficiency of the proposed deep learning model than that of the previous CRF model, also showing that the proposed ensemble model performed better than the single model.