• Proceedings of the Korea Contents Association Conference
• /
• 2012.05a
• /
• pp.15-16
• /
• 2012

본 논문에서는 이동 P2P 환경에서 연속적인 서비스 제공을 위한 k-최근접 질의 처리 기법을 제안한다. 제안하는 기법은 질의 피어와 근접한 k개의 피어를 효율적으로 모니터링하기 위하여 MR(Monitoring-Region)을 배포한다. MR은 주변 피어들의 이동 벡터를 이용하여 시간 범위 내에서 최소 k개의 질의 결과를 보장하기 위해 만들어진 영역이다. 제안하는 기법은 주변 피어들의 벡터를 이용하여 MR을 계산하기 때문에 장시간 유효하며 MR 영역의 무효화에 따른 질의 재수행 비용을 절약한다. 제안하는 기법의 우수성을 입증하기 위해 기존 기법과의 성능평가를 수행한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2007.05a
• /
• pp.61-62
• /
• 2007

본 논문에서는 도로 네트워크 데이터베이스에서 k-최근접 이웃 질의를 효율적으로 처리하기 위한 방안에 대하여 논의한다. 네트워크 거리는 삼각형 부등식 성질(triangular inequality property)을 만족하지 못하므로 기존의 기법들은 네트워크 거리를 기반으로 하는 인덱스를 사용하지 않았다. 이러한 기법들은 질의 처리 시 심각한 성능 저하의 문제를 가진다. 사전 계산된 네트워크 거리를 이용하는 또 다른 기법은 저장 공간의 오버헤드가 크다는 문제를 갖는다. 본 논문에서는 이러한 두 가지 문제점들을 동시에 해결하기 위하여 객체들 간의 네트워크 거리를 근사하여 객체들에 대한 인덱스를 구축하고, 이를 이용하여 k-최근접 이웃 질의를 처리하는 새로운 기법을 제안한다. 실제 도로 네트워크를 이용한 정확도 검증 실험을 통하여 제안된 기법의 우수성을 규명한다.

• The KIPS Transactions:PartD
• /
• v.12D no.5 s.101
• /
• pp.673-682
• /
• 2005

Distance browsing queries, namely k-nearest neighbor queries, are the most important queries in spatial database applications, e.g., Geographic Information Systems(GISs). Recently, GIS applications trends to extend themselves toward wide multi-user environments such as the Web. Since many techniques for such queries, where Hjaltason and Samet's algorithm is the most efficient one, were optimized for only one query, we need to complement them suitable for multi-user environments. It can be a good approach that we store many individual query results in a cache, i.e., query result caching and reuse them in evaluating incoming queries, j.e., query result matching. In this paper, we propose a complementary Hjaltason and Samet's algerian capable of reusing previous query results in a cache for answering distance browsing queries in multi-user GIS environments. Our experimental results conform the efficiency of our approach.

• Spatial Information Research
• /
• v.20 no.4
• /
• pp.83-92
• /
• 2012

Location-based services (LBS) are increasingly popular due to the improvement of geo-positioning capabilities and wireless communication technology. However, in order to enjoy LBS services, a user requesting a query must send his/her exact location to the LBS provider. Therefore, it is a key challenge to preserve user's privacy while providing LBS. To solve this problem, the existing method employs a 2PASS cloaking framework that not only hides the actual user location but also reduces bandwidth consumption. However, 2PASS does not fully guarantee the actual user privacy because it does not take the real user distribution into account. Hence, in this paper, we propose a nearest neighbor query processing algorithm that supports K-anonymity property based on the weighted adjacency graph(WAG). Our algorithm not only preserves the location of a user by guaranteeing k-anonymity in a query region, but also improves a bandwidth usage by reducing unnecessary search for a query result. We demonstrate from experimental results that our algorithm outperforms the existing one in terms of query processing time and bandwidth usage.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• v.9 no.1
• /
• pp.163-166
• /
• 2005

The latest wireless communications technology bring about the rapid developments of Global Position System and Location-Based Service. It is very important for the moving object database to deal with database queries related to the trajectories of a moving objects and the valid time of the query results as well. In this paper, we propose how to get not only the current result of query but also the valid time and the result after the time when a query point and objects are moving at the same time. We would like to predict the valid time by formula because the current results will be incorrect due to the characteristic of the continuous movements of the moving objects and the future results can not be calculated by iterative computations.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.11b
• /
• pp.121-123
• /
• 2005

유클리디언(Euclidean)공간을 가정한 이동객체를 위한 색인구조 연구가 많이 진행되어 왔다. 그러나 LBS(Location Based System) 및 텔레매틱스 응용이 활발해짐에 따라, 이를 효과적으로 지원하기 위해서 도로 철도와 같은 공간 네트워크를 고려한 색인구조 연구가 요구된다. 따라서 본 논문에서는 공간 네트워크상에서 이동객체를 효율적으로 색인할 수 있는 궤적기반 색인구조를 설계한다. 이를 위해 공간 데이터와 시간 데이터를 분리하여 색인하고, 실세계에서 공간네트워크상의 이동객체에 대한 질의를 시공간+궤적 질의, k-최근접 질의, 시간+궤적 질의로 분류하고, 이를 위한 질의 처리 알고리즘을 설계한다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.12
• /
• pp.2919-2925
• /
• 2014

Because of the rapid growth of mobile communication and wireless communication, Location-based services are handled in many applications. So, the management and analysis of spatio-temporal data are a hot issue in database research. Index structure and query processing of such contents are very important for these applications. This paper addressees algorithms that make index structure by using Douglas-Peucker Algorithm and process nearest neighbor search query efficiently on moving objects trajectories. We compare and analyze our algorithms by experiments. Our algorithms make small size of index structure and process the query more efficiently.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2004.11a
• /
• pp.67-70
• /
• 2004

수십, 수백만의 이동 객체가 존재하는 환경에서 전체 이동 객체의 궤적을 중앙 서버가 모두 관리하는 것은 적절한 접근 방법이 아니다. 통신 메시지들이 서버에 집중되기 때문에, 높은 네트워크 대역폭, 처리 능력, 그리고 방대한 저장 공간을 보유한 서버를 필요로 하기 때문이다. 이에 본 논문에서는 중앙 서버없이 각각의 이동 객체들이 자기 자신의 궤적을 관리하는 방식을 통해 앞서의 문제를 해결하려한다. 중앙 서버없이 데이터가 네트워크에 분산되어 있는 경우, 특정 데이터에 효율적으로 접근하기 위해서는 색인이 필요하게 되는데, 본 논문에서는 헤더 객체와 헤더 검색 트리라는 것을 정의하여 IPv6의 모바일 IP를 이용한 P2P방법으로 영역 질의, 최근접 질의, 궤적 질의 처리를 할 수 있는 모델을 제시한다.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.49 no.2
• /
• pp.53-60
• /
• 2012

Nearest-neighbor classification predicts the class of an input data with the most frequent class among the near training data of the input data. Even though nearest-neighbor classification doesn't have a training stage, all of the training data are necessary in a predictive stage and the generalization performance depends on the quality of training data. Therefore, as the training data size increase, a nearest-neighbor classification requires the large amount of memory and the large computation time in prediction. In this paper, we propose a prototype selection algorithm that predicts the class of test data with the new set of prototypes which are near-boundary training data. Based on Tomek links and distance metric, the proposed algorithm selects boundary data and decides whether the selected data is added to the set of prototypes by considering classes and distance relationships. In the experiments, the number of prototypes is much smaller than the size of original training data and we takes advantages of storage reduction and fast prediction in a nearest-neighbor classification.

• Spatial Information Research
• /
• v.20 no.3
• /
• pp.67-81
• /
• 2012

Due to the recent advancement of cloud computing, the research on database outsourcing has been actively done. Moreover, the number of users who utilize Location-based Services(LBS) has been increasing with the development in w ireless communication technology and mobile devices. Therefore, LBS providers attempt to outsource their spatial database to service provider, in order to reduce costs for data storage and management. However, because unauthorized access to sensitive data is possible in spatial database outsourcing, it is necessary to study on the preservation of a user's privacy. Thus, we, in this paper, propose a spatial data encryption scheme to produce outsourced database from an original database. We also propose a k-Nearest Neighbor(k-NN) query processing algorithm that efficiently performs k-NN by using the outsourced database. Finally, we show from performance analysis that our algorithm outperforms the existing one.