• Proceedings of the Korean Information Science Society Conference
• /
• 2000.04b
• /
• pp.134-136
• /
• 2000

본 논문에서는 고차원의 특징 벡터 공간에서의 객체에 대한 효율적인 검색을 지원하는 셀기반 시그니쳐 트리 색인 구조(CS-트리, CI-트리)를 제안한다. 특징 벡터 공간을 셀로써 분할하고 특징 벡터는 셀의 시그니쳐로 표현되며 트리에 저장된다. 특징 벡터 대신 시그니쳐를 사용하여 트리의 깊이가 낮아짐으로서 검색을 효율적으로 수행할 수 있다. 또한 셀에 적합한 새로운 가지치기 거리를 이용한 유사성 검색 알고리즘으로 수행할 수 있다. 또한 셀에 적합한 새로운 가지치기 거리를 유사성 검색 알고리즘을 제시한다. 마지막으로 우수한 고차원 색인 기법으로 알려져 있는 X-트리와 성능 비교를 수행하여, 성능비교 결과 본 논문에서 제안하는 CS-트리와 CI-트리가 검색 시간 측면에서 최대 30%의 검색 성능이 개선됨을 보인다.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.2
• /
• pp.16-20
• /
• 2001

An important aspect of video program indexing and retrieval is the ability to segment video program into meaningful segments, in other words, the ability of content-based video program segmentation. In this paper, a new approach using speech recognition technology has been proposed for content-based video program segmentation. This approach uses speech recognition technique to synchronize closed caption with speech signal. Experimental results demonstrate that the proposed scheme is very promising for content-based video program segmentation.

• Proceedings of the KAIS Fall Conference
• /
• 2008.05a
• /
• pp.205-208
• /
• 2008

보다 효과적인 키워드 추출 및 키워드 가중치 결정을 위하여 문서의 내용뿐 아니라 구조를 이용하여 색인을 추출하는 연구가 이루어지고 있는데, 대부분의 연구들이 XML 단락별 중요도가 아닌, 문맥상의 단락에 대한 중요도를 계산하는게 일반적이다. 이러한 기존 연구들은 대부분이 객관적인 실험을 통해서 중요도를 입증하기보다는 일반적인 관점에서 단순한 수치로 중요도를 결정하고 있다. 본 논문에서는 웹 문서 관리를 위한 표준으로 자리잡아가고 있는 XML 문서의 자동색인을 위하여, 논문을 구성하는 주요 단락을 세분하고, 단락에서 추출된 용어의 가중치를 갱신해 가면서 최종 색인어 가중치를 계산하는 방법을 제안한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2001.04a
• /
• pp.301-304
• /
• 2001

영상내의 문자정보는 색인에 필요한 유용한 정보를 제공하기 때문에 이를 이용한 멀티미디어 데이터의 색인기법이 최근 많이 연구되고 있다. 본 논문에서는 칼라동영상에서 실시간으로 문자를 추출하고 추적하는 방법을 제안하였다. 제안된 방법은 연결성분(connected component)방법을 이용한 문자추출 모듈과 SSD(a Sum Of Squared Difference)를 이용한 문자추적모듈로 구성되어져 있다. 실제 TV영상에 대하여 제안된 방법을 테스트 해본 결과 빠른 문자추출과 추적시간을 가졌다.

• Proceedings of the Korean Society for Information Management Conference
• /
• 2003.08a
• /
• pp.95-103
• /
• 2003

목차 및 권말색인과 같은 인쇄형태의 정보내용에 대한 구조화된 접근방식에서 착안하여 전자 문서의 내용에 대한 새로운 형태의 접근방식을 개발할 수 있는데, 이를 위한 방안으로 용어 자동분류 기법이 있다. 본 연구에서는 용어의 의미모호성 문제를 해결하는 동시에 용어간 계층관계 표현이 가능한 자동분류 기법으로 퍼지 클러스터링 기법을 제안하고, 대표적인 퍼지 클러스터링 알고리즘인 퍼지 c-means 기법에 대해 분석하고자 한다.

• Journal of KIISE:Databases
• /
• v.33 no.5
• /
• pp.495-512
• /
• 2006

Recently, researches on a future position prediction of moving objects have been progressed as the importance of the future position retrieval increases. New index structures are required to efficiently retrieve the consecutive positions of moving objects. Existing index structures significantly degrade the search performance of the moving objects because the search operation makes the unnecessary extension of the node in the index structure. To solve this problem, we propose a space partition based index structure considering the mobility of moving objects. To deal with the overflow of a node, our index structure first merges it and the sibling node. If it is impossible to merge them, our method splits the overflow node in which moving properties of objects are considered. Our index structure is always partitioned into overlap free subregions when a node is split. Our split strategy chooses the split position by considering the parameters such as velocities, the escape time of the objects, and the update time of a node. In the internal node, the split position Is determined from preventing the cascading split of the child node. We perform various experiments to show that our index structure outperforms the existing index structures in terms of retrieval performance. Our experimental results show that our proposed index structure achieves about $17%{\sim}264%$ performance gains on current position retrieval and about $107%{\sim}19l%$ on future position retrieval over the existing methods.

• Journal of KIISE:Databases
• /
• v.30 no.5
• /
• pp.438-450
• /
• 2003

Spatial joins find pairs of objects that overlap with each other. In spatial joins using indexes, original-space indexes such as the R-tree are widely used. An original-space index is the one that indexes objects as represented in the original space. Since original-space indexes deal with sizes of objects, it is difficult to develop a formal algorithm without relying on heuristics. On the other hand, transform-space indexes, which transform objects in the original space into points in the transform space and index them, deal only with points but no sites. Thus, spatial join algorithms using these indexes are relatively simple and can be formally developed. However, the disadvantage of transform-space join algorithms is that they cannot be applied to original-space indexes such as the R-tree containing original-space objects. In this paper, we present a novel mechanism for achieving the best of these two types of algorithms. Specifically, we propose a new notion of the transform-space view and present the transform-space view join algorithm(TSVJ). A transform-space view is a virtual transform-space index based on an original-space index. It allows us to interpret on-the-fly a pre-built original-space index as a transform-space index without incurring any overhead and without actually modifying the structure of the original-space index or changing object representation. The experimental result shows that, compared to existing spatial join algorithms that use R-trees in the original space, the TSVJ improves the number of disk accesses by up to 43.1% The most important contribution of this paper is to show that we can use original-space indexes, such as the R-tree, in the transform space by interpreting them through the notion of the transform-space view. We believe that this new notion provides a framework for developing various new spatial query processing algorithms in the transform space.

• The KIPS Transactions:PartD
• /
• v.8D no.4
• /
• pp.305-312
• /
• 2001

Recently, high-dimensional index structures have been required for similarity search in such database applications s multimedia database and data warehousing. In this paper, we propose a new cell-based signature tree, called CS-tree, which supports efficient storage and retrieval on high-dimensional feature vectors. The proposed CS-tree partitions a high-dimensional feature space into a group of cells and represents a feature vector as its corresponding cell signature. By using cell signatures rather than real feature vectors, it is possible to reduce the height of our CS-tree, leading to efficient retrieval performance. In addition, we present a similarity search algorithm for efficiently pruning the search space based on cells. Finally, we compare the performance of our CS-tree with that of the X-tree being considered as an efficient high-dimensional index structure, in terms of insertion time, retrieval time for a k-nearest neighbor query, and storage overhead. It is shown from experimental results that our CS-tree is better on retrieval performance than the X-tree.

• The Transactions of the Korea Information Processing Society
• /
• v.3 no.5
• /
• pp.1160-1169
• /
• 1996

In this study, an experiment of information retrieval using fuzzy connection matrix of keywords was conducted. A query for retrieval was constructed from each keyword and Boolean operator such as AND, OR, NOT. In a workstation environment, the performance of the fuzzy retrieval system was proved to be considerably effective than that of the system using the crisp set theory. And both recall ratio and precision ratio showed that the proposed technique would be a possible alternative in future information retrieval. Some special features of this experimental system were ; ranking the results in the order of connectivity, making the retrieval results correspond flexibly by changing the threshold value, trying to accord the retrieval process with the retrieval semantics by treating the averse-connectivity (fuzzy value) as a semantic approximation between kewords.

• The KIPS Transactions:PartD
• /
• v.10D no.2
• /
• pp.175-186
• /
• 2003

Mobile clients like a PDA need a cache and a spatial index to search and access map data efficiently. When a server transmits spatial objects to a mobile client which has a low storage capacity, some of them can be duplicated in a cache of the mobile client. Moreover, the cost for strong added data in the cache and reconfiguring spatial index is very high in the mobile client with low computing power. The scheme for processing duplicated objects and disturbing tasks of the mobile client which has low computing power is needed. In this paper, we classfy the method for strorng duplicated objects and present the scheme for the both caching objects and reconfiguring a spatial index of cached objects using the clipping technique. We propose the caching system based on an agent in order to distribute the overhead of a mobile client as well as to provide efficiently map services. We design and implement it, and evaluate the performance.