• The KIPS Transactions:PartB
• /
• v.13B no.5 s.108
• /
• pp.533-540
• /
• 2006

The semantic web environment promise semantic search of heterogeneous data from distributed web page. Semantic search would resuit in an overwhelming number of results for users is increased, therefore elevating the need for appropriate personalized ranking schemes. Culture Finder helps semantic web agents obtain personalized culture information. It extracts meta data for each web page(culture news, culture performance, culture exhibition), perform semantic search and compute result ranking point to base user profile. In order to work efficient, Culture Finder uses five major technique: Machine learning technique for generating user profile from user search behavior and meta data repository, an efficient semantic search system for semantic web agent, query analysis for representing query and query result, personalized ranking method to provide suitable search result to user, upper ontology for generating meta data. In this paper, we also present the structure used in the Culture Finder to support personalized search service.

• Journal of the Korean Society for information Management
• /
• v.20 no.3
• /
• pp.111-127
• /
• 2003

The purpose of this study is to ascertain whether event tracking is more effective in event retrieval than information filtering. This study examined the two techniques for event retrieval to suggest the more effective one. The event-retrieval performances of the event tracking technique based on a kNN classifier and the query-based information filtering technique were compared. Two event tracking experiments, one with the static training set and the other with the dynamic training set , were carried out. Two information filtering experiments, one with initial queries and the other with refined queries, were also carried out to evaluate the event-retrieval effectiveness. We found that the event tracking technique with the static training set performed better than on with the dynamic training set. It was also found that the information fitering technique using intial queries performed better than one using the refined queries. In conclusion, the comparison of the best cases of event tracking and information filtering revealed that the information filtering technique outperformed the event tracking technique in event retrieval.

• Journal of the Korean Society for Library and Information Science
• /
• v.36 no.4
• /
• pp.175-205
• /
• 2002

Information visualization in information retrieval is a creating tool that enables us to observe, manipulate, search, navigate, explore, filter, discover, understand, interact with large volumes of data for more rapidly and far more effectively to discover hidden patterns. The focus of this study is to investigate and analyze information visualization techniques in information retrieval system in the three-schematic levels. In result, it was found that first, scientific data, documents, and retrieval result information are visualized through various techniques. Second, information visualization techniques which facilitate navigation and interaction are zoom and pan, focus+context techniques, incremental exploration, and clustering. Third, the visual metaphors used by the visualization systems are presented in the linear structure, hierarchy structure, network structure, and vector scatter structure.

• The KIPS Transactions:PartC
• /
• v.16C no.2
• /
• pp.223-228
• /
• 2009

Efficient service discovery mechanism is a crucial feature for a hybrid ad-hoc network supporting extension of a wireless ad-hoc network to the Internet. We propose an efficient cross-layer service discovery mechanism using non-disjoint multi-path source routing protocol for hybrid ad-hoc networks. Our scheme has advantages of multi-path routing protocol and cross-layer service discovery. Intuitively, it is not difficult to imagine that the cross-layer service discovery mechanism could result in a decreased number of messages compared to the traditional approach for handling routing independently from service discovery. By simulation, we show that faster route recovery is possible by maintaining multiple routing paths in each node, and the route maintenance overhead can be reduced by limiting the number of multiple routing paths and by maintaining link/node non-disjoint multi-path.

• 한국HCI학회:학술대회논문집
• /
• 2007.02a
• /
• pp.287-292
• /
• 2007

최근 다양한 분야에서(웹 포털, 유료 음원서비스 등) 디지털 오디오의 검색이 사용되고 있다. 이러한 분야에서 디지털 오디오의 검색은 디지털 오디오 데이터가 가지고 있는 자체 메타 정보를 이용하여 이루어진다. 하지만 메타 정보가 다르게 작성 되었거나 작성되지 않은 경우 정확한 검색은 어렵다. 요즘 이러한 문제의 보완 방안으로 내용기반 정보 검색 기법을 이용한 검색이 이루어지고 있다. 본 논문에서는 내용 기반 디지털 오디오 검색 방법에 대해 논하고자 한다. 내용기반으로 디지털 오디오를 검색하기 위해 음성 인식 문야에서 유사도 측정에 사용하는 Dynamic Time Warping 기법을 활용하여 디지털 오디오 간의 유사도 측정을 하였다. 제안된 유사도 측정을 통한 내용기반 디지털 오디오검색 방법의 검증을 위해 같은 장르에서 무작위 추출된 100곡에서 시행한 90번의 검색은 모두 성공했다. 검색에 사용된 90개의 디지털 오디오는 10개의 디지털 오디오를 압축방식과 비트율을 다르게 조합하여 만들었다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.04b
• /
• pp.61-63
• /
• 2001

최근들어, 인터넷상의 E-business나 쇼핑몰사이트와 같은 웹 사이트에서 멀티미디어 정보를 많이 사용하고 있다. 멀티미디어 정보 중에서도 이미지 정보가 가장 많이 사용되고 있으며, 이는 사용자들이 가장 많이 접하는 정보이다. 기존의 이미지 검색 기법은 내용 기반 검색이나 키워드를 이용한 검색 방법을 지원하지만, 사용자의 의도를 적용하지는 못하고 있다. 본 논문에서는 웹에서 사용자가 이미지를 검색하고 접근하는 패턴을 이미지의 칼라와 질감을 특징으로 한 벡터를 기반으로 시스템에 학습 시키고 사용자의 검색 성향을 분석하여 시스템에 적용한다. 이미지 검색의 효율을 높이기 위하여 질감을 기반으로 비트 벡터 인덱스(bit vector index) 기법을 적용하며, 인덱스에 의한 이미지 자동 분류 기법을 제안한다. 또한 이미지 칼라의 정보를 영역별로 추출하여 칼라 부분매칭 검색을 가능하게 한다. 이러한 이미지 검색 시스템을 사용하는 사용자의 정보를 시스템에 학습시키고 학습된 결과를 이용해서 사용자가 검색 하고자 하는 이미지 정보에 편리성을 제공하고 검색의 효율성을 증대시킨다.

• The KIPS Transactions:PartD
• /
• v.9D no.6
• /
• pp.1025-1042
• /
• 2002

The goal of this study is to generate a structured document which improves the performance of an information retrieval system by using thesaurus, information on relations between words (terms), and to study on the technique for searching this structured document. In order to accomplish this goal, we propose a DODI (Depth -Oriented Decomposition Index) technique for the structured document and an algorithm to search for related information efficient]y through this index technique that uses a thesaurus. We establish a storage system by which the structured document generated by this index technique is saved in a database through OpenXML and XML documents are generated through ForXML methods.

• Spatial Information Research
• /
• v.21 no.4
• /
• pp.7-14
• /
• 2013

Recently, R&D(Research and Development) is processing actively on range search in the road network environments. However, the existing representative range search techniques have shortcomings in that the greater the number of POI's, the more increased storage space or the more increased search time due to inefficient search process. Accordingly, In this paper, we proposed a range search technique using QRMP(QR-tree using Middle Point) to solve the problems of conventional range search techniques. In addition, we made a formula to obtain the total size of the storage space for QRMP and proved the excellence of the range search technique proposed in this paper through the experiment using actual road networks and POI data.

• The KIPS Transactions:PartD
• /
• v.11D no.2
• /
• pp.247-258
• /
• 2004

Recently, there have been researches on storage and retrieval technique of moving objects, which are highly concerned by user in database application area such as video databases, spatio-temporal databases, and mobile databases. In this paper, we propose a new signature-based indexing scheme which supports similar sub-trajectory retrieval at well as good retrieval performance on moving objects trajectories. Our signature-based indexing scheme is classified into concatenated signature-based indexing scheme for similar sub-trajectory retrieval, entitled CISR scheme and superimposed signature-based indexing scheme for similar sub-trajectory retrieval, entitled SISR scheme according to generation method of trajectory signature based on trajectory data of moving object. Our indexing scheme can improve retrieval performance by reducing a large number of disk access on data file because it first scans all signatures and does filtering before accessing the data file. In addition, we can encourage retrieval efficiency by appling k-warping algorithm to measure the similarity between query trajectory and data trajectory. Final]y, we evaluate the performance on sequential scan method(SeqScan), CISR scheme, and SISR scheme in terms of data insertion time, retrieval time, and storage overhead. We show from our experimental results that both CISR scheme and SISR scheme are better than sequential scan in terms of retrieval performance and SISR scheme is especially superior to the CISR scheme.

• The KIPS Transactions:PartA
• /
• v.11A no.5
• /
• pp.333-340
• /
• 2004

In order that routers forward a packet to its destination, they perform IP address lookup which determines the next hop according to the packet's destination address. In designing high speed routers, IP address lookup is an important issue. In order to design high speed routers, this paper proposes a parallel IP lookup scheme which consists of several IP lookup engines without any modification of already fabricated indirect IP lookup chipsets. Also, we propose a simple rule for partitioning IP prefix entries In an overall forwarding table among several IP lookup engines. And we evaluate the performance of the proposed scheme in terms of the memory size required for storing lookup information and the number of memory accesses on constructing the forwarding table. With additional hardware logics, the proposed scheme can reduce about 30% of the required memory size and 80% of the memory access counts.