• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.36S no.10
• /
• pp.113-122
• /
• 1999

When we want to retrieve the most similar image from the image database, the color histogram intersection, shape feature and texture feature comparing method are used as a metric to measure the similarity. In order to increase the accuracy of retrievals, we need to integrate two different features. In this paper, the histogram intersection and shape based block histogram intersection method are used. This method results in a high efficient algorithm that meets a similar accuracy and a relatively fast retrieval speed compared to the method of integration of two different features. The Proposed algorithm is tested on retrievals of image database consisting of various 600 images and we implemented that the proposed algorithm gives fast, high efficiency and reliability compared to others.

• Journal of KIISE:Databases
• /
• v.33 no.5
• /
• pp.538-550
• /
• 2006

XML Keyword search enables us to get information easily without knowledge of structure of documents and returns specific and useful partial document results instead of whole documents. Element level query processing makes it possible, but computational complexity, as the number of documents grows, increases significantly overhead costs. In this paper, we present document-level ranking scheme over XML documents which predicts results of element-level processing to reduce processing cost. To do this, we propose the notion of 'keyword proximity' - the correlation of keywords in a document that affects the results of element-level query processing using path information of occurrence nodes and their resemblances - for document ranking process. In benefit of document-centric view, it is possible to reduce processing time using ranked document list or filtering of low scored documents. Our experimental evaluation shows that document-level processing technique using ranked document list is effective and improves performance by the early termination for top-k query.

• The Journal of the Acoustical Society of Korea
• /
• v.23 no.7
• /
• pp.532-539
• /
• 2004

In this Paper, we propose a content-based music information retrieval (MIR) system base on the query-by-example (QBE) method. The proposed system is implemented to retrieve queried music from a dataset where 60 music samples were collected for each of the four genres in Classical, Hiphop. Jazz. and Reck. resulting in 240 music files in database. From each query music signal, the system extracts 60 dimensional feature vectors including spectral centroid. rolloff. flux base on STFT and also the LPC. MFCC and Beat information. and retrieves queried music from a trained database set using Euclidean distance measure. In order to choose optimum features from the 60 dimension feature vectors, SFS method is applied to draw 10 dimension optimum features and these are used for the Proposed system. From the experimental result. we can verify the superior performance of the proposed system that provides success rate of 84% in Hit Rate and 0.63 in MRR which means near 10% improvements over the previous methods. Additional experiments regarding system Performance to random query Patterns (or portions) and query lengths have been investigated and a serious instability problem of system Performance is Pointed out.

• Journal of Korea Multimedia Society
• /
• v.3 no.6
• /
• pp.575-584
• /
• 2000

This paper presents a new image retrieval system with color and morphological informations based on binary sets. Each of them can be obtained from color binary sets and regional segmentation separately. For retrieval processes, the candidate images are decided by comparing color and their image binary sets of the database with query images. Particularly, it is possible that the retrieval of similar-measurements has a weight of color spatial distribution and its objective morphological features. We proposed a new idea for performing simply the complicated similar-measurement of candidated images to improve queried processes. The retrieval method using spatial and morphological features is shown with the effectiveness on the result of implementation on database with 3,000 images.

• The KIPS Transactions:PartB
• /
• v.9B no.5
• /
• pp.701-706
• /
• 2002

A set of color features has been efficiently used to measure the similarity of given images. However, the size of the color features is too large to implement an indexing scheme effectively. In this paper a new method is proposed to retrieve similar images using an interpolated color histogram. The idea is similar to the already reported methods that use the distributions of color histograms. The new method is different in that simplified color histograms decide the similarity between a query image and target images. In order to represent the distribution of the color histograms, the best order of interpolated polynomial has been simulated. After a histogram distribution is represented in a polynomial form, only a few number of polynomial coefficients are indexed and stored in a database as a color descriptor. The new method has been applied to real images and achieved satisfactory results.

• The KIPS Transactions:PartD
• /
• v.10D no.5
• /
• pp.753-762
• /
• 2003

Many Methodologies have proposed for component retrieval. Among them, thesaurus concept has introduced for similar component retrieval. This paper classified classes by concept according to inheritance relation for efficient retrieval of component, and applied fuzzy logic to thesaurus method and constructed object-oriented thesaurus. Proposed method could express category between concepts automatically, and calculate fuzzy degree between classes by comparing matching and mismatching degree to each class and category and construct thesaurus. Component retrieval is that using classes of component, candidate components are retrieved according to priority order using fuzzy similarity. Also, we improved retrieval performance by thesaurus greatly, setting critical of most suitable through simulation.

• Journal of KIISE:Databases
• /
• v.30 no.2
• /
• pp.134-144
• /
• 2003

Spatial relationships among objects are one of the important ingredients for expressing constraints of an image in image or multimedia retrieval systems. In this paper, we propose a unified image retrieval scheme using spatial relationships among objects and their features. The proposed scheme is especially effective in computing similarity between query image and images in the database. Also, objects and their spatial relationships are captured and annotated in XML. It could give better precision and flexibility in retrieving images from database. Finally, we have implemented a prototype system for retrieving images based on proposed technique and showed some of the experiment results.

• Annual Conference on Human and Language Technology
• /
• 2008.10a
• /
• pp.172-176
• /
• 2008

인터넷 쇼핑몰과 같은 웹사이트에서 FAQ 검색의 중요성은 갈수록 증가되고 있다. 일반적으로 FAQ 검색을 하기 위해서 사용자가 직접 FAQ 리스트에서 FAQ를 찾거나 키워드 검색을 통해 FAQ를 찾는다. 하지만 이 방법은 FAQ 리스트를 찾는데 시간이 오래 걸리고 사용자가 만족하는 결과를 보여주지 못하는 경우가 많다. 이를 해결하기 위해 사용자의 자연어 질의에 대해 자동으로 FAQ를 찾아주는 FAQ 시스템이 필요하다. 자동화된 FAQ 시스템은 사용자의 질의 문장에 대해 FAQ 목록 중에 가장 유사한 FAQ 문장을 찾아준다. 이를 위해 각 문장을 유니그램과 바이그램 단어 집합으로 표현하여 문장 간의 유사도를 계산한다. 본 논문에서는 유니그램과 바이그램 단어 집합뿐만이 아니라 한국어 문형을 기반으로 한 문장 정규화를 통해 단어의 문장 성분 정보와 성분 내 위치 정보를 이용하여 문장을 표현함으로써 자동화된 FAQ 시스템의 성능을 향상 시키는 것을 목표로 한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2007.10c
• /
• pp.317-321
• /
• 2007

가공해서 사용하는 정보량이 많아질수록 원하는 정보를 찾는데 더 많은 노력이 필요하게 되었다. 따라서 사람들은 인터넷상에서 원하는 정보를 검색하는 여러 방법들을 고안해 왔으며 이렇게 구현된 검색 알고리즘은 검색 질의와 유사한 문서가 대중에게 얼마나 관심을 받고 있는지 그 정도에 따라 검색순위 상위에 링크된다. 하지만 웹 문서의 폭발적인 증가로 해당 질의에 대한 검색 결과 문서의 수가 급격히 늘어나면서 사용자를 만족시키기가 점점 어렵게 되었다. 이러한 문제를 해결하기 위한 방안으로 네티즌들이 직접 정보를 생산, 공유하고 이들이 모여 활동할 수 있는 커뮤니티를 형성하기 시작했다. 이 논문에서는 정보의 공유를 목적으로 하는 커뮤니티를 인터넷상의 표준화된 웹 서비스(Web Services) 기술인 UDDI에 저장하고, SOAP 프로토콜을 이용하여 플랫폼에 상관없이 사용자 검색 질의와 가장 유사한 커뮤니티를 검색하여 제공하는 방법을 제안한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10b
• /
• pp.247-249
• /
• 1998

최근 들어 내용기반의 이미지 검색을 지원하기 위한 방법으로, 특징 벡터를 이용한 유사 질의 연구가 활발히 진행되고 있다. 이러한 유사 질의를 효율적으로 지원하기 위해서는 고차원 공간상에 존재하는 점 데이터나 공간 데이터를 효과적으로 색인할 수 있는 색인 기법이 필요하다. 하지만 R*-트리를 바탕으로 하는 기존의 방법들은 고차원 데이터에 대해서 차원의 증가함에 따라 검색 시간이 급격하게 증가하는 문제점을 안고 있다. 이러한 문제는 데이터의 클러스터링에 기반을 둔 기존의 방법들이 차원이 증가함에 따라 데이터를 제대로 클러스터링하지 못하기 때문에 발생하며, 따라서 이를 해결하기 위해서는 효과적인 클러스터링 기법이 필요하다. 본 논문에서는 하나의 최소 한계 영역(minimum bounding region)에 속하는 개체들의 응집 정도와 최소 한계 영역들간의 결합 정도를 고려하여 효과적으로 클러스터링하는 방안을 제안한다. 또한 이러한 클러스터링 기법을 수용하기 위한 색인 기법을 간략히 제시한다