• Proceedings of the Korea Database Society Conference
• /
• 1997.10a
• /
• pp.463-471
• /
• 1997

비디오 데이타는 다양하고 방대한 양의 의미를 포함하고 있어 효율적인 내용기반 검색을 지원하기 위해서는 비디오 데이타를 기술하는 구조적이고 체계화된 형태의 메타데이타가 요구된다. 이러한 메타데이타는 검색 시 색인과 같은 역할을 수행하게 되므로 내용 기반검색의 가장 기본적이고 필수적인 데이타이다. 본 논문에서는 뉴스 응용 분야(News On Demand：NOD)를 적용한 비디오 데이터베이스 시스템의 효율적인 내용 기반 검색을 위한 메타데이타를 분류하고, Rambaugh의 OMT기법을 이용하여 메타데이타를 모델링한 후 질의 유형에 따라 모델의 접근 경로를 검사하여 모델을 검증하였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2010.06a
• /
• pp.17-18
• /
• 2010

• The Journal of Information Technology and Database
• /
• v.8 no.1
• /
• pp.117-128
• /
• 2001

Recently there are many researches of text mining to find interesting patterns or association rules from mass textual documents. However, the words extracted from informal documents are tend to be irregular and there are too many general words, so if we use pre-exist method, we would have difficulty in retrieving knowledge information effectively. In this paper, we propose a new feature extraction method to classify mass documents using association rule based on unsupervised learning technique. In experiment, we show the efficiency of suggested method by extracting features and classifying of documents.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2005.05a
• /
• pp.125-128
• /
• 2005

이동 객체 데이터베이스 시스템에서 시공간 조인 연산은 이동 객체들의 결합을 위한 중요한 연산이며 수행 시간은 이동 객체의 수가 증가함에 따라 기하급수적으로 증가한다. 그러므로 효과적인 시공간 조인 연산이 필수적이다. 본 논문에서는 기존의 공간 조인에서 활용되었던 기법들을 이동객체 조인에 적용하였다. 이동 객체의 궤적에 대한 정보를 잘 유지하고 있는 시공간 색인인 TB-Tree 를 이용한 깊이 우선 탐색 기반과 넓이 우선 탐색 기반 TB-Tree 조인에 대한 알고리즘들을 제시하고 구현한 알고리즘들의 성능 비교한 실험 결과를 제시한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2007.10a
• /
• pp.1-2
• /
• 2007

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2000.08a
• /
• pp.297-300
• /
• 2000

본 논문에서는 MPEG 비디오 스트림을 분석하여 DCT DC 계수를 추출하고 이들로 구성된 DC 이미지로부터 제안하는 robust feature를 이용하여 shot을 구하고 각 feature들의 통계적 특성을 이용하여 스트림의 특징에 따라 weight를 부가하여 구해진 characterizing value의 시간변화량을 구한다. 구해진 변화량의 local maxima와 local minima는 MPEG 비디오 스트림에서 각각 가장 특징적인 frame과 평균적인 frame을 나타낸다. 이 순간의 frame을 구함으로서 효과적이고 빠른 시간 내에 key frame을 추출한다. 추출되어진 key frame에 대하여 원영상을 복원한 후, 색인을 위하여 다수의 parameter를 구하고 사용자가 질의한 영상에 대해서 이들 파라메터를 구하여 key frame들과 가장 유사한 대표영상들을 검색한다.

• Proceedings of the Korean Society for Information Management Conference
• /
• 2014.08a
• /
• pp.17-20
• /
• 2014

본 연구는 동시출현 단어분석 기법을 이용하여 최근 전세계적으로 많은 주목을 받고 있는 빅데이터(Big Data) 관련 연구 동향과 연구 영역을 분석하는 것을 목적으로 한다. 이를 위하여 인용색인데이터베이스인 Web of Science SCIE(Science Citation Index Expanded)에서 분석 대상 논문을 수집하였다. 논문 수집을 위한 검색식은 은 Title(논문 제목), Abstract(초록), Author Keywords(저자 키워드), Keywords $Plus^{(R)}$의 네 가지 필드를 동시에 검색하는 주제어(topic)가 "big data"를 포함하고 있는 논문 563편을 대상으로 동시출현단어 분석을 수행하였다.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.43 no.4 s.310
• /
• pp.88-102
• /
• 2006

In multimedia information retrieval, multimedia data are represented as vectors in high dimensional space. To search these vectors effectively, a variety of indexing methods have been proposed. However, the performance of these indexing methods degrades dramatically with increasing dimensionality, which is known as the dimensionality curse. To resolve the dimensionality curse, dimensionality reduction methods have been proposed. They map feature vectors in high dimensional space into the ones in low dimensional space before indexing the data. This paper proposes a method for dimensionality reduction based on a function approximating the Euclidean distance, which makes use of the norm and angle components of a vector. First, we identify the causes of the errors in angle estimation for approximating the Euclidean distance, and discuss basic directions to reduce those errors. Then, we propose a novel method for dimensionality reduction that composes a set of subvectors from a feature vector and maintains only the norm and the estimated angle for every subvector. The selection of a good reference vector is important for accurate estimation of the angle component. We present criteria for being a good reference vector, and propose a method that chooses a good reference vector by using Levenberg-Marquardt algorithm. Also, we define a novel distance function, and formally prove that the distance function lower-bounds the Euclidean distance. This implies that our approach does not incur any false dismissals in reducing the dimensionality effectively. Finally, we verify the superiority of the proposed method via performance evaluation with extensive experiments.

• KIPS Transactions on Computer and Communication Systems
• /
• v.7 no.8
• /
• pp.195-202
• /
• 2018

With the development of location-based applications such as smart phones and GPS navigation, active research is being conducted to protect location and trajectory privacy. To receive location-related services, users must disclose their exact location to the server. However, disclosure of users' location exposes not only their locations but also their trajectory to the server, which can lead to concerns of privacy violation. Furthermore, users request from the server not only location information but also multimedia information (photographs, reviews, etc. of the location), and this increases the processing cost of the server and the information to be received by the user. To solve these problems, this study proposes the EGTC (Enhanced Grid-based Trajectory Cloaking) technique. As with the existing GTC (Grid-based Trajectory Cloaking) technique, EGTC method divides the user trajectory into grids at the user privacy level (UPL) and creates a cloaking region in which a random query sequence is determined. In the next step, the necessary information is received as index by considering the sub-grid cell corresponding to the path through which the user wishes to move as c(x,y). The proposed method ensures the trajectory privacy as with the existing GTC method while reducing the amount of information the user must listen to. The excellence of the proposed method has been proven through experimental results.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.16 no.12
• /
• pp.8649-8653
• /
• 2015

A number of video indexing and retrieval algorithms have been proposed to manage large video databases efficiently. The video similarity measure is one of most important technical factor for video content management system. In this paper, we propose the luminance characteristics model to measure the video similarity efficiently. Most algorithms for video indexing have been commonly used histograms, edges, or motion features, whereas in this paper, the proposed algorithm is employed an efficient similarity measure using the luminance projection. To index the video sequences effectively and to reduce the computational complexity, we calculate video similarity using the key frames extracted by the cumulative measure, and compare the set of key frames using the modified Hausdorff distance. Experimental results show that the proposed luminance projection model yields the remarkable improved accuracy and performance than the conventional algorithm such as the histogram comparison method, with the low computational complexity.