• 정보처리학회논문지B
• /
• 제16B권3호
• /
• pp.247-256
• /
• 2009

대부분의 잠정적 적합피드백기법들은 질의에 대해 검색된 상위검색문서들이 적합하다고 가정하고, 그 문서들을 질의 확장을 위한 피드백 문서로 이용하고 있다. 그러나 초기검색결과에는 상당한 양의 부적합 문서를 포함하고 있는 것이 현실이다. 이 논문에서는 보다 좋은 피드백 문서를 선택하기 위해서 중첩클러스터를 이용한 피드백문서의 재샘플링 기법을 제안한다. 주요 아이디어는 질의 중심적인 초기검색문서집합에 대해서 중첩이 허용된 문서클러스터를 이용하여 문서들 사이의 관계를 반영하여 질의에 핵심역할을 하는 지배적 문서를 찾고, 이 문서들을 반복적으로 피드백 하여 질의가 내포하는 핵심 주제를 강조하는 것이다. 대규모 실험집합인 TREC GOV2와 WT10g에 대한 실험비교에서, 최근 잠정적 적합피드백 기법들 중에서 가장 좋은 성능을 보이고 있는 적합모델보다 재샘플링기법이 우수한 성능향상을 보였다. 제안기법에 대한 검증을 위해서 피드백문서에 포함된 적합문서의 정도를 나타내는 적합밀도를 측정하였다. 재샘플링 기법이 TREC 실험집합에 대해서 적합모델에 비해 높은 적합밀도를 보였고, 이 결과 적합피드백에서 검색성능을 향상시키게 되었다. 이는 제안 기법이 잠정적 적합피드백에서 유효한 방법임을 알 수 있다.

• Korean Journal of Radiology
• /
• 제22권2호
• /
• pp.281-290
• /
• 2021

Objective: To assess the performance of content-based image retrieval (CBIR) of chest CT for diffuse interstitial lung disease (DILD). Materials and Methods: The database was comprised by 246 pairs of chest CTs (initial and follow-up CTs within two years) from 246 patients with usual interstitial pneumonia (UIP, n = 100), nonspecific interstitial pneumonia (NSIP, n = 101), and cryptogenic organic pneumonia (COP, n = 45). Sixty cases (30-UIP, 20-NSIP, and 10-COP) were selected as the queries. The CBIR retrieved five similar CTs as a query from the database by comparing six image patterns (honeycombing, reticular opacity, emphysema, ground-glass opacity, consolidation and normal lung) of DILD, which were automatically quantified and classified by a convolutional neural network. We assessed the rates of retrieving the same pairs of query CTs, and the number of CTs with the same disease class as query CTs in top 1-5 retrievals. Chest radiologists evaluated the similarity between retrieved CTs and queries using a 5-scale grading system (5-almost identical; 4-same disease; 3-likelihood of same disease is half; 2-likely different; and 1-different disease). Results: The rate of retrieving the same pairs of query CTs in top 1 retrieval was 61.7% (37/60) and in top 1-5 retrievals was 81.7% (49/60). The CBIR retrieved the same pairs of query CTs more in UIP compared to NSIP and COP (p = 0.008 and 0.002). On average, it retrieved 4.17 of five similar CTs from the same disease class. Radiologists rated 71.3% to 73.0% of the retrieved CTs with a similarity score of 4 or 5. Conclusion: The proposed CBIR system showed good performance for retrieving chest CTs showing similar patterns for DILD.

• 한국정보과학회논문지:컴퓨팅의 실제 및 레터
• /
• 제9권2호
• /
• pp.158-165
• /
• 2003

영상간의 유사도는 일반적으로 영상으로부터 추출한 특징벡터간의 벡터공간상의 거리를 계산해서 판단한다. 그러나 이러한 특징벡터가 유사도 계산을 위한 하나의 방법이지만 항상 인간의 유사도 개념을 충실히 반영하지는 않는다. 그러므로 현존하는 대부분의 영상검색시스템들은 각 특징간의 중요도를 선정하여 유사도에 반영하는 방법을 사용하고 있다. 본 논문에서는 영상검색을 위한 새로운 초기 가중치 설정과 갱신 알고리즘을 제안한다. 이를 위해서 먼저 데이터 베이스 영상을 인간의 인지도 판단에 의해 그룹화 한 후, 내부질의와 외부질의를 수행하고, 검색된 영상중 유사한 영상이 어느 그룹에 속하는지 알아내어 각 영상별로 유사도 계산에 필요한 최적 특징 가중치를 계산한다. 2000개의 영상 데이타에 대한 실험을 통해서 제안된 알고리즘의 우수성을 보인다.

• 한국측량학회지
• /
• 제31권6_2호
• /
• pp.585-592
• /
• 2013

CBIR is an effective tool to search and extract image contents in a large remote sensing image database queried by an operator or end user. However, as imaging principles are different by sensors, their visual representation thus varies among image modality type. Considering images of various modalities archived in the database, image modality difference has to be tackled for the successful CBIR implementation. However, this topic has been seldom dealt with and thus still poses a practical challenge. This study suggests a cross modality CBIR (termed as the CM-CBIR) method that transforms given query feature vector by a supervised procedure in order to link between modalities. This procedure leverages the skill of analyst in training steps after which the transformed query vector is created for the use of searching in target images with different modalities. Current initial results show the potential of the proposed CM-CBIR method by delivering the image content of interest from different modality images. Despite its retrieval capability is outperformed by that of same modality CBIR (abbreviated as the SM-CBIR), the lack of retrieval performance can be compensated by employing the user's relevancy feedback, a conventional technique for retrieval enhancement.

• 한국정보통신학회논문지
• /
• 제9권8호
• /
• pp.1828-1833
• /
• 2005

이질적인 생물 데이터베이스의 통합은 데이터간의 연계 분석의 필요성이 높아짐에 따라 중요한 문제로 대두되고 있다. 그러나 이러한 데이터베이스들은 초기에 이질적 환경에서 각기 다른 목적에 의해 생성되므로 포맷, 설계자가 불일치하는 등 여러 가지 문제점으로 인해 통합하는데 어려움이 따른다. 그러므로 이질적인 데이터베이스의 통합을 위해서는 초기단계의 설계가 무엇보다도 중요하다. 본 논문에서는 대표적인 핵산 데이터베이스인 Genbank와 단백질 데이터베이스인 Swiss-Prot을 통합하기 위해 ER 모델을 사용하여 개념적 모델을 보인 후, 이를 합병하여 통합모델을 제시한다. 또한, 핵산-단백질 자료로 연계되는 정보를 통합 서비스할 수 있는 시스템 구조를 제안한다. 제안된 바이오 게이트웨이 시스템은 개념적 설계 단계에서 가장 원자적인 단위로 분할하여 모델링 함으로써 정교한 질의 처리가 가능하고, 사용자가 상세 조건을 알고 있을 경우에 기존의 검색시스템과 달리 여러 번의 검색 과정을 거치지 않고, 단시간에 원하는 결과를 얻을 수 있다는 장점을 지닌다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권3호
• /
• pp.894-907
• /
• 2022

The features extracted by convolutional neural networks are more descriptive of images than traditional features, and their convolutional layers are more suitable for retrieving images than are fully connected layers. The convolutional layer features will consume considerable time and memory if used directly to match an image. Therefore, this paper proposes a feature weighting and region integration method for convolutional layer features to form global feature vectors and subsequently use them for image matching. First, the 3D feature of the last convolutional layer is extracted, and the convolutional feature is subsequently weighted again to highlight the edge information and position information of the image. Next, we integrate several regional eigenvectors that are processed by sliding windows into a global eigenvector. Finally, the initial ranking of the retrieval is obtained by measuring the similarity of the query image and the test image using the cosine distance, and the final mean Average Precision (mAP) is obtained by using the extended query method for rearrangement. We conduct experiments using the Oxford5k and Paris6k datasets and their extended datasets, Paris106k and Oxford105k. These experimental results indicate that the global feature extracted by the new method can better describe an image.

• 한국CDE학회논문집
• /
• 제19권4호
• /
• pp.390-401
• /
• 2014

ISO 15926 is an international standard for integration of lifecycle data for process plants including oil and gas facilities. From the viewpoint of information modeling, ISO 15926 Parts 2 provides the general data model that is designed to be used in conjunction with reference data. Reference data are standard instances that represent classes, objects, properties, and templates common to a number of users, process plants, or both. ISO 15926 Parts 4 and 7 provide the initial set of classes, objects, properties and the initial set of templates, respectively. User-defined reference data specific to companies or organizations are defined by inheriting from the initial reference data and the initial set of templates. In order to support the extension of reference data and templates, an editor that provides creation, deletion and modification functions of user-defined reference data is needed. In this study, an editor for reference data based on ISO 15926 was developed. Sample reference data were encoded in OWL (web ontology language) according to the specification of ISO 15926 Part 8. iRINGTools and dot15926Editor were benchmarked for the design of GUI (graphical user interface). Reference data search, creation, modification, and deletion functions were implemented with XML (extensible markup language) DOM (document object model), and SPARQL (SPARQL protocol and RDF query language).

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권6호
• /
• pp.2429-2449
• /
• 2018

Dynamic data possession verification is a common requirement in cloud storage systems. After the client outsources its data to the cloud, it needs to not only check the integrity of its data but also verify whether the update is executed correctly. Previous researches have proposed various schemes based on Merkle Hash Tree (MHT) and implemented some initial improvements to prevent the tree imbalance. This paper tries to take one step further: Is there still any problems remained for optimization? In this paper, we study how to raise the efficiency of data dynamics by improving the parts of query and rebalancing, using a new data structure called Rank-Based Merkle AVL Tree (RB-MAT). Furthermore, we fill the gap of verifying multiple update operations at the same time, which is the novel batch updating scheme. The experimental results show that our efficient scheme has better efficiency than those of existing methods.

• 한국도서관정보학회지
• /
• 제22권
• /
• pp.373-401
• /
• 1995

This study aims to improve subject retrieval by constructing the hypertext-based thesaurus to provide a browsing interface to a thesaurus. The experimental system used IBM 486 DXII as a hardware, C++ language as a programming language and Hangul Window 3.1 version as a user interface. The results of this study are summarized as follows : (1) The experimental hypertext-based thesaurus can be used as an efficient search aid for query formulation for the retrieval of bibliographic information. (2) The initial access to the hypertext-based thesaurus is via a keyword index. This index is consist of all the words used to form thesaurus terms, whether descriptors or nondescriptors. (3) Hypertext-based thesaurus allow bookmark button and history button to alleviating the problem of disorientation. (4) This system allow an end-user to view a rich variety of inter-term relationships and a complete conceptus of associations through the information space in a nonsequential manner.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송공학회 2009년도 IWAIT
• /
• pp.164-169
• /
• 2009

As one of the most interesting scenes, landmarks constitute a large percentage of the vast amount of scene images available on the web. On the other hand, a specific "landmark" usually has some characteristics that distinguish it from surrounding scenes and other landmarks. These two observations make the task of accurately estimating geographic information from a landmark image necessary and feasible. In this paper, we propose a method to identify landmark location by means of landmark recognition in view of significant viewpoint, illumination and temporal variations. We use GPS-based clustering to form groups for different landmarks in the image dataset. The images in each group rather fully express the possible views of the corresponding landmark. We then use a combination of edge and color histogram to match query to database images. Initial experiments with Zubud database and our collected landmark images show that is feasible.