• KIPS Transactions on Software and Data Engineering
• /
• v.3 no.12
• /
• pp.531-536
• /
• 2014

With the very large quantity of information available on the Internet, techniques for dealing with the abundance of documents have become increasingly necessary but the problem of processing information in the documents is still technically challenging and remains under study. Automatic document summary techniques have been considered as one of critical solutions for processing documents to retain the important points and to remove duplicated contents of the original documents. In this paper, we propose a document summarization technique that uses a fuzzy theory. Proposed summary technique solves the ambiguous problem of various features determining the importance of the sentence and the experiment result shows that the technique generates better results than other previous techniques.

• Korean Management Science Review
• /
• v.27 no.3
• /
• pp.15-31
• /
• 2010

We consider a reverse logistics network design problem for recycling. The problem consists of three stages of transportation. In the first stage products are transported from retrieval centers to disassembly centers. In the second stage disassembled modules are transported from disassembly centers to processing centers. Finally, in the third stage modules are transported from either processing centers or a supplier to a manufacturer, a recycling site, or a disposal site. The objective is to design a network which minimizes the total transportation cost. We design a cooperative coevolutionary algorithm to solve the problem. First, the problem is decomposed into three subproblems each of which corresponds to a stage of transportation. For subproblems 1 and 2, a population of chromosomes is constructed. Each chromosome in the population is coded as a permutation of integers and an algorithm which decodes a chromosome is suggested. For subproblem 3, an heuristic algorithm is utilized. Then, a performance evaluation procedure is suggested which combines the chromosomes from each of two populations and the heuristic algorithm for subproblem 3. An experiment was carried out using test problems. The experiments showed that the cooperative coevolutionary algorithm generally tends to show better performances than the previous genetic algorithm as the problem size gets larger.

• Journal of Korea Multimedia Society
• /
• v.7 no.3
• /
• pp.293-307
• /
• 2004

The SMIL(Synchronized Multimedia Integration Language) documents are represented as logical structure information, spatial layout structure information, temporal synchronization structure information and hyperlink structure information, according as the structural characteristics of SMIL documents based on XML. This paper proposes the effective modeling and query method for the multi -structure information of inherent SMIL documents. In particular, we present the object-oriented modeling by using UML class diagram in order to represent the objects classes for the structured information of SMIL documents, and the hierarchical structure and the relationships for the objects classes. In addition, the objects classes definition is specified in compliance with SQL3 for database standard language. We also propose the access method and the query representation for hierarchical structure in order to retrieve efficiently the structural objects of SMIL documents.

• Proceedings of the Korea Contents Association Conference
• /
• 2003.05a
• /
• pp.356-360
• /
• 2003

There have been many considerations to develop new content services that integrate a variety of contents databases being already constructed and then produce new content services which are more valuable than existing services in many applications such as Internet portal, EC, and CRM. By doing the above thing, the burden of searching databases to access interesting databases and service applications can be reduced and the database availability of users is also enhanced through a single view integrating multiple contents database. This paper presents implementation details of the query processing system that is a core component of the database integration system, which can construct a virtual database that integrates databases being managed by multiple heterogeneous database systems using XML data model and support a quay facility on the integrated database.

• Journal of KIISE:Software and Applications
• /
• v.30 no.3_4
• /
• pp.259-267
• /
• 2003

Though naive Bayes text classifiers are widely used because of its simplicity, the techniques for improving performances of these classifiers have been rarely studied. In this paper, we propose and evaluate some general and effective techniques for improving performance of the naive Bayes text classifier. We suggest document model based parameter estimation and document length normalization to alleviate the Problems in the traditional multinomial approach for text classification. In addition, Mutual-Information-weighted naive Bayes text classifier is proposed to increase the effect of highly informative words. Our techniques are evaluated on the Reuters21578 and 20 Newsgroups collections, and significant improvements are obtained over the existing multinomial naive Bayes approach.

• Proceedings of the KSRS Conference
• /
• 2004.10a
• /
• pp.590-593
• /
• 2004

Extraction of ocean surface current velocity offers important physical oceanographic parameters especially on understanding ocean environment. Although Remote Sensing techniques were highly developed, the investigation of ocean surface current using Synthetic Aperture Radar (SAR) is not an easy task. This paper presents the results of ocean surface current observation using Doppler Centroid of ERS-1 SAR data obtained off the coast of Korea peninsula. We employed the concept, in which Doppler frequency shift and the ocean surface current are closely related, to evaluate ocean surface current. Moving targets cause Doppler frequency shift of the back scattered radar waves of SAR, thus the line-of-sight velocity component of the scatters can be evaluated. The Doppler frequency shift can be measured by estimating the difference between Doppler Centroid of raw SAR data and reference Doppler Centroid. Theoretically, the Doppler Centroid is zero; however, squinted antenna which is affected by several physical factors causes Doppler Centroid to be nonzero. The reference Doppler Centroid can be obtained from measurements of sensor trajectory, attitude and Earth model. The estimated Doppler Centroid was compensated by considering the accurate attitude estimation of ERS-1 SAR. We could verify the correspondence between the estimated ocean surface current and observed in-situ data in the error bound.

• Proceedings of the KSRS Conference
• /
• v.2
• /
• pp.610-614
• /
• 2006

In this paper the advanced method of geodesic active contours was developed for the task of building detection from aerial and satellite images. Automatic extraction of man-made structures including buildings, building blocks or roads from remote sensing data is useful for land use mapping, scene understanding, robotic navigation, image retrieval, surveillance, emergency management procedures, cadastral etc. A level set method based on a region-driven segmentation model was implemented with which building boundaries were detected, through this curve propagation technique. The essence of this approach is to optimize the position and the geometric form of the curve by measuring information along that curve, and within the regions that compose the image partition. To this end, one can consider uniform intensities inside objects and the background. Thus, given an initial position of the curve, one can determine global, region-driven functions and provide a statistical description of the inside and outside object area. The calculus of variations and a gradient descent method was used to optimize the variational functional by an iterative steady state process. Experimental results demonstrate the potential of the proposed processing scheme.

• ETRI Journal
• /
• v.38 no.3
• /
• pp.518-528
• /
• 2016

The searching of an extensive document database for documents that are locally similar to a given query document, and the subsequent detection of similar regions between such documents, is considered as an essential task in the fields of information retrieval and data management. In this paper, we present a framework for such a task. The proposed framework employs the method of short-read mapping, which is used in bioinformatics to reveal similarities between genomic sequences. In this paper, documents are considered biological objects; consequently, edit operations between locally similar documents are viewed as an evolutionary process. Accordingly, we are able to apply the method of evolution tracing in the detection of similar regions between documents. In addition, we propose heuristic methods to address issues associated with the different stages of the proposed framework, for example, a frequency-based fragment ordering method and a locality-aware interval aggregation method. Extensive experiments covering various scenarios related to the search of an extensive document database for documents that are locally similar to a given query document are considered, and the results indicate that the proposed framework outperforms existing methods.

• Annual Conference on Human and Language Technology
• /
• 2001.10d
• /
• pp.183-190
• /
• 2001

본 논문은 확률기반 자연어 검색 시스템 POSNIR/E를 이용한 여러 가지 유사 적합성 피드백 방법들이 검색 시스템의 성능 향상에 기여할 수 있는 정도를 보여주고, 확률 기반 정보 검색 시스템에 적합한 유사 적합성 피드백 수행 방법을 제시한다. POSNIR/E는 한국어 자연어 검색 시스템, POSNIR를 기반으로 만들어진 영어 자연어 검색 시스템이다. 이 시스템은 성능 향상을 위한 질의 확장의 방법으로 검색 단계에서 유사 적합성 피드백을 사용한다. 검색 단계에서 영어 태거에 의해 태깅된 사용자 질의로부터 질의어를 추출하고 초기 검색을 수행한다. 유사 적합성 피드백을 위하여 초기 검색 결과 중 상위 5개의 문서에 나타나는 키워드를 중요도에 따라 내림차순 정렬하여 상위 10개의 키워드를 초기 질의어에 확장한다. 이렇게 확장된 질의어로 최종 검색을 수행한다. TREC 평가용 테스트 컬렉션 WT10g와 TREC-9의 질의 적합문서 집합을 이용하여 여러 가지 TSV 함수를 사용하여 검색 성능을 평가 하였다. 실험 결과 유사 적합성 피드백을 사용할 경우 TSV 함수에 확률 모델의 CF 요소 뿐만 아니라 TF 요소 등을 적용 시킬 경우 성능 향상에 기여할 수 있음을 알 수 있었다. 또한 색인어와 검색어로 단일어 뿐만 아니라 복합어도 사용할 경우 성능이 향상됨을 알 수 있다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.04b
• /
• pp.556-558
• /
• 2001

전체 영상을 이용하지 않고 영상 안에 포함된 특정 객체 혹은 영역만을 이용하는 "영역에 의한 질의(query-by-region)" 방법은 내용기반 영상 검색 중 상위개념의 방법 이지만, 영상 분할의 한계, 여러개로 분할된 영역을 모두 검색하기 위한 인덱싱 문제, 유사성 측정 시 선형적으로 분리되지 않는 특징 값들에 대한 무리한 선형 조합으로 인한 검색 오류와 같은 많은 문제점을 안고 있다. 따라서 본 논문에서는 영역 기반 영상 검색 시스템인 FRIP에 대하여 영상 분할의 한계를 극복하고, 사용자의 주관성을 영상 검색에 적용하기 위해 확률적 연관성 학습 모델(MPFRL)을 유사성 측정 단계에서 적용 하였고, 아울러 검색 모델로는 기존에 일반적으로 사용되어 오던, 선형 모델을 사용하지 않고 선형 모델보다 유연한 검색 결과를 보여주는 확률적 이접 부울 모델(PDB)을 사용하였다. 또한, 검색 시간을 단축 시키기 위해, 선형 검색 방법에 부울 AND 연산자를 적용 시킴으로써, 검색 시간을 상당부분 단축 할 수 있었다. 실험 결과, 본 논문에서 제안하는 방법(MPFRL+PDB)을 사용할 경우 검색 결과가 선형 조합 보다 향상되는 것을 알 수 있었다. 아울러 사용자 피드백을 통해 사용자가 특징 가중치를 일일이 조절하지 않더라도 단순한 몇 번의 클릭만으로 사용자의 주관성을 반영하고 보다 정확한 검색 결과를 보여 줄 수 있는 시스템을 설계 할 수 있었다.