• Journal of Korean Library and Information Science Society
• /
• v.45 no.3
• /
• pp.5-36
• /
• 2014

The Web is rich with various sources of information that go beyond the contents of documents, such as hyperlinks and manually classified directories of Web documents such as Yahoo. This research extends past fusion IR studies, which have repeatedly shown that combining multiple sources of evidence (i.e. fusion) can improve retrieval performance, by investigating the effects of combining three distinct retrieval approaches for Web IR: the text-based approach that leverages document texts, the link-based approach that leverages hyperlinks, and the classification-based approach that leverages Yahoo categories. Retrieval results of text-, link-, and classification-based methods were combined using variations of the linear combination formula to produce fusion results, which were compared to individual retrieval results using traditional retrieval evaluation metrics. Fusion results were also examined to ascertain the significance of overlap (i.e. the number of systems that retrieve a document) in fusion. The analysis of results suggests that the solution spaces of text-, link-, and classification-based retrieval methods are diverse enough for fusion to be beneficial while revealing important characteristics of the fusion environment, such as effects of system parameters and relationship between overlap, document ranking and relevance.

• Annual Conference on Human and Language Technology
• /
• 2003.10d
• /
• pp.108-112
• /
• 2003

웹 문서의 가장 큰 특징 중 하나는 링크 구조이다. 이 링크들을 이용하여 전체 웹 문서를 커다란 하나의 네트워크로 구성할 수 있으며 이러한 네트워크를 분석함으로써 보다 중요한 문서, 보다 유용한 사이트를 찾아낼 수 있다. 전통적인 검색 모델인 벡터 모델의 성능 개선을 위해 이러한 링크 분석 기법을 활용하여 검색 정확도를 향상시키기 위한 방법을 제안한다. 또한 하이퍼 텍스트는 보다 정확한 키워드를 포함할 확률이 높으므로, 이를 가중치 계산에 적용하여 보다 정확한 결과를 산출한다.

• Proceedings of the Korea Society for Industrial Systems Conference
• /
• 2004.06a
• /
• pp.20-25
• /
• 2004

본 논문은 하이퍼링크를 활용한 2 단계 스팸 메일 필터링에 관한 방법을 제시한다. 일반적으로 스팸 메일의 본문에는 텍스트 문장보다는 그림이 더 많이 포함되어 있기 때문에 단어의 블랙리스트와 같은 전형적인 방법으로 스팸 메일을 구분하기에는 많은 어려움이 따른다. 이러한 문제를 해결하기 위하여 본 논문에서는 스팸 메일에 포함되어 있는 하이퍼링크를 추출하여 해당 웹 페이지를 가져온 후, 이를 확장된 형태의 메일 본문이라 간주하여 텍스트 정보를 추출하였다. 또한 스팸 메일을 구분하기 위한 정보를 두 가지로 구분하여 사용하였는데, 메일 송신자의 정보와 확실한 스팸 키워드 리스트를 확실한 정보군으로 구분하여 먼저 적용하고, 이보다 덜 명확한 정보들은 토로 구분하여 속성벡터를 만들어 SVM 알고리즘을 적용하였다. 실험결과 하이퍼링크를 통하여 웹페이지를 가져온 방법이 그냥 원본 메일만 사용한 방법보다 F-measure 값이 평균 2.8%의 성능향상을 보였다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.23 no.2
• /
• pp.126-132
• /
• 2013

This study proposed a method to classify internet websites based on occurrence frequency of medical terms in the webpages and website structure composed with webpages and hyperlinks. The classification was done by using the suitability measure defined by three factors: (1)occurrence frequency of medical terms in the whole terms involved in a webpage, (2)occurrence frequency of medical terms in de-duplicated terms involved in the webpage, and (3)the number of hyperlinks to reach to a specific webpage from homepage. We conducted an experiment to verify the proposed method with the 80 websites registered in directories related to medical field and 127 websites in nonmedical field directories, and the experiment result showed 82.5 % of accuracy of the classification.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.18 no.1
• /
• pp.86-93
• /
• 2008

The construction of ontology defines as complicated semantic relations needs precise and expert skills. For the well defined ontology in real applications, plenty of information of instances for ontology classes is very critical. In this study, crawling algorithm which extracts the fittest topic from the Web overflowing over by a great number of documents has been focused and developed. Proposed crawling algorithm made a progress to gather documents at high speed by extracting topic-specific Link using URL patterns. And topic fitness of Link block text has been represented by fuzzy sets which will improve a precision of the focused crawler.

• Journal of Korean Library and Information Science Society
• /
• v.38 no.4
• /
• pp.257-276
• /
• 2007

Back-of-book-style indexes have a similar function as back-of-book indexes. The best advantage o4 back-of-book-style indexes for Information access on the web is to give direct access to specific subjects of interest. Though back-of-book-style indexes are alphabetically arranged as back-of-book indexes, they have linked index entries to contents on the site by using a anchor tag of HTML. In this research, I have created back-of-book-style indexes in two separated ways, by hand-crafted and semi-automatic Indexing. We have utilized back-of-book-style indexes, that is similar to back-of-book index of traditional information organization method of library and information science, in library circumstances.

• Proceedings of the Korea Society for Industrial Systems Conference
• /
• 2002.06a
• /
• pp.306-315
• /
• 2002

With the rapid growth of WWW, the existing systems are integrated into web and various web-based systems are developed. Unlike the general applications, web aplications are developed by combining the various technologies and have their own complexities. So, we have much difficulties in the development and maintenance of web applications. To accommodate to the rapidly changing business environments and user requirements, the continuos evolution is required. In this paper, we present a method for supporting the effective development and maintenance of web applications. Our method involves the extraction of web application's structure information and analyzes web log ales containing the useful information about web site. We also describe a web testing method using the attracted information and our system developed for extracting hyperlink information and analyzing web log.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.07a
• /
• pp.67-69
• /
• 2005

웹을 기반으로 한 프로그래밍 교육 시스템에 관하여 많은 연구가 이루어져 왔으나 대부분의 연구는 강의 노트의 제공 방법, 또는 하이퍼링크 상에서의 효과적인 네비게이션 방법과 같은 원격 교육 시스템에 국한된 내용이었다. 그러나 프로그래밍 교육은 그 특성상 강의노트의 제공 방법이나 하이퍼링크 상에서의 효과적인 네비게이션 방법보다도 많은 양의 과제를 통한 연습 기회의 제공과 각 과제에서의 교수자와 학습자간의 친밀하고 개인적인 상호작용이 학습자의 학업성취도에 더욱 큰 영향을 미친다. 본 논문에서는 원격 교육 시스템의 관점에서 더 나아가 교수자가 학습자에게 많은 양의 과제를 통한 연습의 기회를 부여 할 수 있고 각 과제당 교수자와 학습자간의 상호작용을 극대화할 수 있는 웹 기반 프로그래밍 교육 시스템을 구현하였다. 웹 기반 프로그래밍 교육 시스템은 학습자에게 시간과 장소에 구애됨이 없이 웹 브라우저를 이용하여 프로그램을 코딩하거나 컴파일할 수 있고 SPLINT를 통한 분석결과를 통해 즉시 피드백을 받을 수 있는 환경을 제공한다. 또한 교수자에게 많은 수의 학생들의 과제를 관리하고 교수자의 평가 기준에 의해 한꺼번에 자동으로 평가할 수 있는 환경을 제공하여 준다. 이러한 웹 기반 프로그래밍 교육 시스템은 웹을 이용한 프로그래밍 교육을 면대면 학습에 더욱 가깝게 만들어준다. 현재 웹 기반 프로그래밍 교육시스템은 K대학교 컴퓨터교육과 1학년 재학생들을 대상으로 한 C 프로그래밍 강좌에 사용되고 있다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.364-366
• /
• 2022

The web uses an asynchronous web method to provide various information having different processing speeds together. The asynchronous method has the advantage of being able to respond to other events even before the task is completed, but a typical crawler has difficulty collecting information provided asynchronously by collecting point-of-visit information on a web page. In addition, asynchronous web pages often do not change their web address even if the page content is changed, making it difficult to crawl. In this paper, we propose a web crawling algorithm considering asynchronous page movement by analyzing links in the web. With the proposed algorithm, it was possible to collect dictionary information on TTA terms that provide information asynchronously.

• Proceedings of the Korea Contents Association Conference
• /
• 2006.11a
• /
• pp.478-482
• /
• 2006

This paper implemented speech-library and the Web Services that conversion the Web page for the speech recognition. The system is consisted of Web services consumer and Web services providers. The Web services consumer has libraries that Speech-library and proxy-library. The Speech -library has functions as follows from the user's speech extracted speech-data and searching the URL in link-table that is mapped with user's speech. The proxy-library calls two web services and is received the returning result. The Web services provider consisted of Parsing Web Services and Speech-Recognition Web Services. Parsing Web Services adds ActiveX control and reconstructs web page using the speech recognition. The speech recognizer is the web service providers that implemented in the previous study. As the result of experiment, we show that reconstructs web page and creates link-Table. Also searching the URL in link-table that is mapped with user's speech. Also confirmed returning the web page to user by searching URL in link-table that is mapped with the result of speech recognition web services.