• Proceedings of the Korea Information Processing Society Conference
• /
• 2003.05a
• /
• pp.427-430
• /
• 2003

멀티미디어 특히 비디오 데이터는 대용량 파일 단위의 저장 특성을 가지며 사용자 접근 패턴이 기존의 햄 객체인 이미지 또는 텍스트와 다르기 때문에 기존의 웹 캐슁 정책은 멀티미디어 프락시 서버에는 적절하지 않다. 본 연구에서는 멀티미디어를 위한 프락시 서버의 캐쉬 히트율을 높이고 파일 단위 캐쉬 관리 정책을 세그먼트 관리에 적용하여 세그먼트 관리의 과부하를 줄이는 고정 크기 세그먼트 기법과 개선된 캐쉬 제거 기법을 제안하였다. 또한 시뮬레이션을 통해 제안 기법의 성능 개선 정도를 분석 제시하였다.

• Proceedings of the Korea Contents Association Conference
• /
• 2018.05a
• /
• pp.361-362
• /
• 2018

본 연구는 대학 웹페이지의 Q&A(질의응답) 게시판에 게재된 비정형화 된 데이터를 수집한 후 텍스트마이닝과 네트워크 분석을 활용하여 자주 등장하는 키워드 간 연관 패턴을 파악하고자 한다. 분석결과를 바탕으로 FAQ(자주하는 질문) 게시판을 구성한다면 반복적인 질문에 대한 민원을 간소화함으로써 수요자의 편의성과 행정의 효율성 향상에 기여하고 나아가 원활한 양방향 소통이 가능할 것으로 기대한다.

• Review of KIISC
• /
• v.18 no.3
• /
• pp.34-40
• /
• 2008

웹상에서 표시되는 다양한 정보의 유출을 막기 위해서 접근 방법으로 캡처 프로그램이 필요로 하는 시스템 기능의 패턴을 분석한 후 분석된 시스템 기능을 제어하는 방식을 제안한다. 이것은 기존의 다양한 캡처 프로그램을 등록하는 방식과 비교하여 보안 위험 기간을 획기적으로 줄여줄 뿐만 아니라 미래의 잠재 위험에 대해서도 근본적인 대책을 가져다 준다는 점에서 강점이 있다. 향후 시스템 기능 제어 기술을 확장하여 안전한 가상 환경을 제공하는 가상화 보안 기술로 확대 연구를 진행할 계획이다.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2002.05c
• /
• pp.112-117
• /
• 2002

웹의 급속한 발전과 함께 오디오 컨텐츠의 양이 증가하게 되었고 수많은 오디오 데이터베이스로부터 특정한 오디오 컨텐츠를 찾아내는 오디오 검색 시스템의 필요성이 증가했다. 본 논문에서는 효율적인 오디오 검색 시스템을 위해 사용자들이 질의하는 멜로디 패턴을 이용해서 FAI(Frequently Accessed Index)라는 새로운 인덱싱 메카니즘을 제안하고 이 FAI를 기반으로 구현한 프로토타입 시스템의 구조와 원리에 대해서 설명한다. 실험을 통해서 FAI 기반 오디오 검색 시스템의 성능을 일반적인 오디오 검색 방법과 비교해 보았다.

• The KIPS Transactions:PartD
• /
• v.17D no.4
• /
• pp.253-258
• /
• 2010

Traditional frequent pattern mining considers equal profit/weight value of every item. Weighted Frequent Pattern (WFP) mining becomes an important research issue in data mining and knowledge discovery by considering different weights for different items. Existing algorithms in this area are based on fixed weight. But in our real world scenarios the price/weight/importance of a pattern may vary frequently due to some unavoidable situations. Tracking these dynamic changes is very necessary in different application area such as retail market basket data analysis and web click stream management. In this paper, we propose a novel concept of dynamic weight and an algorithm DWFPM (dynamic weighted frequent pattern mining). Our algorithm can handle the situation where price/weight of a pattern may vary dynamically. It scans the database exactly once and also eligible for real time data processing. To our knowledge, this is the first research work to mine weighted frequent patterns using dynamic weights. Extensive performance analyses show that our algorithm is very efficient and scalable for WFP mining using dynamic weights.

• Convergence Security Journal
• /
• v.11 no.6
• /
• pp.9-15
• /
• 2011

It is difficult to identify attack requests from normal ones when those attacks are based on CSRF which enables an attacker transmit fabricated requests of a trusted user to the website. For the protection against the CSRF, there have been a lot of research efforts including secret token, custom header, proxy, policy model, CAPTCHA, and user reauthentication. There remains, however, incapacitating means and CAPTCHA and user reauthentication incur user inconvenience. In this paper, we propose a method to detect CSRF attacks by analyzing the structure of websites and the usage patterns. Potential victim candidates are selected and website usage patterns according to the structure and usage logs are analyzed. CSRF attacks can be detected by identifying normal usage patterns. Also, the proposed method does not damage users' convenience not like CAPTCHA by requiring user intervention only in case of detecting abnormal requests.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.6
• /
• pp.645-651
• /
• 2019

At a time when large amounts of data are being poured out, there are many changes in software architecture or data storage patterns because of the nature of the data being written, rather more read-intensive than writing. Accordingly, in this paper, the query model of Command Query Responsibility Segmentation (CQRS) pattern separating the responsibilities of commands and queries is used to implement an efficient high-capacity data lookup system in users' requirements. This paper uses the 2018 temperature, humidity and precipitation data of the Korea Meteorological Administration Open API to store about 2.3 billion data suitable for RDBMS (PostgreSQL) and NoSQL (MongoDB). It also compares and analyzes the performance of systems with CQRS pattern applied from the perspective of the web server (Web Server) implemented and systems without CQRS pattern, the storage structure performance of each database, and the performance corresponding to the data processing characteristics.

• Journal of KIISE:Databases
• /
• v.35 no.3
• /
• pp.231-242
• /
• 2008

W3C has selected OWL as a standard for Web ontology description and a necessity of research on storage models that can store OWL ontologies effectively has been issued. Until now, relational model-based storage systems such as Jena, Sesame, and DLDB, have been developed, but there still remain several issues. Especially, they lead inefficient query processing performance. The structural problems of their low query processing performance are as follow: Jena has a simple structure which is not normalized and also stores most information in a single table. It exponentially decreases the performance because of comparison with unnecessary information for processing queries requiring join operations as well as simple search. The structures of storages(e.g., Sesame) have been completely normalized. Therefore it executes many join operations for query processing. The storages require many join operations to find simply a specific class. This paper proposes a storage model to resolve the problems that the query processing performance is decreased because of non-normalization or complete normalization of the existing storages. To achieve this goal, we analyze the problems of existing storage models as well as relations of OWL elements and query patterns. The proposed model, defined with the analysis results, provides an optimal normalized structure to minimize join operations or unnecessary information comparison. For the experiment of query processing performance, a LUBM data sets are used and query patterns are defined considering search targets and their hierarchical relations. In addition, this paper conducts experiments on correctness and completeness of query results to verify data loss of the proposed model, and the results are described. With the comparative evaluation results, our proposal showed a better performance than the existing storage models.

• Journal of Internet Computing and Services
• /
• v.13 no.1
• /
• pp.65-74
• /
• 2012

Cloud services are used for improving business. Moreover, customer relationship management(CRM) approaches use social networking as tools to enhance services to customers. However, most cloud systems do not support the semantic structures, and because of this, vital information from social network sites is still hard to process and use for business strategy. This paper proposes a collaboration framework based on social semantic web for cloud system. The proposed framework consists of components to support social semantic web to provide an efficient collaboration system for cloud consumers and service providers. The knowledge acquisition module extracts rules from data gathered by social agents and these rules are used for collaboration and business strategy. This paper showed the implementations of processing of social network site data in the proposed semantic model and pattern extraction which was used for the virtual grouping of cloud service providers for efficient collaboration.

• Journal of Korea Society of Industrial Information Systems
• /
• v.10 no.3
• /
• pp.15-20
• /
• 2005

Since the invention of Web, it became part of our daily life replacing the routine information search and effecting many activities which otherwise could have been done off-line without it. It was a natural evolution of the web technology, which had started out as a simple test based pattern matching, to be based on the optimized match process for its multi-media Web environment, like still images, music and movies that we are to face today. In this paper, we proposed and implemented the model which the multimedia resources can be efficiently shared in semantic web. After converting multimedia resource information(metadata) into RDF, for efficient management, the model separate and allotment actual multimedia resource and corresponding metadata changed to RDF to each server. The proposed model could be applied in all multimedia resources. For easy explanation and implementation, however, we applied it to the digital photograph in example.