시계열 스트림 시퀀스 매칭을 위한 효율적인 내장형 알고리즘

문양세,김진호 강원대학교 정보통신연구소 2006 정보통신논문지 Vol.10 No.-

For handheld devices, minimization of CPU operations is the most important factor in performance. In this paper, we propose the efficient algorithms that extract sequences similar to the given query sequence from the time-series stream sequence matching as the problem of finding similar to the given query sequence from the time-series stream such as network traffic data, stock prices, and sensor data. First, we formally define th stream sequence matching as the problem of finding similar sequences from the time-series stream. Second, we propose an efficient window-based approach by using the window construction mechanism of traditional subsequence matching methods. Third, we provide the notion of window MBR and propose two different stream sequence matching algorithms based on the notion. Fourth, we formally prove correctness of the proposed algorithms by presenting related theorems. Last, through extensive experiments, we show that our approach improves performance significantly compared with the naive approach. Overall, we believe that our methods would be very suitable for handheld devices as the embedded algorithms.

온라인 교육 사이트의 교육 환경 및 효과 분석

이용주,문양세,김혜숙,김진호 강원대학교 정보통신연구소 2008 정보통신논문지 Vol.12 No.-

In recent years, people are gradually interested in the Internet-based educational programs (also known as e-learning and cyber lectures), and online education is now widely used in various educational areas. In this paper, considering the environment and achievement aspects we analyze online education by dividing it into public education and private education. We analyze the online education as the following steps. First, we review the recent online educational methods by comparing them to the traditional offline educational ones. Second, we compare EBS, a representative public educational site, and Megastudy, a representative private educational site, by investigating their operation, human, economic, hardware, and software resources. Third, we compare EBS and Megastudy in aspects of environment and achievement through a survey on high-school students after providing their sample lectures. Analytical results show that the students of using Megastudy get good grades compared with those of using Megastudy. This is because Megastudy provides the specialized lectures that focus on obtaining high scores in examinations. We believe that our research results will provide important directions in improving the quality of online educational sites.

모바일 사전을 위한 효율적인 선인출 기법

홍순정,문양세,김혜숙,김진호,정영준 강원대학교 정보통신연구소 2007 정보통신논문지 Vol.11 No.-

In the mobile Internet environment, we need to perform frequent accesses between a mobile device and a content server for searching or downloading learning materials. In this paper, we propose an efficient prefetching technique to reduce the network cost and to improve the communication efficiency in the mobile dictionary. Contributions of the paper can be summarized as follows. First, we propose an overall framework for the prefetching-based mobile dictionary. Second, we present a method of determining the amount of prefetching data for each of packet-based and flat-rate billing cases. Third, by focusing on the English-Korean mobile dictionary for middle or high school students, we propose a method of determining the words to be prefetched in advance. Fourth, based on these determination methods, we propose an efficient prefetching algorithm. Fifth, through experiments, we show superiority of our prefetching-based method. Experimental results show that, compared with the traditional on-demand method, our prefetching-based method improves the average performance by 9.8%~19.7% for the packet-based billing and by 19.0%~33.2% for the flat-rate billing. These results indicate that our framework can be widely used not only in the mobile dictionary but also in other mobile Internet applications that require the prefetching technique.

RSS기반 블로그 에이전트를 사용한 교육용 블로그 구현

황의경,문양세,김혜숙,김진호,이상민 강원대학교 정보통신연구소 2007 정보통신논문지 Vol.11 No.-

More than 80% of the Internet users are familiar with blogs and exploit blogs as their primary communication tools in Korea. These blogs are widely used as a representative personal media service in educational applications as well as business marketing or information sharing applications. In this paper, we present an efficient mechanism of using blogs for educational applications. For this, we first investigate the major problems that the current educational blog sites have. We then propose a new notion of the blog agent that solves the problems and makes it easy to exploit blogs for educational applications. Contributions of the paper can be summarized as follows. First, we investigate the problems with respect to the homework management by analyzing the existing educational blog sites. Second, we present how we can solve the problems by using blog agent. Third, we show that, if exploiting the blog agent, we can easily design blog sites for homework, consultation, or lecture material management. Fourth, through a survey on current blog-related techniques, we select the RSS-based technique to implement our intelligent blog agent. Fifth, we implement the blog sites for homework, consultation, and lecture material management by using "Cyworld's Paper," which is one of RSS-based blog techniques.

시계열 매칭 기반의 윤곽선 이미지 매칭 시스템 설계 및 구현

김범수,문양세,김진호 강원대학교 정보통신연구소 2008 정보통신논문지 Vol.12 No.-

In this paper we exploit time-series matching for boundary image matching. The motivation is based on that images can be converted to time-series. Thus, we implement a boundary image matching system using time-series matching techniques. We use an index-based matching method that efficiently performs boundary image matching. We implement our boundary matching system as a client-server model. Experimental results show that our index-based method outperforms the sequential scan by one to three orders of magnitude. Our system is practically implemented for image matching through the appropriate techniques in the time-series domain, and we thus believe that our approach can be widely used in image matching areas.

XML 소스 데이터로부터 스타 스키마를 생성하기 위한 XML2Star 알고리즘

최은하,김진호,문양세 강원대학교 정보통신연구소 2006 정보통신논문지 Vol.10 No.-

Star Schema is widely used for OLAP(On-Line Analytical Processing) systems to represent multidimensional data in relational database management systems which can be used to analyze for an enterprise's decision making. The star schema consists of a fact table and a set of dimension tables surrounding the fact table. Many researchers have investigated the methodologies for designing the Star Schema from the source data in relational databases. Meanwhile, XML is emerging as a new standard representing data and documents in the Internet. We also need to analyze a huge amount of data represented by XML files, thus we present an algorithm, XML2Star, for designing Star Schema from XML source data. The XML2Star extracts the components of star schema, which are measurements (i.e., facts), dimensions, and dimension hierarchies, from DTDs in XML documents. Based on th star schema, this algorithm also builds star schema instances (i.e., multidimensional cube) from XML documents. The star schema instances will be used for users' multidimensional analyses by using OLAP tools. Therefore, the proposed algorithm can help user deriving star schema from XML source data in the Internet easily and analyzing the schema over various dimensions effectively.

효과적인 다차원 분석을 지원하는 스트레드시트 방식의 OLAP 인터페이스 설계 및 구현

윤민희,김진호,문양세 강원대학교 정보통신연구소 2006 정보통신논문지 Vol.10 No.-

OLAP systems support effectively data analysis over a large volume of business data to make decisions for an enterprise. In order to process data analysis efficiently, OLAP systems pre-compute the summary of business data over various dimensions and store them into data warehouses. Several specialized tools have been developed to display and to browse the results of complex analytical queries on data warehouses. In this paper, we developed a spreadsheet-like tool to support OLAP operations and to browse the results of analytical queries. Spreadsheet is one of the most popular and user-friendly data analysis tool. It also has lots of advantages, which is interactive, easy to learn, etc. Thus this paper implemented an OLAP tool using spreadsheet-style interface to analyze large OLAP data on World-Wide Web. Through this tool, users can process OLAP activities easily and effectively by using the same ways as well-known shreadsheet tools like Microsoft Excel.

효과적인 다차원 분석을 지원하는 스프레드시트 방식의 OLAP 인터페이스 설계 및 구현

윤민희,김진호,문양세 강원대학교 정보통신연구소 2006 정보통신논문지 Vol.10 No.-

OLAP systems support effectively data analysis over a large volume of business data to make decisions for an enterprise. In order to process data analysis efficiently, OLAP systems pre-compute the summary of business data over various dimensions and store them into data warehouses. Several specialized tools have been developed to display and to browse the results of complex analytical queries on data warehouses. In this paper, we developed a spreadsheet-like tool to support OLAP operations and to browse the results of analytical queries. Spreadsheet is one of the most popular and user-friendly data analysis tool. It also has lots of advantages, which is interactive, easy to learn, etc. Thus this paper implemented an OLAP tool using spreadsheet-style interface to analyze large OLAP data on World-Wide Web. Through this tool, users can process OLAP activities easily and effectively by using the same ways as well-known shreadsheet tools like Microsoft Excel.

벼 기계이앙 1회용 종이 육묘상자 개발

박성태,황정동,김상열,문헌팔,양세준,김순철,유남진 한국국제농업개발학회 2003 韓國國際農業開發學會誌 Vol.15 No.3

벼 기계이앙 육묘에 주로 사용하고 있는 플라스틱 육묘상자의 문제점을 개선하고 친환경 농업 실현에도 기여할 수 있는 벼 기계이양 종이 육묘상자개발 및 육묘시험 결과를 요약하면 다음과 같다. 1. 폐지 등 혼합물을 고열압착한 종이 육묘상자는 내부 길이가 가로 58㎝, 세로 28㎝, 높이는 2㎝이고 무게는 250g이다. 2. 종이 육묘상자 모의 파종 후 10일~30일 묘소질은 플라스틱 육묘상자 모에 비하여 초장이 1.1~1.8㎝ 짧고 본당 건물중은 0.3~3.7㎎이 떨어졌으나 엽수와 모 충실도는 비슷하였다. 3. 종이 육묘상자 모의 파종 후 30일(중묘) 이앙 시 결주율은 3.3%로 플라스틱 육묘상자 모 결주율 3.2%와 거의 같았고, 출수기, 벼 생육특성 및 수량도 플라스틱 육묘상자와 비슷하였다. 4. 종이 육묘상자의 이점은 플라스틱 육묘상자에 비하여 육묘상자 비용이 약 20% 절감되고 파종 후 운반, 치상 등 작업이 용이하며, 이앙 후 회수, 세척 및 보관관리 노력이 생략된다. 그리고 이앙 후는 논에서 유기물화되므로 친환경농업에도 부합된다. A light and bio-degradable paper tray for seedling raising in machine transplanting rice was developed for the first time and a possibility for rice seedling growth was evaluated. The paper tray was made of mixture of recycling paper and sawdust, and molded under high temperature and pressure. The tray was 58㎝ length x28㎝ width x2㎝ height in size and 250g in weight per tray. The seedling height and dry weight per plant grown at the paper tray were generally 1.1 to 1.8㎝ shorter and 0.3 to 3.7㎎ lighter than those grown at the conventional plastic tray, respectively. However, other characters such as leaf number, seedling health score, missing hill, heading date, panicle and spiketlet number, and milled rice yield were similar to those in the plastic tray. Based on the result of research, the paper tray had following advantages in comparison of the plastic tray. The paper tray could be reduced seedling raising cost by 20% as compared with the plastic tray. Since the paper tray was of lightness even after seedbed soil fill-up, it was more handy and convenient to carry over. Labor input for collecting, washing, and storing the empty tray after seedling transplanting was not needed as the plastic tray. The paper tray was also bio-degraded and transformed into an organic matter, once it was incorporated into soil, which was considered as an environmentally- safe cultivation technology.

트랙타 부착 대형이앙기 개발 및 재배적 실용성

박성태,장익주,정창화,김상열,김재현,문헌팔,양세준,김순철 한국국제농업개발학회 2002 韓國國際農業開發學會誌 Vol.14 No.3

쌀 생산비를 줄이고 대경지 규모화 영농에 적응할 수 있는 농기계 개발을 목적으로 연구개발한 대규모 경지적응 트랙터 부착 대형벼이앙기의 구성과 특징 및 개발시작기의 재배적 실용성 평가결과를 요약정리하면 다음과 같다. 1. 트랙터 부착 12조 대형벼이앙기는 크게 바퀴자국 제거로타리, 문짝형 정지판, 이앙장치부 등으로 구성되어 있다. 2. 트랙터 부착 12조 이앙기는 트랙터 동력을 이용하여 로타리, 정지 및 이앙작업이 동시에 이루어 질 수 있도록 개발하였다. 3. 바퀴자국을 없애는 날은 기존의 로타리 날을 교체하면 되도록 반달형 날로 제작하였다. 4. 이앙기는 이동시는 접을 수 있도록 설계 제작되어 대형화가 가능하므로 기존 6조 승용이앙기에 비하여 이앙작업 능력이 1.5배 이상이었다. 5. 트랙터 부착 대형벼이앙기는 트랙터와 같은 큰 동력원을 이용할 수 있으므로 측조시비기, 긴매트모 탑재 등 다양한 작업기 및 기능을 추가로 부착 가능하도록 설계되었다. 6. 트랙터 부착 대형벼이앙기는 어린모 및 중묘이앙시 결주율은 3.0∼4.3%로 기계이앙기 결주 허용범위 5% 미만이었다. 하지만 무경운 논에서는 이앙작업은 가능하나 부묘율이 다소 높은 점은 보완이 필요하다. To develop a labor saving rice cultivation technology, a new tractor attached rice transplanter has been developed for large scale farming and its field performance was evaluated. The rice transplanter consists of three major parts such as rotary for removing tractor tire furrow, door-shaped soil leveling plate and transplanting equipment. The rice transplanter is operated using the power of the tractor and performs rotary, harrow and seedling transplanting simultaneously. A half moon shaped rotary blade was found to be very efficient to remove tire furrow developed during seedling transplanting and the original rotary blade can be replaced easily with existed rotary blade. The seedling transplanting part of the transplanter was designed to fold into small size and the transplanting efficiency of the transplanter was found to be 1.5 times higher than that of the conventional riding type six row transplanter. Since the transplanter is operated through a strong power of the tractor, the additional attachment such as fertilizer application apparatus and long mat rice seedling carrier is possible. Based on the field performance of the transplanter, although the missing hill percentage showed roughly with 3.0∼4.3% for 10-day and 30-day old seedlings but the percentage is less than 5% which is maximum missing hill permit in the transplanting rice. Seedling transplanting in no-tillage field condition is acceptable but floating seedling percentage is still high and thus further technology improvement for minimizing floating seedling percontage is needed.