大學綜合試驗 出題에 關한 硏究

卞烘圭,吳鎭坤,申鎔鎭,金泳喆,鄭求福,柳哲鍾,洪允杓 全北大學校 1977 論文集 Vol.19 No.-

The Comprehensive Examination for the undergraduate students is planned to adopt the Jeonbug National University for the purpose of enhancing the quality of its educational outputs. In order to effectively measure and evaluate the level of student progress toward the minimun standard as the students who have finished their general education and basic education for their major field study, the university has decided to make an item pool related to the subjects for the examination. This study is an attempt to make good items of five required courses among general education subjects which are Korean Language, English, Korean History, National Ethics. and Natural Science and are expected to be the subjects of the examination. Instead of making those items separately from the regular examination during the academic semesters, the university plans to collect and use good items from the usual mid-term and final examinations of those subjects. A key concern, in this study, therefore, is to analyze all the items used for the mid-term and final examinations of those subjects during the regular academic semesters. From the data of an about 100 sample in each subject at each time of total three tests, the difficulty index, discrimination index, the response rate of students on each item, and other general descriptive statistics are calculated and then good items are pooled for the future use for the Comprehensive Examination. Anhother important concern in this study is this study is given to the collection and analysis of controversial issues and problems on the process of operating the examination at other universities where the Comprehensive Examination is already adopted. The data from four institutions and from opinions of six participants of this study are used. The following main findidings are summarized : 1. The items have generally shown to be easy while the degree of discrimination is low. 2. Among the five subjects, Korean Language shows the highest rate of right answer (difficulty index) while English the strongest one in the discrimination index. 3. When comparing two types of test between the multiple choice and the short-answer, the latter seems to be powerful in both difficulty and discriminative. 4. The only one question of essay type test reveals clear limitations in scoring and evaluating students progress. 5. All items are constructed under the consideration of content-referenced base, not of the basis on behavioral aspect. 6. The analysis of issues and problems related to the Comprehensive Examination suggests that it is acceptable to test in terms of required subjects of general education and a few most basic subjects of major areas of study. 7. The examination can be functionally operated when it is held at the end of the first year of and before the beginning of the second year of the college years. The length of the examination is usally held from three to five hours. 8. Since the examination should be existed on the idea of criterion-referenced base rather than of norm-referenced one because of its testing the qualification of the entering behavior of students for the further study in their major field, the standard of the pass-and-failure should be set on at least 60 percent of correct answers in each subject. But there must be provided at least two additional opportunities for the make-up examination for those who failed in the first one. 9. A special committee must be established in order to deal with all the details related to the Comprehensive Examination.

사용자 프로파일과 상황정보를 이용한 여행지 및 방문경로 추천시스템

김영민,변영철,이상준 제주대학교 공과대학 첨단기술연구소 2004 尖端技術硏究所論文集 Vol.15 No.2

In this study we design and implement tour information recommendation and path scheduling system which recommends the places to visit and guides a path for a traveler who wants an information in a strange place. System gets user profile information to recommend some preferable places for the user and also it collects context informations, for example weather, season. And then the system classifies the user and calculates which places are preferable for the user. After the user selects places to visit from the list of recommendation, finally the system shows a reasonable path to visit sequentially. Through this experimentation, we found that this recommending system can work as an human expert to guide tour informations.

역전파 신경망을 이용한 Kornet 사용자 질의 메일 자동 분류에 관한 연구

김영민,변영철,이상준,홍영보 제주대학교 공과대학 첨단기술연구소 2003 尖端技術硏究所論文集 Vol.14 No.1

For increasing Internet Users, the increasingly asked E-Mail is needed to be classified automatically. This thesis is proposed to automatically classify the asked E-Mail of KORNET queried by users using backpropagation neural network. This experiment is use of the 490 asked E-mail. When selected keyword number is 70, the recognition rate is highest, 73.1%. Even though the Classification rate of the asked E-Mail is less than the iris recognition and fingerprint recognition, it is affirmativefor an asking User to satisfy with a quick reply.

규칙기반 온톨로지 인스턴스 자동생성

윤현주,변상용,김장형,변영철 제주대학교 공과대학 첨단기술연구소 2004 尖端技術硏究所論文集 Vol.15 No.2

In this paper, we propose a method generating of instance ontology from web pages for the lodge. First, we analyze non-structured web documents and generate information extraction rules in order to extract specific informations from web sources. And we extract the informations from various web pages related to lodge using generated rules. And then ontology instances are automatically generated through adding OWL semantic information to the extracted informations. Our method is able to cut down on time and cost than a method which generates various ontology instances manually. Also agents can search the information corresponding to semantically what the user needs in the various web pages using numerous vocabulary having same meaning and provide good quality of informations to user.

규칙기반 CC/PP 프로파일 생성 방법

송재경,변영철,이상준 제주대학교 공과대학 첨단기술연구소 2005 尖端技術硏究所論文集 Vol.16 No.2

In the ubiquitous computing environment, service providers should able to deliver their services to various devices, such as PDA, Smart Phone, Telematics devices. For this purpose, W3C announced 'Composite Capabilities / Preference Profile (CC/PP) Standard' for delivery-context description. At present, only a small number of mobile phone supports the CC/PP standard, But most of contents servers do not. In this paper, by analyzing request headers which are received from various ubiquitous devices, a rule-based CC/PP profiling method in ubiquitous environment is developed. Service providers can use these profiles for providing optimized services to the various devices which have different capabilities.

다중 장치 지원 모바일 응용서버 설계

정은경,변영철,이상준 제주대학교 공과대학 첨단기술연구소 2003 尖端技術硏究所論文集 Vol.14 No.1

In this paper, we propose the design of the application server which provides internet services for the mobile devices such as a mobile phone and PDA, etc. The neutral form of mobile application which can be requested by various kinds of mobile devices is created and deployed to the proposed application server. The mobile application server supplies the content transformed according to the constraints and features of a mobile device. By using the server, a user can use many kinds of mobile devices at the same time, and only one session can be created and maintained at the server. The mobile application server is composed of the tree modules: Mobile Request Generator, Mobile Engine, and Environmental Adoptation module. The Mobile Request Generator module receives the mobile request from a mobile device and creates the mobile request information supplied to the Mobile Engine module. The Mobile Engine module processes the mobile application requested and creates a content which will be sent to a mobile device. The Environmental Adaptation module transforms the output of Mobile Engine into the contents that a mobile device can display. In this way, the proposed mobile application server can support various kinds of mobile devices using a neutral form of internet service.

시멘틱 웹 환경에서의 온톨로지 기반의 정보검색

정은경,김영민,변영철,이상준,박경린 濟州大學校 基礎科學硏究所 2003 基礎科學硏究 Vol.16 No.2

온톨로지는 시멘틱 웹을 구현하기 위한 이론이다. 온톨로지는 웹자원의 효과적인 검색, 통합, 재사용을 위한 새로운 기술이고 사람과 이질의 응용 시스템 간에 지식을 공유하고 재사용하기 위한 방법을 제공한다. 온톨로지의 역할은 특별한 개념을 명세화하는 것이다. 도메인과 일반적인 온톨로지는 정보와 지식 시스템을 분석하고 디자인하는 측면에서 공유, 통합, 재사용될 수 있다. 본 연구는 제주도의 숙박과 관광정보를 위한 온톨로지를 디자인하는데 목적이 있다. 그리고 제안된 온톨로지 기반으로 인터넷 웹 검색 시스템을 만드는데 목적이 있다. 본 연구에 의해 클라이언트가 사용하는 표현은 다르지만 뜻이 같은 동의어에 의해 같은 결과를 얻을 수 있다는 것을 알게 되었다. Ontologies are formal theories that are suitable for implementing the semantic web. Which is a new technology that attempts to achieve effective retrieval, integration, and reuse of web resources. Ontologies provide a way of sharing and reusing the knowledge among the people and the heterogeneous application systems. The role of ontologies is that of making explicit specified conceptualizations. In this context, domain and generic ontologies can be shared, reused, and integrated in the analysis and design stage of information and knowledge systems. This study aims to design an ontology for juju lodges and tour, and build an Internet web retrieval system based on the proposed ontologies. This study found that we can take the same result for synonym that cilent use.

온톨로지 매핑 정보를 이용한 웹 서비스 자동발견

송재경,김도현,변영철,이상준 제주대학교 공과대학 첨단기술연구소 2004 尖端技術硏究所論文集 Vol.15 No.2

Web Service is a new model based on web to offer dynamic information for a user's request. Currently, the web service discovery mechanism is not efficient because it does not support automatic discovery based on the semantics or capabilities of a web service. The language for semantic web service, such as OWL-S, is being developed to express the capability of a web service semantically. Such semantics can support the more flexible discovery of web services. Web service discovery occurs when the provider's description matches to that of requester. But the matching can fail when the ontologies of service description differ from that of the requester although the requester wants that Web Service. In distributed environment service providers develope their semantic web services independently and tend to use their ontologies to describe web services. So a new approach must be required to avoid service discovery failure. In this paper, a method of annotating the query of requester and discovering semantic web services without service discovery failure is proposed.

Load Balancing for ALE-compliant Middleware

Yung-Cheol Byun,Young-Sik Noh 한국콘텐츠학회 2011 ICCC International Digital Design Invitation Exhib Vol.2011 No.12

Load Balancing for ALE-compliant Middleware

Yung-Cheol Byun,Young-Sik Noh 한국콘텐츠학회 2011 한국콘텐츠학회 ICCC 논문집 Vol.- No.-