• Proceedings of the Korean Information Science Society Conference
• /
• 2004.04b
• /
• pp.544-546
• /
• 2004

최근 자연언어를 이용하여 정보를 제공하거나 업무를 수행하는 대화 에이전트가 활발히 연구되고 있다. 보다 다양한 사용자 질의를 분석하고 적절한 답변을 제공하기 위해서는 대규모의 답변 데이터베이스를 구축하여야 한다. 기존 답변 데이터베이스의 구축은 설계자가 수동으로 입력질의의 패턴을 분석하고 이에 대한 답변을 작성하여 패턴-답변쌍을 제작하여 이루어졌다. 따라서 패턴의 분석이 설계자에 의존적이어서 일반적이지 못하며 중복되거나 쓸모없는 패턴-답변쌍이 생성되기도 한다. 또한 초기에 구축된 답변 데이터베이스에 의해 성능이 제한되어 답변 성능의 향상을 위해서는 답변 데이터베이스를 수동으로 추가해야한다. 본 논문에서는 대화를 통해 필요한 정보론 수집하여 자동으로 괘턴-답변쌍을 생성하는 방법을 제안하다. 사용자 입력문장을 화행별로 구분하고 각 화행별 답변 템플릿을 이용하여 패턴-답변쌍을 완성한다. 기존의 수동제작 방식과 비교 실험을 통해 제안하는 방법이 지식구조 구축 속도나 사용자 평가 면에서 훨씬 우수함을 확인하였다.

• Asia pacific journal of information systems
• /
• v.15 no.2
• /
• pp.195-218
• /
• 2005

The emergence of ubiquitous computing environment will change the service architecture of business information systems such as Decision Support System(DSS), which will be a new application. Recent mobile DSSs allow the decision makers to be benefited from web and mobile technology. However, they seldom refer to context data, which are useful for proactive decision support. Meanwhile, ubiquitous applications so far provide restricted personalization service using context and preference of the user, that is, they do not fully make use of decision making capabilities. Hence, this paper aims to describe how the decision making capability and context-aware computing are jointly used to establish ubiquitous applications. To do so, an amended DSS paradigm: CKDDM(Context-Knowledge-Dialogue-Data-Model) is proposed in this paper. What will be considered for the future decision support systems when we regard ubiquitous computing technology as an inevitable impact that enforces the change of the way of making decisions are described. Under the CKDDM paradigm, a framework of ubiquitous decision support systems(ubiDSS) is addressed with the description of the subsystems within. To show the feasibility of ubiDSS, a prototype system, CAMA-myOpt(Context-Aware Multi Agent System-My Optimization) has been implemented as an illustrative example system.

• Journal of Preventive Medicine and Public Health
• /
• v.29 no.1 s.52
• /
• pp.1-14
• /
• 1996

To evaluate the health risk of the workplace environment of a general hospital toward the hospital workers, a questionnaire survey on the perception of risk at the workplace environment and environmental measurements at 27 locations with 240 workers in the hospital were made from October 25th, 1993 to October 30th, 1994. The results were as follows ; 1. By the environmental measurements, 86 workers(36%) were found to be exposed to poor or harmful degree of general illumination, 193 workers(80%) were exposed to poor or harmful local illumination, 34 workers(14%) were exposed to poor or harmful degree of thermal condition and 180 workers(75%) were exposed to poor or harmful noise level, but nobody was exposed to poor or harmful dust and toluene concentration. Also nobody was exposed to poor or harmful level of radiation or anticancer/antibiotic agents. 2. The subjective perceptions on the environmental conditions felt by the workers were different from the objective findings by the environmental measurements. The workers underscored the poor illumination state and overscored the dust concentrations. Also workers oversensitized about the thermal condition, the noise level, the radiation level, the toluene concentration and anticancer/antibiotic agents 3. The sources of noise were the dialogue(179 workers) and the office instruments(131 workers). The sources of dust came from the clothes(108 workers) and the building materials(79 workers). 4. The questionnaire survey showed that the 28 workers were exposed to poor or harmful level of the antibiotics, 10 workers to alkali or acid and 6 workers to drug dust in the pharmacy but the above findings could not be proved by the environmental measurements.