Group collaborative systems are recently emerging to support a group of users engaged in common tasks such as group decision making, engineering design, group scheduling, or collaborative writing. This paper provides an change management framework for the group collaborative system to facilitate managing dependency relationship between shared objects and dependent user views, and coordinating change and propagation activities between the two in distributed computing environments. Specifically, the framework adopts an object-oriented database paradigm and presents several object constructs capturing dependency management and change notification mechanisms. First, it introduces change management mechanisms with transient shared objects and secondly, it extends them into presistent object computing environment by integrating transaction management mechanisms and change notification mechanisms. A prototype change management framework is developed on a commercial object-oriented database management system.

Recently, neural network models have been employed as an alternative to regression analysis for point estimation or function fitting in various field. Thus far, however, no theoretical or empirical guides seem to exist for selecting the tool which the most suitable one for a specific function-fitting problem. In this paper, we evaluate performance of three major function-fitting techniques, regression analysis and two neural network models, back-propagation and linear-Hebbian-learning neural networks. The functions to be fitted are simple linear ones of a single independent variable. The factors considered are size of noise both in dependent and independent variables, portion of outliers, and size of the data. Based on comutational results performed in this study, some guidelines are suggested to choose the best technique that can be used for a specific problem concerned.

The maximum Origin-Destination Flow Path Problem (MODFP) in an Acyclic Network has known as NP-hard. K. S. Sung has suggested on Optimal Algorithm for MODFP based on the Pseudo flo or arc and the K-th shortest path algorithm. When we try to solve MODFP problem by general Branch and Bound Method (BBM), the upper and lower bounds of subproblems are so weak that the BBM become very inefficient. Here we utilized the Pseudo flow of arc' for the tight bounds of subproblems so that it can produce an efficient BBM for MODFP problem.

This paper presents a method for estimating the population variability distribution of the failure parameter (failure rate or failure probability) for each failure mode considered in PSA (Probabilistic Safety Assessment). We focus on the utilization of generic estimates from various industry compendia for the estimation. The estimates are complicated statistics of failure data from plants. When the failure data referred in two or more sources are overlapped, dependency occurs among the estimates provided by the sources. This type of problem is first addressed in this paper. We propose methods based on ML-II estimation in Bayesian framework and discuss the characteristics of the proposed estimators. The proposed methods are easy to apply in real field. Numerical examples are also provided.

In a microcellular personal communication service (PCS) it is required to minimize the paging and location updating signals. We propose a multi-layered location registration scheme to reduce the paging and updating signals. In this scheme the subscribers are grouped by their characteristics (velocity and call arrival rate) and are served by appropriately sized location registration area. In order to group the subscriber, we define subscriber grouping problem (SGP). Proposition are examined to solve the grouping problem. The performance of the proposed subscriber grouping algorithm is tested with examples. Simulation results indicate that the subscriber grouping procedure is effective for designing the multi-layered location registration scheme.

Intelligent Decision System support the decision analysis process in the managerial problems with decision analytic knowledge as well as domain specific knowledge. Influence Diagram has been one of the major knowledge representation in the intelligent decision system. In the development of intelligent decision system, knowledge acquisition is also known to be difficult. This paper suggests a developing tool using an influence diagram and Verbal Protocol Analysis which facilitates knowledge acquision for intelligent decision system. An ennvironmental decision making problem is used as an illustrative example and validation of the suggested developing tool is discussed. The suggested tool is very flexible to be expanded or applied to similar problems.

Based on linear models, the inference about the true measurement x$_{f}$ and the optimal designs x (nx1) for the calibration experiments are considered via Baysian statistical decision analysis. The posterior distribution of x$_{f}$ given the observation y$_{f}$ (qxl) and the calibration experiment is obtained with normal priors for x$_{f}$ and for themodel parameters (.alpha., .betha.). This posterior distribution is not in the form of any known distributions, which leads to the use of a numerical integration or an approximation for the calculation of the overall expected loss. The general structure of the expected loss function is characterized in the form of a conjecture. A near-optimal design is obtained through the approximation nof the conditional covariance matrix of the joint distribution of (x$_{f}$ , y$_{f}$ $^{T}$ )$^{T}$ . Numerical results for the univariate case are given to demonstrate the conjecture and to evaluate the approximation.n.

Several quality control schemes have been extensively compared using multivariate normal data sets simulated with various correlation structures. They include multiple univariate CUSUM charts, multivariate EWMA charts, multivariate CUSUM charts and Shewhart T$^{3}$ chart. This paper considers a new approach of the multivariate EWMA chart, in which the smoothing matrix has full elements instead of only diagonal elements. Performance of the schemes is measured by avaerage run length (ARL), coefficient of variation of run length (CVRL) and rank in order of signaling of off-target shifts in the process mean vector. The schemes are also compared by noncentrality parameter. The multiple univariate CUSUM charts are generally affected by the correlation structure. The multivariate EWMA charts provide better ARL performance. Especially, the new EWMA chart shows remarkable results in small shifts.