Abstract
Abstract In recent years, new product development becomes more complex and needs more collaborative work than before. In order to support efficient product design, Workflow Management System (WfMS) plays an important role to help manager to control and analyze the status of workflow. Also, Project Risk Management (PRM) is used to collect and analyze the risk factors of tasks in the projects. This thesis proposes Bayesian Belief Network (BBN) methodology for building risk factors relationship graph based on quantitative risk analysis. Program Evaluation and Review Technique (PERT) is used to estimate the probability of completion time and the sum of cost. Therefore, factors of the tasks in the workflow are collected, analyzed, and recorded in the database. Bayesian Belief Network use probabilities and network map to display the relationships between each node, and infer the influences of the workflow by calculating expected values of time and cost. PERT uses the result from BBN analysis to adjust processing time of tasks. PERT displays the probability of completed days by normal distribution charts and calculates the cost of manpower and tools. A project manager can use the report of risk analysis to make decision about execution or rejection of workflow designs. If a workflow is executed, executing members of each task needs to input data of factors and update the record to database. Else, a project manager must redesign the workflow and upload a new workflow to make analysis till to conform the requirements of projects. Keywords: Project Risk Management (PRM), Workflow Management System (WfMS), PERT, Bayesian Belief Network (BBN), risk factors