A Rule-Based Guidance (RBG) system with graphical representation of uncertainty

摘要

Automated expert systems provide decision support for certain subject domain by prompting users for answers. The answers could either be a `yes', `no', or a selected response from multiple choice items. To date, there is no account of any expert system which allows a user to defer or revise his/her response to a particular question. Our Rules-Based Guidance (RBG) System provides such flexibility to the user. At the same time, its recommendation (or conclusion) is updated accordingly to the revised or new input by user. Legislation compliance is complex and thus rules represented in the knowledge are complex Boolean expressions. Easily comprehensible and intuitive information visualization is necessary to help a lay user understand the reasoning process, derivation of the final recommendation, and also to revisit the state of his/her responses to a fixed set of questions. Thus, RBG System has provided functionality for transforming Boolean expressions (with user associated inputs) into a Directed Acyclic Graph (DAG).