| 摘要 |
A method that includes: obtaining a computer code usable to process insurance claims; building a computer readable directed graph representing a control flow of the code, by identifying decisions and actions in the code, the graph comprising nodes connected by edges, some of the nodes being decision nodes associated with the decisions and some of the nodes being action nodes associated with the actions; determining, on the graph, benefit action nodes that are each associated with at least one monetary outcome of a specified insurance claim; identifying all logic paths that lead to each benefit action node by traversing the graph from each benefit action node backwards, each logic path comprising a sequence of preceding decision nodes and action nodes connected by edges, each set of paths being associated with a specified benefit action node representing a benefit rule; and outputting all benefit rules by presenting each specified benefit action vis à vis grouped logic paths associated with the specified benefit action.
|
