A manual approach involves the hand-writing of rules on the basis of subject matter expert interviews and the inspection of source code, job flows, data structures and observed behavior.
Manually extracting rules is complicated by the difficulty of locating and understanding highly interdependent logic that has been interwoven into millions of lines of software code.
An automated approach utilizes repository-based software to locate logical connections inherent within applications and extract them into a predetermined business rules format.
With automation, an effective approach is to apply semantic structures to existing applications.
System integrators, software vendors, rules mining practitioners, and in-house development teams have developed technologies, proprietary methodologies and industry-specific templates for application modernization and business rule mining.