An Expert System Algorithm for Computer System Diagnostics

In troubleshooting Computer Systems the two most common causes of delay are Trial and Error and having Incomplete Information. The problems in Computer Systems will be fixed faster if the Possible Cause of the Problem is already known. A solution to this is to use an Expert System. This system can reproduce the ability of an expert to diagnose by giving an accurate recommendation on the possible cause of the problem for effective troubleshooting. To know the Possible Cause of a problem there must be a complete set of information. These data will be the one to be inputted in the Expert System to give an accurate recommendation. A problem is that in reality a complete set of data will not always be obtained. There will be instances when the information gathered will be incomplete. This research solved the two most causes of delay which are Trial and Error and having Incomplete Information. This is done by developing an Expert System Algorithm that creates the rules of an Expert System. The rules created from the algorithm are nominal in terms that only the necessary information needs to be inputted. In instances that the data gathered are incomplete the correct Possible Cause can still be suggested. A theorem is also presented in this research about and the Information Dependency of Data which can be used with Incomplete Information Systems and unknown data. Formal Proof of the theorem is provided and its correctness was verified with actual data