Hi,

in order to measure the accuracy of a XPS you need some cases which function as "gold standard", i.e. cases for which the solution is known, but which where not yet used for the development of the system (that is also some "good evalution practice" in machine learning), otherwise any evaluation of the accuracy would be biased.

Measuring the accuracy then means to run the XPS on those test cases, record the result and compare it with the expected result for building up a confusion matrix. For details of evaluation of this confusion matrix and start learning about evaluation measures, you can use Wikipedia and search for "accuracy", "recall", "precision" or "F-measures". The later three are usually prefered over accuracy, since they carry more information about "false positives" and "false negatives".

Hope this helps for a start.

you can use the Confusion Matrix .

you can see this link: https://en.wikipedia.org/wiki/Precision_and_recall

Thank you so much Mr. Thomas Hoppe , Hossein and Ms. Noura Aboudi for your help and information.

Sir Hoppe and Hossein your guide and references are helpful.

Thanks again!

Regards,

Noor