PREDICTING SERVICE REQUEST BREACHES

An approach for prioritizing work requests to resolve incidents in an information technology (IT) infrastructure is presented. Historical data of work requests to resolve incidents in the IT infrastructure is divided into first and second data sets. A first set of data fields of work requests in the first data set is used to generate incident concept(s). The incident concept(s) are combined with a second set of data fields of the work requests in the first data set to form a set of predictive variables. Utilizing the predictive variables, a statistical model is generated for predicting whether or not work requests will be resolved in accordance with a service level target. The statistical model is validated using the second data set. The statistical model is deployed to the IT infrastructure.