System and method for analysis of information technology production service support metrics

A system for analysis of Information technology (IT) production service support metrics, the system comprising of:

• at least one processor;
• and
• a memory communicatively coupled to the at least one processor, wherein the memory comprises an analysis module to:
• receive a plurality of metrics from an associated IT service management system;
• provide a systematic view to analyze the plurality of metrics by designing a causal loop diagram (CLD) that represents the IT service management system and includes a plurality of CLD variables mapped to metrics of IT service management system, wherein the CLD variables include knowledge management, problems, total incidents, new incidents, recurring incidents, and application maturity, wherein the metric that explain the knowledge management are related to the metrics that explain the problems and the total incidents due to connection into the knowledge management, wherein the metrics related to the problems include a number of problems resolved, average problem resolution time, and a number of problems reopened;
• obtain the CLD comprising of the plurality of CLD variables to define inter-intra relationship between the plurality of metrics associated with processes of the IT service management system;
• wherein each metric is connected to the one or more CLD variables;
• determine a root cause of deviations of the plurality of metrics from a tolerance value to enable proactive steps for controlling and negating the deviations;
• determine one or more deviant metrics by comparing a monitored value of a metric with the tolerance value and tracing an associated CLD variable from the causal loop diagram;
• determine for the associated CLD variable a first set of metrics from the plurality of metrics, that caused deviation in the deviant metrics and determining for the associated CLD variable, a second set of metrics from the plurality of metrics that are to be impacted due to the deviation in the deviant metrics;
• iteratively perform factor analysis on the first set of metrics and the second set of metrics to determine cogent factors for quantifying deviation of the deviant metrics wherein iteratively performing factor analysis comprises eliminating the one or more of the plurality of metrics with a variance less than threshold, thereby eliminating loosely correlated metrics and maintaining highly correlated metrics, wherein each cogent factor creates a classification axis on which measurement variables are plotted, and coordinates of each measurement variable projections on the classification axis provides a factor score and a factor loading, wherein the factor score represent a coordinate value, and the factor loading provide correlation of the variable with the factor and an amount of variance accounted by the variable, wherein a first multidimensional space is created for each of the first set of metrics and the factor score, factor loading are determined by projecting the deviant metric on the first multidimensional space, wherein a second multidimensional space is created for each of the second set of metrics and the factor score, factor loading are determined by projecting a measurement variable list built based on outward traversals of the CLD from CLD variables connected to the deviant metric.

A non-transitory computer readable medium embodying a program executable in a computing device for analysis of Information technology (IT) production service support metrics, the program comprising:

• a program code for receiving a plurality of metrics from an associated IT service management system;
• a program code for providing a systematic view to analyze the plurality of metrics by designing a causal loop diagram (CLD) that represents the IT service management system and includes a plurality of CLD variables mapped to metrics of IT service management system wherein the CLD variables include knowledge management, problems, total incidents, new incidents, recurring incidents, and application maturity, wherein the metric that explain the knowledge management are related to the metrics that explain the problems and the total incidents due to connection into the knowledge management, wherein the metrics related to the problems include a number of problems resolved, average problem resolution time, and a number of problems reopened;
• a program code for obtaining CLD comprising of the plurality of CLD variables to define inter-intra relationship between the plurality of metrics associated with processes of the IT service management system, wherein each metric is connected to the one or more CLD variables;
• a program code for determining a root cause of deviations of the plurality of metrics from a tolerance value to enable proactive steps for controlling and negating the deviations;
• a program code for determining one or more deviant metrics by comparing a monitored value of a metric with the tolerance value and tracing an associated CLD variables from the causal loop diagram;
• a program code for determining, for the associated CLD variables, a first set of metrics from the plurality of metrics, that caused deviation in the deviant metrics and determining for the associated CLD variables, a second set of metrics from the plurality of metrics that are to be impacted due to the deviation in the deviant metrics;
• a program code for iteratively performing factor analysis on the first set of metrics and the second set of metrics to determine cogent factors for quantifying deviation of the deviant metrics, wherein iteratively performing factor analysis comprises eliminating one or more of the plurality of metrics with a variance less than threshold, thereby eliminating loosely correlated metrics and maintaining highly correlated metrics, wherein each cogent factor creates a classification axis on which measurement variables are plotted, and coordinates of each measurement variable projections on the classification axis provides a factor score and a factor loading, wherein the factor score represent a coordinate value, and the factor loading provide correlation of the variable with the factor and an amount of variance accounted by the variable, wherein a first multidimensional space is created for each of the first set of metrics and the factor score, factor loading are determined by projecting the deviant metric on the first multidimensional space, wherein a second multidimensional space is created for each of the second set of metrics and the factor score, factor loading are determined by projecting a measurement variable list built based on outward traversals of the CLD from CLD variables connected to the deviant metric.