PARETO PRINCIPLE TO IMPROVE ANOMALY DETECTION ON SOFTWARE ASSET MANAGEMENT

Software Asset Management (SAM) is essential for a large company with centralized software distribution system. Unfortunately, the operationalization of SAM has various problems. These problems become even more complicated when it comes to managing big data. This research proposes Pareto Principle reduce data dimensions solve problem sizes without losing dataset characteristics before conducting anomaly detection. detection mandatory identify and invalid due misalignment or misclassification. Therefore, this study compares state-of-arts algorithms: I-forest, KNN, SVM. As result, we found that SVM best algorithm, an accuracy rate 78.4%. In addition, using total population name variation effectively reduces number observations 20% instances only 16.5% features compromising characteristics. algorithm based on experimental results, use increases by 3.2% processing time efficiency 20%.