Improvement of incident management model using machine learning methods
Abstract
Technical support of IT infrastructure is a crucial aspect of organizational operations, with the most challenging task being ensuring service continuity. Quality support guarantees high IT efficiency, but complex incidents reduce support quality and require effective management. Incident management includes configuration processes and control of technical solutions. To improve technical support, adhering to both quantitative and qualitative standards and considering system specifics is necessary. According to service level agreements (SLA), the resolution time of incidents is important. „Service Desk“ tools, applying machine learning methods, can help optimize these processes. Incorrectly classified user requests lead to additional work for the IT team and delay incident resolution. Machine learning methods, such as K-means clustering, Random Forest regression, and classification, can optimize incident management and speed up resolution time. The research analyzes „Service Desk“ incident data to model resolution times and improve incident management.
Article in Lithuanian.
Incidentų valdymo modelio tobulinimas, taikant mašininio mokymosi metodus
Santrauka
IT infrastruktūros techninis palaikymas yra esminis organizacijos veiklos aspektas, kurio sudėtingiausia užduotis yra užtikrinti veikimo tęstinumą. Kokybiškas palaikymas garantuoja aukštą IT efektyvumą, tačiau sudėtingi incidentai sumažina palaikymo kokybę ir reikalauja veiksmingo valdymo. Incidentų valdymas apima konfigūracijų procesus ir techninių sprendimų kontrolę. Siekiant pagerinti techninį palaikymą, būtina laikytis tiek kiekybinių, tiek kokybinių standartų ir atsižvelgti į sistemų specifiką. Pagal paslaugų lygio sutartis (SLA) svarbus incidentų sprendimo laikas. „Service Desk“ įrankiai, taikant mašininio mokymosi metodus, gali padėti optimizuoti šiuos procesus. Naudotojų neteisingai klasifikuotos užklausos lemia papildomą IT komandos darbą ir vilkina incidentų sprendimą. „K-means“ klasterizacijos, „Random Forest“ regresijos ir klasifikacijos mašininio mokymosi metodai gali optimizuoti incidentų valdymą ir pagreitinti sprendimo laiką. Tyrimo tikslas yra analizuoti „Service Desk“ incidentų duomenis, siekiant modeliuoti sprendimų laiką ir pagerinti incidentų valdymą.
Reikšminiai žodžiai: IT infrastruktūra, techninis palaikymas, incidentų valdymas, incidentų sprendimo laikas, Service Desk, mašininio mokymosi metodai, užklausos klasifikavimas.
Keyword : IT infrastructure, technical support, incident management, incident resolution time, Service Desk, machine learning methods, request classification
This work is licensed under a Creative Commons Attribution 4.0 International License.
References
Agarwal, S., Sindhgatta, R., & Sengupta, B. (2012). SmartDispatch: Enabling efficient ticket dispatch in an IT service environment. In The 18th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 1393–1401), Beijing, China. https://doi.org/10.1145/2339530.2339744
Altintas, M., & Tantug, A. C. (2014). Machine learning based ticket classification in issue tracking systems. In Proceedings of the International Conference on Artificial Intelligence and Computer Science (AICS) (pp. 195–207), Bandung, Indonesia.
Arthur, D., & Vassilvitskii, S. (2007). k-means++: The advantages of careful seeding. In Proceedings of the Eighteenth Annual ACM-SIAM Symposium on Discrete Algorithms (pp. 1027–1035). Society for Industrial and Applied Mathematics.
Bartolini, C., Stefanelli, C., & Tortonesi, M. (2009). Business-impact analysis and simulation of critical incidents in IT service management. In 2009 IFIP/IEEE International Symposium on Integrated Network Management (pp. 9–16), New York, NY, USA. https://doi.org/10.1109/INM.2009.5188781
Breiman, L. (1996). Bagging predictors. Machine Learning, 24(2), 123–140. https://doi.org/10.1007/BF00058655
Costa, J., Pereira, R., & Ribeiro, R. (2019). ITSM automation-using machine learning to predict incident resolution category. In K. S. Soliman (Ed.), Proceedings of the 33rd International Business Information Management Association Conference, IBIMA 2019: Education Excellence and Innovation Management through Vision 2020 (pp. 5819–5830). International Business Information Management Association, IBIMA.
Dasgupta, G. B., Nayak, T. K., Akula, A. R., Agarwal, S., & Nadgowda, S. J. (2014). Towards auto-remediation in services delivery: Context-based classification of noisy and unstructured tickets. In Proceedings of the International Conference on Service-Oriented Computing (SOC) (pp. 478–485). Springer. https://doi.org/10.1007/978-3-662-45391-9_39
Eckerson, W. (2010). Performance dashboards: Measuring, monitoring, and managing your business. Wiley.
Jain, A. K., Murty, M. N., & Flynn, P. J. (1999). Data clustering: A review. ACM Computing Surveys, 31(3), 264–323. https://doi.org/10.1145/331499.331504
Ng, A. (2018). Machine learning yearning. https://nessie.ilab.sztaki.hu/~kornai/2020/AdvancedMachineLearning/Ng_MachineLearningYearning.pdf
Paramesh, S., & Shreedhara, K. (2019). IT help desk incident classification using classifier ensembles. ICTACT Journal on Soft Computing, 9(4), 1980–1987.
Revina, A., Buza, K., & Meister, V. G. (2021). Designing explainable text classification pipelines: Insights from IT ticket complexity prediction case study. In Interpretable artificial intelligence: A perspective of granular computing (pp. 293–332). Springer. https://doi.org/10.1007/978-3-030-64949-4_10
Zuev, D., Kalistratov, A., & Zuev, A. (2018). Machine learning in IT service management. Procedia Computer Science, 145, 675–679. https://doi.org/10.1016/j.procs.2018.11.063
Xu, J., He, R., Zhou, W., & Li, T. (2018). Trouble ticket routing models and their applications. In IEEE Transactions on Network and Service Management (pp. 530–543). IEEE. https://doi.org/10.1109/TNSM.2018.2790956