Incident Management Optimization through the Reuse of Experiences and Natural Language Processing

Abstract

Information Technology (IT) Governance need for tools and processes that assist the qualification, retention and reuse of solutions to leverage the management of incidents. In this sense, this work proposes a process for resolving incidents, whose differential is the use of CBR (Case Based Reasoning) and NLP (Natural Language Processing) techniques, aiming the reuse of experience in handling similar incidents. Specially, NLP techniques, such as morph syntactic analysis and semantic textual similarity, are used in the retrieval phase of the CBR system. We applied the proposed process in a real scenario after the deployment of an ERP system in a mixed capital company, in order to validate if the use of CBR and NLP components in an Incident Management Process optimizes the handling of incidents.