Unleashing Unstructured Data's Value for Enterprise IT Asset Management

Within the enterprise, domain-specific knowledge is often recorded by IT processes or employees in the form of noisy, unstructured data. The question then arises on how to gain actionable insight from the volumes of un-structured data in order to improve the bottom line in an effective and timely manner. In this paper we propose a method on how to approach the issue within the realm of software license management (SLM). After providing some background materials in the early sections, we will describe the processes, business logic, and data model-ing components of our Ontology based solution. The first technical section ("CMDB as Semantic model and Se-mantic Reconciliation Framework") defines the CMDB information hierarchy needed to support various domain relationships that ultimately reconcile the semantic in-formation models. The following section describes how to identify and extract valid software product licenses and conditions of use from the noisy, unstructured pur-chase order data housed in legacy procurement sources. The subsequent section annotates and reconciles the unstructured data using semantic models with temporal contexts and a robust semantic reconciliation mediator. Following that we describe the application and use cas-es. Finally, we will present our observations and make recommendations gleaned from our experience and fu-ture development efforts.