A key aspect of Information Technology Service Management (ITSM) is the monitoring and evaluation of service performance - a task that is complicated by the presence of interrelationships among different service processes in a multiservice contract. While success in the service arrangement requires participant organizations' knowledge about the nature of service dependency and their subsequent effect on performance measures; such information is often tacitly present in the service level agreement/contract documents. In this context, the aim of our research is extracting information that might be hidden in the service contracts to assist in better process management and contract (re)negotiation. We propose an information extraction driven framework for analyzing Service Level Agreements (SLA) for IT services. Our framework consists of three stages - 1) Service Entity Recognition, 2) Service Entity Context Recognition, and 3) Service Interdependency Analysis. In this article the focus is on stage 1, where we identify interrelationships by using domain ontology on a set of annotated industry-standard SLAs. Our ongoing research is aimed towards the creation and subsequent validation of process models from the information extracted from SLAs that will help both customer and service provider organizations in contract and compensation formulation, resource allocation, and SLA life cycle management.