The data required for automatic optimization of user services usually exists in current systems, but that data is not modelled or linked in a way that facilitates automation. Knowledge engineering is a promising approach for managing the disparate communication service quality management information data sets and the links across those data sets. Once a knowledge base is in place, semantic techniques can be used to analyse and suggest optimizations to service quality. This paper describes our work in building, populating and evaluating a knowledge base for an IPTV service in Home Area Networks. Population of the knowledge base was implemented using terminal reports. The characteristics of the approach were evaluated through experimentation and the evaluation results are presented in this paper.