Correlation of component information provided by diverse detection and data collection methods is used to discover additional components and identify and record relationships between these and still other components. Each detection and data collection method can be individually useful to discover components, and to provide at least some information about one or more discovered components to identify at least one relationship between the discovered components. When the information and component discoveries of a first detection and data collection method are properly combined with those of a second detection and data collection method, there can be a synergistic effect that greatly improves the utility and efficiency of a relationship-based network management system by enabling automated discovery of additional components and identification of additional component relationships, by automatically merging diverse views of a single component into a coherent picture of that component, and by automated prediction of component state changes based on state changes in related components.