As you were understanding of the jigsaw puzzle increases with every piece of the puzzle that you add in a greater bigger picture a clear bigger picture starts to emerge. This is a probabilistic kind of evolving picture a picture with increasing clarity. As the pictures clarity increases a more holistic view of The domain starts to emerge this is called context. Context will encourage you to look for data. In fact you are not looking for any kind of data, you are looking for very specific data that is related to the existing context that you have established.
The context you establish can be viewed as a knowledge network a network of data that is connected via relationships you can think of it as a set of entities and the interrelationships between those entities but in fact that is only a static view of this knowledge graph of this context graph.
The context graph is formed by starting with some static data entities and semantically relating them using relationships that make sense for that domain.
As the jigsaw puzzle of context gets clear for that given domain and you see potential patterns emerging or at the very least you see certain gaps that are appearing as anomalies in an otherwise potentially repeating set of patterns, you will start to think about traversals.
Graph traversals are the dynamic component of context crafts a context graph consists of an entity relationship diagram along with the meta-data that is necessary to determine for a given domain in a given use case what traversal pads will most likely yield probably Stickley significant pieces of relevant information.
contextual computing 