ABSTRACT
Construction is the first process in inC's cascade. It builds up the current conceptual representation (ccr) by reading perceived entities (pes) from the preprocessing unit and calling the concept matcher to find more complex concepts. Depending on the concept matcher's results, construction modifies the ccr and sends a notification to selection. As the concept matcher interacts exclusively with construction, it could also be modeled as part of the construction task. Due to this close connection, this chapter not only describes the construction process but also the concept matcher and the concept storage (cs). Construction waits until the first pe is available in the input_stream (the wait statement). As soon as this is the case, the pe is fetched and inserted into the ccr. The cs is inC's second knowledge representation besides the ccr. In contrast to the ccr, it does not change over time; in particular, nothing is added, which would correspond to learning of categories.
