The analyst should assure herself that the interpretations of the data and models and
the conclusions drawn are conform to the knowledge of the problem owner or, if
available, domain expert. In particular the evidences, findings, and conclusions (not
only a final model) need to be documented throughout the process and presented at
this evaluation phase, where the results are discussed to decide if and how they may
be deployed later. There are at least three more reasons for documentation: First, it
is important that all steps are revised from the perspective of the project’s owner to
guarantee flawless interpretations and conclusions. Secondly, the findings represent
important resources for future projects, where data understanding and data preparation
phases may exhibit a large overlap with the current project. For instance, the
cognitive map may be revised or extended and possibly linked to evidence found in
the data. Finally, the drawbacks and problems faced may initiate improved data entering
procedures (to improve data quality where it is crucial) or establish new data
collections (data acquisition) such that future projects can benefit from an improved
situation.