The article presents an approach to the automatic derivation of conceptual
database models from heterogeneous source artifacts. The approach is based
on the integration of conceptual database models that are derived from
source artifacts of one single type by already existing tools, whereby those
models possess limited certainty given their limited completeness and
correctness. The uncertainty of the automatically derived models from
specific source artifacts is expressed and managed through the effectiveness
measure of the generation of specific concepts of the input conceptual
database models. The approach is implemented by the DBomnia tool - the
first online web-based tool enabling automatic derivation of conceptual
database models from heterogeneous source artifacts (business process models
and textual specifications). DBomnia employs other pre-existing tools to
derive conceptual models from sources of the same type and then integrates
those models. The case study-based evaluation proves that the implemented
approach enables effective automatic derivation of the conceptual database
model from a set of heterogeneous source artifacts. Moreover, the automatic
derivation of the conceptual database model from a set of heterogeneous
source artifacts is more effective than each independent automatic
derivation of the conceptual database model from sources of one single type
only.