A Method for Data Warehouse Design - Data Warehouse Systems: Design and Implementation - page 415

Determining the activities that could improve these evaluation
criteria required the participation of users at various organizational
levels. Interviews with users allowed us to conclude that, in the first
step, information related to international conferences, projects, and
publications was necessary to better understand the participation of
the university's staff in international forums.
Participation in international conferences helps the university's staff
to meet international colleagues working in the same or a similar
area. In this way, not only can new strategic contacts be established
(which may lead to international projects) but also the quality of the
university's research can be improved.
Further, international projects promote the interaction of the
university staff with peers from other universities working in the same
area and thus could help to improve the peer review score .Thereare
several sources of funding for research projects: the university, industry,
and regional, national, and international institutions. Independently of
the funding scheme, a project may be considered as being international
when it involves participants from institutions in other countries.
Finally, knowledge about the international publications produced by
the university's staff is essential for assessing the citation per faculty
criterion . Publications can be of several types, namely, articles in
conference proceedings or in journals, and topics.
Basedonthedescriptionabove,weaskyouto
(a) Produce a requirements specification for the design of the data
warehouse using the analysis-driven approach . For this, you must
￿ Identify users.
￿ For each goal and subgoal, write a set of queries that these users
would require. Refine and prioritize these queries.
￿ Define facts, measures, and dimensions based on these queries.
￿ Infer dimension hierarchies.
￿ Build a table summarizing the information obtained.
(b) Produce a conceptual schema, using the analysis-driven approach
and the top-down design . Discuss data availability conditions and
how they impact on the design. Identify and specify the necessary
mappings.
(c) Produce a conceptual schema, using the analysis-driven approach ,
and the bottom-up design . For this, you must build three data marts:
one for the analysis of conferences, another one for the analysis of
publications, and the third one for the analysis of research projects.
Then, merge the three of them, and compare the schema produced
with the one obtained through the top-down approach above.
(d) Produce a requirements specification for the design of the data
warehouse using the source-driven approach , given the entity-