From Conventional to Multiversion Data Warehouse - Evolving Application Domains of Data Warehousing and Mining - page 57

Merge into 'Vy.z';
Conventional database management systems
cannot process clauses like version and merge.
Therefore Morzy et. al. have implemented a
prototype tool to support this kind of extended
features, as shown in Figure 10.

without specifying any information about versions
in from clause of the query. ii) analysis transpar-
ency, by this transparency a user can retrieve
data from multiple versions without specifying
any information about versions in where clause
of the query. iii) projection transparency, by this
transparency a user can retrieve data from multiple
versions without specifying any information about
versions in select clause of the query.
Conventional database management systems
cannot provide transparency, therefore a prototype
is implemented by Nasir, et. al. 2006, as shown
below in Figure 11.

transparent querying Approach

According to this approach, extensions to query
language are not required for retrieval from
multiple versions of the data warehouse, and the
whole approach revolves around the concept of
transparency. 'Transparency refers to separation
of high-level semantics of a system from lower-
level implementation issues' (Ozsu, 1991). By
using this approach, a user can query MVDW
without paying any attention to versions and data
associated with each version.
In a query, transparency is provided at three
levels i) source transparency, by this transparency
a user can retrieve data from multiple versions

comparison of Approaches

Both SQL extension and transparent querying
approaches have advantages and disadvantages
of their own. Brief descriptions of these are as
follows:

Figure 11. Synthetic warehouse builder (adapted from Nasir, 2006)