Use this wizard to specify the algorithms to use to find relationships
between a source and target in the mapping editor.
Specify Algorithms page
Use this page to select
one or more algorithms to control how relationships are found.
- Algorithm
- Select the check box for the algorithm that you want to use to find relationships.
You can select multiple algorithms and configure the values for each one.
The default algorithm is Lexical similarity. If the
target in the mapping editor is an .xsd file, then the algorithms that require
data sampling are disabled.
- Configuration for Lexical similarity algorithm
- Set the rejection threshold for the Lexical similarity algorithm. The
recommended value is 1.
- Composition by sequence
- If you selected more than one algorithm, specify the order in which you
want the algorithms applied to find relationships. To change the order, highlight
an algorithm and click the Move Up push button or the Move Down push
button. The recommended order is to place Lexical similarity first. Composition
by sequence is the default setting.
- Composition by weight
- If you selected more than one algorithm, select this radio button to specify
the percentage for each algorithm. By default, each algorithm is weighted
equally.
Specify Connection Information for Data Sampling and Caching
page
Use this page to specify the connection information for data
sampling and caching.
- Data sampling connection information
- Click the ellipsis push button to open the Data Sampling Connection wizard
so that you can specify the connection information for the .dbm file that
is shown in the DBM files column. You must specify this information
if one of the algorithms that you selected requires data sampling.
- Cache sample data
- Select to cache sample data. If you do not select this option, the data
is gathered from the data sources for each discovery run.
- In memory
- Select to cache the sample data in memory. This option shortens the time
to gather sample data for consecutive discovery runs. This setting is only
recommended for up to 50 tables.
- In database
- Select to cache the sample data in the database that you select from the
database list. Only DB2 for Linux, UNIX, and Windows databases can be selected
as a cache database. This setting is recommended if you have more than 50
tables.
- Clear Cache
- Click to drop all sample cache schemas and tables from the database that
is selected in the database list. Schemas whose name begins with "IBM_RDA_DISCOVERY_CACHE_"
and all of their contained tables are dropped from this database. By default,
the sample cache data is also dropped when closing the mapping editor.