Managing Case Analyzer data fields
To use the values of case and workflow data fields in Case Analyzer reports, you must identify which data fields will be exposed, specify their properties as dimensions or measures, and specify the appropriate OLAP cube to store the data. The values for case and workflow data fields are stored in the Case Analyzer store.
About this task
Before you make a case or task property or a data field available to those who will use this information to analyze workflows and cases, the property or field must be captured in the workflow system event log or object store audit log.
- In Process Designer, define a data field in a workflow definition. For more information, see Workflow properties - data fields.
- In the administration console, create a database field in the event log for that data field. For more information, see Managing user database fields.
To make case or task properties available from Case Manager
systems, see Integrating IBM case analytics tools
If a data field that you want to make available occurs in both the event log and audit log, then you need to create only a single Case Analyzer data field to retrieve the data values from both logs. To enable this, the data field name and type must be the same in the event log and audit log. For example, if a data field named LoanAmt of data type float is exposed in the event log and audit log, then you will create a single Case Analyzer data field of type float named LoanAmt to pull the data from both sources to the Case Analyzer database.
Data fields can be created as dimensions or measures:
- Dimensions provide meaningful statistical information about an
item of business significance. A large dimension (a dimension with
many members) is hard for a user to comprehend unless the dimension
provides meaningful data. For example, defining the social security
number (SSN) as a dimension results in a large number of dimension
members, with little or no statistical value per member. On the other
hand, defining a dimension as the first three numbers of the SSN,
which indicate the issuing state, can provide meaningful groupings
of statistical information where there are many workflow events with
different SSNs. Statistical analysis can then be performed on the
resulting groups.
Any data field type can be a dimension. For data fields of type float, integer, and time, you have the option of aggregating the data. For example, if a data field is an amount, you can categorize the amount field into ranges of 0-10, 10-100, 100-1000, and above 1000. Aggregating dimension data saves on storage space; if you choose not to aggregate the data, all the values are stored as members in the dimension, which yields large dimensions.
Important: Large dimensions (even less than 64,000 members) can be problematic to Excel. Consider a third-party application if Excel does not serve your purpose with large dimensions. Large dimensions also increase the memory footprint of Analysis Services. - Measures provide an aggregate value for a data field, such as a sum or average. Because measures are used for aggregation functions, only data fields of type integer or float can be created as measures. The default aggregation function for the measure is Sum.