The InfoSphere® MDM Probabilistic Matching Engine (PME) is an engine that can be used to perform probabilistic matching and comparison on record data.
Probabilistic matching algorithms calculate scores based upon weights associated with values for specific attributes and are used across all attributes configured for comparison. This method of matching takes into consideration the frequency of data occurrence. The algorithm calculates one composite weight that is a result of a probabilistic calculation of match and no-match weights for individually compared fields.
The PME configuration data is maintained within a Configuration project as a bundle that comprises a subset of the overall configuration. This bundle can be included in a composite bundle, which provides a means of grouping and deploying a set of artifacts to a server. The default naming convention for PME Bundle projects is the name of the Configuration project with the .pmeConfig suffix, though you can change the name of the PME bundle (not the suffix) to what you want. All Composite Bundle projects are suffixed with .cba.
It is also possible to regenerate a PME Configuration. You can right-click on either a Configuration or PME Bundle project and select Generate PME Configuration, or access this in the wizard and select the relevant project names. The PME Bundle project contents are refreshed from its associated Configuration, and all Composite Bundle projects that this specific Bundle project contributes to are also refreshed.