Understanding the InfoSphere MDM Probabilistic Matching Engine data model

The InfoSphere® MDM Probabilistic Matching Engine uses Person and Organization functions to compare party records and determine their degree of similarity. However, the parties that are being compared must also be defined formally in the InfoSphere MDM Probabilistic Matching Engine data model.

The InfoSphere MDM Probabilistic Matching Engine data model is defined using three constructs:

• Records – An entity being compared; such as a person or an organization. A record is divided into attributes.
• Attributes – A major identifying component of a record. For example, an address or a name represents an attribute of a record. An attribute is further subdivided into fields.
• Attribute fields – A data component of an attribute. For example, an address can consist of a street name and number, a city, a province or state, a country, and a postal code.

The figure above represents the default InfoSphere MDM Probabilistic Matching Engine data model definitions for the Person and Organization entities and their constituent attributes that are assigned to matching functions. These correspond to the Person and Organization party entities in InfoSphere MDM Server. Also illustrated is a subset of attributes that comprise each record type.

The full data models for both Person and Organization consist of many more fields that are not assigned to matching functions. For a complete listing of the attributes comprising the InfoSphere MDM Probabilistic Matching Engine data model, see the section Mapping InfoSphere MDM Server business objects to the InfoSphere MDM Probabilistic Matching Engine data model.