|Bookshelf Home | Contents | Index | PDF|
When data matching is performed in real time or in batch mode, SDQ searches in the database for records that potentially match the current record (the record entered by a real-time user or the active record in the batch job). These records are called candidate records. When comparing the current records with existing records in the database, SDQ does not use raw data, but instead uses match key values.
Match keys are calculated by applying an algorithm to specified fields in customer records. Typically keys are generated from a combination of name, address, and other identifier fields, for example, a person's name (first name, middle name, last name) for prospects and contacts, or the account name for accounts. These fields are specified in a set of user properties for the appropriate Account, Contact, or Prospect business component. The way in which match keys are calculated differs for the Matching Server and Universal Connector as described in the following topics.
You generate match keys for records in the database by using batch jobs, as described in Generating or Refreshing Keys Using Batch Jobs.
Because key data can become out of sync with the base tables, you need to refresh the key data periodically. A key refresh updates keys only for records that are new or have been modified since your last key generation. Key refresh is much faster than key generation.
Typically, an administrator generates and refreshes keys on a periodic basis by running batch jobs. In such batch jobs, keys are generated for all account keys, all contact keys, all prospect keys, or subsets defined by search specifications that include an object WHERE clause.
For instructions about using batch jobs to generate or refresh keys, see Generating or Refreshing Keys Using Batch Jobs.
Additionally, if real-time data matching is enabled for users, keys are automatically generated (or refreshed) for a record whenever the user saves a new Account, Contact, or List Mgmt Prospective Contact record or modifies and commits an existing record to the database.
|Siebel Data Quality Administration Guide||Copyright © 2006, Oracle. All rights reserved.|