This topic describes how to attach domains/composite domains in a data quality knowledge base to a reference data service in Azure Marketplace to build knowledge against the high-quality reference data. Each reference data service contains a schema (data columns). After attaching a domain or a composite domain to a reference data service, you must map the attached domain or the individual domains within the attached composite domain to the appropriate columns in a reference data service schema. Attaching a composite domain to a reference data service enables you to attach just one domain to a reference data service, and then map the individual domains within the composite domain to appropriate columns in the reference data service schema.

This topic describes how to back up and restore the DQS databases.

A knowledge base in [!INCLUDE[ssDQSnoversion](../includes/ssdqsnoversion-md.md)] (DQS) is a repository of knowledge about your data that enables you to understand your data and maintain its integrity. A knowledge base consists of domains, each of which represents the data in a data field. The knowledge base is used by DQS to perform data cleansing and deduplication on a database. To prepare the knowledge base for data cleansing, you can run a computer-assisted analysis of a data sample, and interactively manage values in the domains. DQS enables you to import knowledge, create rules and relationships, change data values directly, and leverage a default database.

This topic describes how to change and augment the metadata in a knowledge base in [!INCLUDE[ssDQSnoversion](../includes/ssdqsnoversion-md.md)] (DQS). After you generate knowledge by knowledge discovery, import knowledge into the knowledge base or domains, or base a knowledge base upon another knowledge base, you can interactively change the data values. Knowledge base generation not only leverages computer-assisted processes, but gives you the means to use your own knowledge to verify data values and change them in the following ways:

This topic provides information about cleansing of composite domains in [!INCLUDE[ssDQSnoversion](../includes/ssdqsnoversion-md.md)] (DQS). A composite domain consists of two or more single domains, and maps to a data field that consists of multiple related terms. The individual domains in a composite domain must have a common area of knowledge. For detailed information about composite domains, see [Managing a Composite Domain](../data-quality-services/managing-a-composite-domain.md).

This topic describes how to cleanse your data by using a data quality project in [!INCLUDE[ssDQSnoversion](../includes/ssdqsnoversion-md.md)] (DQS). Data cleansing is performed on your source data using a knowledge base that has been built in DQS against a high-quality data set. For more information, see [Building a Knowledge Base](../data-quality-services/building-a-knowledge-base.md).

This topic describes how to cleanse data using knowledge from the reference data providers. While all the steps of running a cleansing activity remains the same for cleansing your data using knowledge from the reference data providers as explained in the [Cleanse Data Using DQS (Internal) Knowledge](../data-quality-services/cleanse-data-using-dqs-internal-knowledge.md), this topic provides information specific to data cleansing using reference data service in [!INCLUDE[ssDQSnoversion](../includes/ssdqsnoversion-md.md)] (DQS).

This topic describes how to configure advanced settings for [!INCLUDE[ssDQSServer](../includes/ssdqsserver-md.md)] and [!INCLUDE[ssDQSClient](../includes/ssdqsclient-md.md)] log files, such as set the rolling file size limit of the log files, set the time stamp pattern of the events, and so on.

This topic describes how to configure [!INCLUDE[ssDQSnoversion](../includes/ssdqsnoversion-md.md)] (DQS) to use reference data for cleansing your data. You could either use reference data from Azure Marketplace or from direct online third-party reference data providers.