3. AFD-definition: introduction

An AFD-definition is used to exchange data with third parties and to store data in a database. The use of AFD-definitions leads to more efficient and effective processes in data processing for storage and messaging. The AFD-definition Standard ensures uniformity in the data and messages within the financial sector.

*An AFD-definition is defined as a set of specifications for a product or payload of a service, written in a common machine-readable language and based on AFD. The purpose is to validate messages and data against these specifications.

The AFD-definition Standard focuses on standardizing various aspects of the definition of services and/or products. The aim is that data storage and message traffic between partners in the insurance industry can take place unambiguously. This standard is suitable for use by all parties within the insurance industry and for several purposes. It can be applied for web services, for data storage, for batch processing, and for references to specific purposes (more information will become available in manual SIVI All Finance Standard). For more information about domain-specific use by different parties, see paragraph AFD-definition Protocols.

The AFD-definition Standard specifies the response to three focus areas:

• Data: which data should be communicated?
• Structure: what should be the structure of these data?
• Validations: what checks should be performed?
  

When using the AFD-definition Standard you validate two things: firstly the structure of the data, and secondly the relationship between the data:

• Data structure: checking whether the data elements are defined in accordance with product definitions. Example: is the mandatory data element ‘insured sum’ actually present and in the right place in the structure? This structure check takes place on the basis of a technical message schema (XML or JSON). This validation goal corresponds to the above-mentioned areas Data and Structure in the AFD-definition Standard.
• Data relation: validation rules to check data elements in relation to each other. Example: if data element ‘insured sum’ is higher than 40,000 euros, then data element ‘excess’ must be higher than 500 euros. XPath (in AFD 1.0) or JMESPath (in AFD 2.0) is used to create these validation rules. This corresponds to the above-mentioned area Validations in the AFD-definition Standard.
  

The AFD-definition Standard uses schemas and AFD validation rules to validate AFD-messages (in XML or JSON).

An AFD-definition consists of different components which are further specified in the following paragraphs.