Dataformatet

Dataformatet is a central concept in digital information systems, referring to the set of rules and conventions used to encode, store, and transport data. A data format specifies how data is serialized: the arrangement of bits or characters, the meaning of fields, and how metadata is represented and accessed. Formats can be text-based or binary, and they can be designed for human readability, machine efficiency, or interoperability.

Text-based formats include CSV, JSON, XML, YAML and INI files. Binary formats include Parquet, Avro, Protobuf, MessagePack, HDF5 and many specialized formats for images, audio, and scientific data. Some formats are self-describing (for example JSON and XML, which include structure and data types within the file), while others rely on external schemas (for instance CSV with a header or a database schema). Data formats are often designed with a particular use case in mind: simple data interchange, high-performance analytics, streaming, or archival.

Key properties of data formats include schema or metadata support, portability across platforms and programming languages, and support for validation and versioning. Interoperability and open standards are important considerations, as they affect long-term accessibility and ecosystem tooling. When choosing a data format, practitioners consider factors such as data size, speed of encoding/decoding, human readability, compatibility with existing systems, and the availability of libraries and validators.

The term "dataformatet" is used across computing disciplines to discuss data modeling, serialization, and interchange, and it overlaps with topics in data modeling, data governance, and information architecture.