The term “unstructured data,” in computer science, goes to define digital information in raw form, that is, having a structure that cannot be identified.
Unlike Structured Data, the information cannot thus be mapped into schemas and databases and is therefore difficult to use in the case of automated analysis processes.
This category includes, for example, comments on social media, product reviews, audio, video and image content.
Depending on their degree of structuring, data can belong to high two different macro-categories than the one previously illustrated.
Let us analyse them in detail:
- structured data: they respond to a well-defined framework and structure. Precisely because of this nature they can be easily processed and modified by computer systems. Indeed, information has precise positions within rows and columns and can be easily edited and queried;
- semi-structured data: such data respond to a basic structural framework, that is, they possess metadata that help define particular characteristics that facilitate their cataloging.
Unstructured data, as is readily apparent can be stored and saved more quickly but requires more effort in terms of both storage and protection and security.
Typically, structured data is commonly stored in CRMs and data warehouses, while unstructured data is stored in data lakes.
