Labelling

Labelling, or labeling, is the process of attaching a label that conveys identifying information, instructions, or categorization to an object, dataset, or phenomenon. Labels can be physical, such as tags on products, or digital, such as metadata or annotations in a dataset. The purpose is to communicate meaning, enable identification, and support organization and safety.

In commerce and manufacturing, labels provide product name, ingredients or materials, origin, expiration or freshness indicators, and handling instructions. Safety and regulatory labels use standardized symbols and text to warn of hazards or ensure compliance with laws and standards. Packaging labels also support traceability and recall if needed.

In information management and computing, labeling enables indexing, searchability, access control, and governance. In data science and machine learning, data labeling or annotation involves human or automated tagging of data to create ground truth for training models. Tasks include image or video annotation, text classification, and transcription. Quality control, inter-annotator agreement, and bias mitigation are important considerations.

In linguistics and cognitive science, labeling assigns grammatical or discourse categories to linguistic units, facilitating analysis and parsing.

Spelling variants include labelling (British English) and labeling (American English). See also label, tagging, annotation.