Actwhat

Actwhat is a term used in information science and AI discourse to describe a framework for capturing and interpreting the actions of agents—whether human, software, or robotic—and the surrounding context and resulting outcomes. The term functions as a conceptual shorthand to denote a triad consisting of the act performed, the contextual factors at the time, and the what or outcome produced.

Etymology and scope: Actwhat is not the name of a formal standard; it is a portmanteau created in online discussions to emphasize documenting acts and what happened. Its usage varies; some authors treat it as a methodology, others as a data schema. In practice, actwhat is used to frame how action data should be recorded and analyzed across different domains.

Definition and components: Actwhat seeks to standardize how actions are logged and interpreted by recording three elements: the action taken (act), the situational context (context), and the resulting data or outcome (what). This framework is intended to support auditing, reproducibility, and interoperability across systems.

Applications: In user experience research, actwhat can structure action logs tied to interface context. In autonomous systems, it can annotate agent decisions with environmental conditions and outcomes. In data governance, actwhat-like records support provenance and accountability by linking actions to their contexts and results.

Relation to other concepts: It intersects with data provenance, event schemas, audit trails, and activity recognition. Unlike simple event logs, actwhat emphasizes the connection among act, context, and outcome, aiming to improve clarity and traceability of actions in complex systems.

History and status: The term appears in a limited set of online discussions and speculative writings; there is no widely adopted standard or formal specification as of now. See also data provenance, event schema, audit trail, and activity recognition.