Faceted Model for Transparent AI-Assisted Text Attribution
A recent publication on arXiv introduces a faceted framework aimed at enhancing the clarity of AI-assisted text attribution. This model tackles the issue that existing disclosure methods typically reveal the use of AI but fail to clarify how, where, or to what degree it was utilized. The fundamental model comprises three facets: Form, Generation, and Evaluation, while an expanded version incorporates Intent, Control, and Traceability. Intended as a basic operational standard, it can be adapted for more detailed representations. An example illustrating the article's creation showcases its practical use. This paper falls under the category of Computer Science > Computers and Society.
Key facts
- The paper proposes a faceted model for transparent attribution of AI-assisted text production.
- Current disclosure practices usually indicate AI use but not how it was used.
- The core model includes Form, Generation, and Evaluation facets.
- The extended model adds Intent, Control, and Traceability.
- The model is positioned as a minimal operational baseline with extensibility.
- A worked example based on the production of the article demonstrates applicability.
- The paper is categorized under Computer Science > Computers and Society.
- The paper is available on arXiv.
Entities
Institutions
- arXiv