Deconstructing NLG Evaluation: Evaluation Practices, Assumptions, and Their Implications
This work addresses the problem of unclear evaluation practices in NLG for researchers and practitioners, but it is incremental as it builds on existing mapping of the evaluation landscape.
The paper tackles the difficulty of evaluating natural language generation (NLG) systems due to variability in expression and deployment-specific criteria, by conducting interviews and surveys with practitioners to uncover their goals, practices, and assumptions, revealing how these factors shape evaluations and embody ethical considerations.
There are many ways to express similar things in text, which makes evaluating natural language generation (NLG) systems difficult. Compounding this difficulty is the need to assess varying quality criteria depending on the deployment setting. While the landscape of NLG evaluation has been well-mapped, practitioners' goals, assumptions, and constraints -- which inform decisions about what, when, and how to evaluate -- are often partially or implicitly stated, or not stated at all. Combining a formative semi-structured interview study of NLG practitioners (N=18) with a survey study of a broader sample of practitioners (N=61), we surface goals, community practices, assumptions, and constraints that shape NLG evaluations, examining their implications and how they embody ethical considerations.