Designing a Symbolic Intermediate Representation for Neural Surface Realization
This addresses error reduction in neural natural language generation for applications like dialogue systems, though it is incremental as it builds on existing multi-stage approaches.
The paper tackles the problem of errors in neural NLG outputs by designing a symbolic intermediate representation for multi-stage generation, showing that it produces high-quality surface realization and outperforms the E2E challenge winner on the E2E dataset.
Generated output from neural NLG systems often contain errors such as hallucination, repetition or contradiction. This work focuses on designing a symbolic intermediate representation to be used in multi-stage neural generation with the intention of reducing the frequency of failed outputs. We show that surface realization from this intermediate representation is of high quality and when the full system is applied to the E2E dataset it outperforms the winner of the E2E challenge. Furthermore, by breaking out the surface realization step from typically end-to-end neural systems, we also provide a framework for non-neural content selection and planning systems to potentially take advantage of semi-supervised pretraining of neural surface realization models.