Understanding the Properties of Generated Corpora
This work addresses the problem of analyzing generated text for researchers in natural language processing, but it is incremental as it focuses on characterization rather than new generation methods.
The paper tackled the challenge of understanding the properties of automatically generated text corpora by proposing a set of tools to examine them, revealing significant differences in corpora produced by two leading generative models.
Models for text generation have become focal for many research tasks and especially for the generation of sentence corpora. However, understanding the properties of an automatically generated text corpus remains challenging. We propose a set of tools that examine the properties of generated text corpora. Applying these tools on various generated corpora allowed us to gain new insights into the properties of the generative models. As part of our characterization process, we found remarkable differences in the corpora generated by two leading generative technologies.