amazon.science commerce natural language processing
This dataset provides a set of non-shoppable items, which are items that can't be purchased via a virtual assistant (love, vampires, etc). In addition, for each non-shoppable item, the dataset contains humorous responses generated by two different large language models, and a template-based generation solution that uses a commonsense knowledge graph called ConceptNet. Finally, each row contains a score provided by human annotators that judge how funny each response is. The columns provided for each datapoint are as follows: question- purchase request of the non-shoppable item answer- a generated response model- the name of the model or solution that generated the response non_shoppable_item- the name of the non shoppable item relation- commonsense relation that was used to generate the response in the template-based approach prompt_or_template- the prompt that was used to generate the responses score- the average response score of the human evaluation
Not currently being updated
Dataset is published under CDLA-Permissive.
https://shopping-humor-generation.s3.amazonaws.com/README.txt
See all datasets managed by Amazon.
orenk@amazon.com, alibov@amazon.com, sofiato@amazon.com
Shopping Humor Generation was accessed on DATE from https://registry.opendata.aws/shopping-humor-generation.
arn:aws:s3:::shopping-humor-generationus-east-1aws s3 ls --no-sign-request s3://shopping-humor-generation/