README.md

metadata

license: cc-by-nc-2.0
task_categories:
  - text-generation
language:
  - en
tags:
  - prompt-learning
  - explainable-recommendation
  - aspect-learning

Dataset for MAPLE: Enhancing Review Generation with Multi-Aspect Prompt LEarning in Explainable Recommendation

• Note that the train.index only contain the data index belonging to the train split (same for other splits); they need to work with reviews.pickle to obtain the data.

// since the file is very large, the compressed format is uploaded and huggingface cannot detect the splits
// adding splits here for clarity
- config_name: default
  data_files:
  - split: train
    path: "nete_format_data/*/*/train.index"
  - split: validation
    path: "nete_format_data/*/*/validation.index"
  - split: test
    path: "nete_format_data/*/*/test.index"