---
dataset_info:
  features:
  - name: reviewerID
    dtype: string
  - name: asin
    dtype: string
  - name: reviewerName
    dtype: string
  - name: reviewText
    dtype: string
  - name: overall
    dtype: float64
  - name: summary
    dtype: string
  - name: cleanReviewTime
    dtype: string
  - name: reviewTime
    dtype: string
  splits:
  - name: train
    num_bytes: 349520
    num_examples: 965
  download_size: 218178
  dataset_size: 349520
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
task_categories:
- text-classification
language:
- fra
---

## Description
Cleaned version of the `Arts Crafts and Sewing` subset (`reviews` folder) of [XMRec dataset](https://xmrec.github.io/data/fr/).  

Columns meaning:
- `asin` is a code which can be used as a key to merge the dataset with [XMRec_metadata_fr_Arts_Crafts_and_Sewing](https://huggingface.co/datasets/CATIE-AQ/XMRec_metadata_fr_Arts_Crafts_and_Sewing) containing all the informations about the product to which the customer review refers
- `summary` is the term used in the original dataset, can be seen as a preview of the review
- `reviewText` contains the text of the review
- `overall` is the customer's rating of the product

This dataset is intended to be used for text classification.


## Original paper citation
```
@inproceedings{bonab2021crossmarket,
	author = {Bonab, Hamed and Aliannejadi, Mohammad and Vardasbi, Ali and Kanoulas, Evangelos and Allan, James},
	booktitle = {Proceedings of the 30th ACM International Conference on Information \& Knowledge Management},
	publisher = {ACM},
	title = {Cross-Market Product Recommendation},
	year = {2021}}
```