---
dataset_info:
  features:
  - name: title
    dtype: string
  - name: description
    dtype: string
  splits:
  - name: train
    num_bytes: 920689032
    num_examples: 1021106
  download_size: 630135488
  dataset_size: 920689032
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
language:
- en
size_categories:
- 1M<n<10M
---

# Goodreads Book Descriptions

A dataset of English book titles and descriptions from Goodreads.

The original dataset has 2.3 million books total with many more fields.

There may exist a small number of non-English books in this dataset.

# Citations

- Mengting Wan, Julian McAuley, ["Item Recommendation on Monotonic Behavior Chains"], in RecSys'18.
- Mengting Wan, Rishabh Misra, Ndapa Nakashole, Julian McAuley, ["Fine-Grained Spoiler Detection from Large-Scale Review Corpora"], in ACL'19.

["Item Recommendation on Monotonic Behavior Chains"]: https://mengtingwan.github.io/paper/recsys18_mwan.pdf
["Fine-Grained Spoiler Detection from Large-Scale Review Corpora"]: https://mengtingwan.github.io/paper/acl19_mwan.pdf