---
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
dataset_info:
  features:
  - name: image
    dtype: image
  - name: words
    sequence: string
  - name: bboxes
    sequence:
      sequence: float64
  - name: labels
    sequence: int64
  - name: grouped_words
    sequence:
      sequence: int64
  - name: linked_groups
    sequence:
      sequence: int64
  splits:
  - name: train
    num_bytes: 183288640.158
    num_examples: 1026
  - name: test
    num_bytes: 20706650
    num_examples: 113
  download_size: 195177090
  dataset_size: 203995290.158
extra_gated_prompt: >-
  You agree to not attempt to determine the identity of individuals in this
  dataset.
  You agree to the terms and conditions of the [FUNSD+ license](https://huggingface.co/datasets/konfuzio/funsd_plus/blob/main/LICENSE).
extra_gated_fields:
  Name: text
  Company: text
  Country: text
  Email: text
  I agree to the terms and conditions of the FUNSD+ license: checkbox
license: other
language:
- en
pretty_name: FUNSD+
size_categories:
- 1K<n<10K
tags:
- funsd
---
# Dataset Card for "funsd_plus"

## Table of Contents
- [Dataset Description](#dataset-description)
  - [Homepage](#homepage)
  - [Point of Contact](#point-of-contact)
  - [Languages](#languages)
- [Additional Information](#additional-information)
  - [Licensing Information](#licensing-information)
  - [Citation Information](#citation-information)

## Dataset Description

- **Homepage:** [FUNSD+ | A larger and revised FUNSD dataset by Konfuzio](https://konfuzio.com/en/funsd-plus/)
- **Point of Contact:** [https://konfuzio.com/en/contact/](https://konfuzio.com/en/contact/)
- **Languages:** `English`

## Additional Information

The [FUNSD dataset](https://guillaumejaume.github.io/FUNSD/) provides ground truth data for form understanding.

Following research, the [FUNSD+ dataset](https://konfuzio.com/de/en/funsd-plus/) addresses inconsistencies in labeling found in the original [FUNSD Vu and Nguyen (2020)](https://arxiv.org/abs/2010.05322), which could hinder its effectiveness for key-value extraction.

FUNSD+ significantly expands the document count from 199 to 1,113, while enhancing header, question, and answer statistics as outlined in the provided table. 

Besides the increase from 199 documents to 1113 documents we summarize the characteristics of both datasets below. Statistics of the FUNSD dataset are retrieved from the Paper by Jaume et al. (2019).

|  | FUNSD | FUNSD+ |
| --- | --- | --- |
| Documents | 199 | 1113 |
| headers | 563 | 1604 |
| questions | 4343 | 14695 |
| answers | 3623 | 12154 |
| questions with no answers | 720 (16.6%) | 2691 (18.3%) |
| answers without questions* | 0 | 114 (0.9%) |

Key differences emerge in the annotation process between FUNSD and FUNSD+. 

### Visual Example

![FUNSD-Plus-Document-Preview](https://konfuzio.com/wp-content/uploads/2022/10/image-3-1536x819.png)

Visit [https://app.konfuzio.com/d/303962/](https://app.konfuzio.com/d/303962/) to see an example.


The dataset is accessible via our [Python SDK](https://github.com/konfuzio-ai/konfuzio-sdk) or can be customized through the [Konfuzio Server](https://konfuzio.com/de/dokumenten-ki/). Users are encouraged to explore additional examples relevant to various document types on the official website. The latest research can be cited as follows:

Zagami, D., & Helm, C. (2022, October 18). FUNSD+: A larger and revised FUNSD dataset. Retrieved November 5, 2022, from [https://konfuzio.com/de/en/funsd-plus/](https://konfuzio.com/de/en/funsd-plus/) 


### Licensing Information

[FUNSD+ license](https://huggingface.co/datasets/konfuzio/funsd_plus/blob/main/LICENSE)

### Citation Information

```
@misc{zagami_helm_2022,
title = {FUNSD+: A larger and revised FUNSD dataset},
author = {Zagami, Davide and Helm, Christopher},
year = 2022,
month = {Oct},
journal = {FUNSD+ | A larger and revised FUNSD dataset},
publisher = {Helm & Nagel GmbH},
url = {http://konfuzio.com/funsd-plus/}
}
```