README.md

---
task_categories:
- token-classification
- image-to-text
- object-detection
tags:
- invoices
- finance
- document-understanding
- layoutlm
- ner
features:
  id:
    dtype: int32
  words:
    dtype: string
    sequence: string
  boxes:
    dtype: int64
    sequence:
      sequence: int64
  ner_tags:
    class_label:
      names:
        0: O
        1: B-BALANCE_DUE
        2: B-BANK_NAME
        3: B-CLIENT_ADDRESS
        4: B-CLIENT_EMAIL
        5: B-CLIENT_NAME
        6: B-CLIENT_PHONE
        7: B-CLIENT_TAX_ID
        8: B-CURRENCY
        9: B-DUE_DATE
        10: B-IBAN
        11: B-INVOICE_DATE
        12: B-INVOICE_NUMBER
        13: B-ITEM_DESC
        14: B-ITEM_NO
        15: B-LINE_DISCOUNT
        16: B-LINE_GROSS_WORTH
        17: B-LINE_NET_WORTH
        18: B-LINE_UNIT_PRICE
        19: B-LINE_VAT
        20: B-ORDER_NUMBER
        21: B-QTY
        22: B-REFERENCE_NUMBER
        23: B-SELLER_ADDRESS
        24: B-SELLER_EMAIL
        25: B-SELLER_NAME
        26: B-SELLER_PHONE
        27: B-SELLER_TAX_ID
        28: B-SUBTOTAL
        29: B-TOTAL_GROSS_WORTH
        30: B-TOTAL_NET_WORTH
        31: B-TOTAL_VAT
        32: I-BANK_NAME
        33: I-CLIENT_ADDRESS
        34: I-CLIENT_EMAIL
        35: I-CLIENT_NAME
        36: I-CLIENT_PHONE
        37: I-INVOICE_NUMBER
        38: I-ITEM_DESC
        39: I-LINE_NET_WORTH
        40: I-LINE_UNIT_PRICE
        41: I-SELLER_ADDRESS
        42: I-SELLER_EMAIL
        43: I-SELLER_NAME
        44: I-SELLER_PHONE
        45: I-SELLER_TAX_ID
        46: I-TOTAL_NET_WORTH
  image:
    dtype: image
---

# Merged Invoice NER Dataset

## Dataset Description

This dataset consists of processed invoice documents intended for **Document Layout Analysis** and **Named Entity Recognition (NER)** tasks (e.g., training LayoutLM, LayoutLMv2, LayoutLMv3, or LiLT).

The data was created by merging multiple invoice datasets. It includes the document images, OCR-extracted words, bounding boxes, and NER tags.

### Supported Tasks
- **Token Classification:** Extracting entities like `INVOICE_NUMBER`, `TOTAL_GROSS_WORTH`, `SELLER_NAME`, etc.
- **Key Information Extraction (KIE)**

## Data Structure

Each example in the dataset contains the following fields:

- **`id`**: A unique identifier for the document.
- **`image`**: The PIL Image of the invoice.
- **`words`**: List of strings (tokens) obtained via OCR.
- **`boxes`**: List of bounding boxes corresponding to the words.
  - **Format:** `[x_min, y_min, x_max, y_max]`
  - **Note:** These boxes are likely in **absolute pixel coordinates** (based on the image size) and **NOT normalized** to 0-1000. You must normalize them before inputting them into models like LayoutLM.
- **`ner_tags`**: List of class IDs (integers) in BIO (Beginning, Inside, Outside) format.

## Important Usage Warnings

### 1. Bounding Box Normalization Required
The `boxes` in this dataset are likely stored in absolute pixel values (e.g., `0-2500`). Models like LayoutLM expect boxes normalized to a `0-1000` scale.

**How to handle this:**
When loading the dataset, you should normalize the boxes relative to the image size:

```python
def normalize_box(box, width, height):
    return [
        int(1000 * (box[0] / width)),
        int(1000 * (box[1] / height)),
        int(1000 * (box[2] / width)),
        int(1000 * (box[3] / height)),
    ]

def preprocess(example):
    w, h = example['image'].size
    # Normalize boxes to 0-1000
    example['boxes'] = [normalize_box(box, w, h) for box in example['boxes']]
    return example

dataset = load_dataset("your-username/dataset-name")
dataset = dataset.map(preprocess)

2. Mixed OCR Sources
This dataset was created by merging sources that may have used different OCR engines (e.g., Tesseract, EasyOCR, or commercial APIs).
Some documents may have slightly different bounding box tightness or word-splitting logic.
Ensure your training pipeline is robust to slight variations in OCR quality.
3. Missing I-Tags for some entities
The label map includes B- (Beginning) tags for all entities, but some entities (like BALANCE_DUE, TOTAL_VAT, IBAN) do not have a corresponding I- (Inside) tag in the schema.
If a value spans multiple words (e.g., "Total VAT"), the model may only be able to label the first word, or you may need to map the subsequent words to O or a generic tag during training.
4. Privacy / PII
This dataset contains invoice data. While processed, users should handle the data with care regarding Personally Identifiable Information (PII) such as names, addresses, and financial details.