license: apache-2.0
language:
- en
tags:
- NER
- finance
- NLP
pretty_name: Financial Named Entity Recognition Natural Language Processing
Dataset Card for Financial-NER-NLP Dataset Summary
The Financial-NER-NLP Dataset is a derivative of the FiNER-139 dataset, which consists of 1.1 million sentences annotated with 139 XBRL tags. This new dataset transforms the original structured data into natural language prompts suitable for training language models. The dataset is designed to enhance models’ abilities in tasks such as named entity recognition (NER), summarization, and information extraction in the financial domain.
The Financial-NER-NLP Dataset retains the financial domain’s specificity, focusing on numeric tokens and context-based tagging, while providing a more accessible and intuitive format for training natural language processing models. Key Features:
Natural Language Formatting: The dataset converts the original structured annotations into conversational prompts, making it suitable for training in a question-answering or dialogue-based format.
Multiple Formats: Available in TXT, JSONL, and CSV formats for ease of use in various machine learning pipelines.
Retains Financial Domain Focus: Preserves the emphasis on context-based tagging of numeric tokens, critical for financial document processing.
Supported Tasks
This dataset supports various NLP tasks, including but not limited to:
Named Entity Recognition (NER): Identifying and classifying entities within financial text.
Question Answering (QA): Training models to extract relevant information from financial documents based on user queries.
Summarization: Assisting in the generation of concise summaries of financial reports.
Data Augmentation: Providing a rich source of natural language data for augmenting existing financial NLP datasets.
Languages
This dataset is compiled from English-language financial reports and maintains the language characteristics of the original dataset. Acknowledgments
This dataset is based on the FiNER-139 dataset, created and released by nlpaueb. We would like to express our gratitude to the original creators for their valuable work, which has enabled the development of this derivative dataset. Citation
If you use this dataset, please cite the original FiNER-139 dataset as follows:
bibtex
@inproceedings{nlpaueb_finer139_2023, title={FiNER-139: Financial Named Entity Recognition with Extensive Business Reporting Language Tags}, author={NLPAUEB}, year={2023}, publisher={Hugging Face}, url={https://huggingface.co/datasets/nlpaueb/finer-139} }