ESG-DLT-NER / README.md
walterhernandez's picture
Upload dataset
86de2ca verified
metadata
language:
  - en
license: cc-by-4.0
size_categories:
  - 10K<n<100K
task_categories:
  - token-classification
tags:
  - Distributed Ledger Technology
  - Blockchain
  - ESG
  - Named Entity Recognition
  - Environmental, Social, and Governance
dataset_info:
  features:
    - name: text
      dtype: string
    - name: input_ids
      sequence: int64
    - name: attention_mask
      sequence: int64
    - name: labels
      sequence:
        class_label:
          names:
            '0': O
            '1': B-Blockchain_Name
            '2': I-Blockchain_Name
            '3': B-Codebase
            '4': I-Codebase
            '5': B-Consensus
            '6': I-Consensus
            '7': B-ChargingAndRewardingSystem
            '8': I-ChargingAndRewardingSystem
            '9': B-ESG
            '10': I-ESG
            '11': B-Extensibility
            '12': I-Extensibility
            '13': B-Identifiers
            '14': I-Identifiers
            '15': B-Identity_Management
            '16': I-Identity_Management
            '17': B-Miscellaneous
            '18': I-Miscellaneous
            '19': B-Native_Currency_Tokenisation
            '20': I-Native_Currency_Tokenisation
            '21': B-Security_Privacy
            '22': I-Security_Privacy
            '23': B-Transaction_Capabilities
            '24': I-Transaction_Capabilities
    - name: ner_tags
      sequence:
        class_label:
          names:
            '0': O
            '1': B-Blockchain_Name
            '2': I-Blockchain_Name
            '3': B-Codebase
            '4': I-Codebase
            '5': B-Consensus
            '6': I-Consensus
            '7': B-ChargingAndRewardingSystem
            '8': I-ChargingAndRewardingSystem
            '9': B-ESG
            '10': I-ESG
            '11': B-Extensibility
            '12': I-Extensibility
            '13': B-Identifiers
            '14': I-Identifiers
            '15': B-Identity_Management
            '16': I-Identity_Management
            '17': B-Miscellaneous
            '18': I-Miscellaneous
            '19': B-Native_Currency_Tokenisation
            '20': I-Native_Currency_Tokenisation
            '21': B-Security_Privacy
            '22': I-Security_Privacy
            '23': B-Transaction_Capabilities
            '24': I-Transaction_Capabilities
    - name: tokens
      sequence: string
    - name: paper_name
      dtype: string
    - name: __index_level_0__
      dtype: int64
  splits:
    - name: train
      num_bytes: 108441309
      num_examples: 5813
  download_size: 11419998
  dataset_size: 108441309
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

Dataset Card for ESG/DLT Named Entity Recognition Dataset

This dataset contains named entities related to Distributed Ledger Technology (DLT) and Environmental, Social, and Governance (ESG) topics created to support research in these areas and at the intersection of these domains.

Dataset Details

Dataset Description

  • Curated by: Walter Hernandez Cruz, Kamil Tylinski, Ali Irzam Kathia, Alastair Moore, Niall Roche, Nikhil Vadgama, Horst Treiblmaier, Jiangbo Shangguan, Jiahua Xu, Paolo Tasca
  • Language(s) (NLP): English
  • Number of Entity: 12
  • Entity Types: Blockchain Name, Consensus, Identifiers, Security Privacy, ESG, Transaction Capabilities, ChargingAndRewardingSystem, Extensibility, Identity Management, Native Currency Tokenisation, Native Currency Tokenisation, Miscellaneous
  • License: CC BY-NC 4.0

Dataset Sources

Use

This dataset can be used for training and evaluating Named Entity Recognition models focused on DLT and ESG topics. It's particularly useful for researchers and practitioners working on text mining and information extraction in these domains.

Dataset Structure

The dataset contains 39,427 named entities organized into 12 top-level categories with 136 labels in a tree structure. It includes entities related to blockchain names, consensus mechanisms, transaction capabilities, security and privacy, and ESG concepts.

Label ID

The label2id dictionary is:

    {
        "O": 0,
        "B-Blockchain_Name": 1,
        "I-Blockchain_Name": 2,
        "B-Codebase": 3,
        "I-Codebase": 4,
        "B-Consensus": 5,
        "I-Consensus": 6,
        "B-ChargingAndRewardingSystem": 7,
        "I-ChargingAndRewardingSystem": 8,
        "B-ESG": 9,
        "I-ESG": 10,
        "B-Extensibility": 11,
        "I-Extensibility": 12,
        "B-Identifiers": 13,
        "I-Identifiers": 14,
        "B-Identity_Management": 15,
        "I-Identity_Management": 16,
        "B-Miscellaneous": 17,
        "I-Miscellaneous": 18,
        "B-Native_Currency_Tokenisation": 19,
        "I-Native_Currency_Tokenisation": 20,
        "B-Security_Privacy": 21,
        "I-Security_Privacy": 22,
        "B-Transaction_Capabilities": 23,
        "I-Transaction_Capabilities": 24
    }

Dataset Creation

Curation Rationale

The dataset was created to address the scarcity of labeled NLP data for blockchain research, focusing on the intersection of DLT and ESG topics.

Source Data

Data Collection and Processing

The dataset was created by manually annotating 80 publicly available publications using the brat tool and argilla. The taxonomy framework from Tasca and Tessone (2019) was extended to include ESG-related concepts.

Annotations

Annotation process

The annotation process involved manual labeling using the brat tool and argilla, following an extended version of the Tasca and Tessone (2019) taxonomy. Inter-labeler consistency was improved through systematic processes and programmatic cleaning.

Who are the annotators?

The annotators are the research paper's authors and other collaborators involved in the project.

Personal and Sensitive Information

The dataset does not contain personal or sensitive information as it is based on publicly available academic publications.

Bias, Risks, and Limitations

Recommendations

Users should be aware of potential biases in the dataset due to the selection of source publications and the annotation process.

Glossary

  • DLT: Distributed Ledger Technology
  • ESG: Environmental, Social, and Governance
  • NER: Named Entity Recognition

More Information

For more details about the dataset creation process and its applications, please refer to the associated research paper: https://arxiv.org/abs/2308.12420

Citation Information

@misc{hernandez2024evolutionesgfocuseddltresearch,
      title={Evolution of ESG-focused DLT Research: An NLP Analysis of the Literature}, 
      author={Walter Hernandez and Kamil Tylinski and Alastair Moore and Niall Roche and Nikhil Vadgama and Horst Treiblmaier and Jiangbo Shangguan and Paolo Tasca and Jiahua Xu},
      year={2024},
      eprint={2308.12420},
      archivePrefix={arXiv},
      primaryClass={cs.IR},
      url={https://arxiv.org/abs/2308.12420}, 
}

Contributions

Thanks to Ali Irzam Kathia for his contribution to labeling this dataset.