Skip to content

Latest commit

 

History

History
109 lines (97 loc) · 3.11 KB

File metadata and controls

109 lines (97 loc) · 3.11 KB

German news quotation attribution dataset

This repository contains the dataset and evaluation code for the paper "Dataset of Quotation Attribution in German News Articles" by Fynn Petersen-Frey and Chris Biemann (2024).

Data Format:

The data is available as JSON files where document (news article) is a individual JSON file.

Each JSON file has the following format:

{
    "annotations": [
        {
            "addressee": [],
            "cue": {
                "spans": [
                    {
                        "begin": 1,
                        "charBegin": 5,
                        "charEnd": 9,
                        "end": 2
                    }
                ],
                "text": "said",
                "tokenIds": [
                    1
                ]   
            },
            "frame": {
                "spans": [
                    {
                        "begin": 0,
                        "charBegin": 0,
                        "charEnd": 9,
                        "end": 2
                    }
                ],
                "text": "They said",
                "tokenIds": [
                    0, 1
                ]   
            },
            "hasNested": false,
            "id": 0,
            "isNested": false,
            "medium": "SW",
            "quote": ["0:2", "0:3", "0:4", "0:5"],
            "quote": {
                "spans": [
                    {
                        "begin": 2,
                        "charBegin": 10,
                        "charEnd": 29,
                        "end": 7
                    }
                ],
                "text": "this is a sentence.",
                "tokenIds": [
                    2, 3, 4, 5
                ]   
            },
            "speaker": {
                "spans": [
                    {
                        "begin": 0,
                        "charBegin": 0,
                        "charEnd": 4,
                        "end": 1
                    }
                ],
                "text": "They",
                "tokenIds": [
                    0
                ]   
            },
            "type": "Indirect"
        },
    ],
    "documentName": "Title of the document",
    "originalText": "They said this is a sentence.",
    "sentences": [
        {
            "begin": 0,
            "charBegin": 0,
            "charEnd": 29,
            "end": 7,
            "id": 0,
            "text": "They said this is a sentence.",
            "tokens": ["They", "said", "this", "is", "a", "sentence", "."]
        },
    ]
}

Data Source

The data for includes news articles from the German WIKINEWS, extracted from the articles XML dump. The entire dump from April 2022 consists of 13,001 published articles, from which we sampled 1000 articles to annotate. These annotated articles contain almost 250,000 tokens.

License

The data is provided under the CC BY-NC-SA 4.0 license.

Evaluation

The evaluation script can be found under the eval directory.