Attend the Women in ML Symposium on December 7 Register now

guardian_authorship

参考:

cross_topic_1

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_topic_1')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:1.0.0
  • 拆分
拆分 样本
'test' 207
'train' 112
'validation' 62
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_genre_1

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_genre_1')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:13.0.0
  • 拆分
拆分 样本
'test' 269
'train' 63
'validation' 112
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_topic_2

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_topic_2')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:2.0.0
  • 拆分
拆分 样本
'test' 179
'train' 112
'validation' 90
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_topic_3

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_topic_3')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:3.0.0
  • 拆分
拆分 样本
'test' 152
'train' 112
'validation' 117
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_topic_4

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_topic_4')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:4.0.0
  • 拆分
拆分 样本
'test' 207
'train' 62
'validation' 112
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_topic_5

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_topic_5')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:5.0.0
  • 拆分
拆分 样本
'test' 229
'train' 62
'validation' 90
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_topic_6

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_topic_6')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:6.0.0
  • 拆分
拆分 样本
'test' 202
'train' 62
'validation' 117
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_topic_7

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_topic_7')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:7.0.0
  • 拆分
拆分 样本
'test' 179
'train' 90
'validation' 112
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_topic_8

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_topic_8')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:8.0.0
  • 拆分
拆分 样本
'test' 229
'train' 90
'validation' 62
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_topic_9

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_topic_9')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:9.0.0
  • 拆分
拆分 样本
'test' 174
'train' 90
'validation' 117
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_topic_10

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_topic_10')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:10.0.0
  • 拆分
拆分 样本
'test' 152
'train' 117
'validation' 112
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_topic_11

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_topic_11')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:11.0.0
  • 拆分
拆分 样本
'test' 202
'train' 117
'validation' 62
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_topic_12

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_topic_12')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:12.0.0
  • 拆分
拆分 样本
'test' 174
'train' 117
'validation' 90
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_genre_2

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_genre_2')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:14.0.0
  • 拆分
拆分 样本
'test' 319
'train' 63
'validation' 62
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_genre_3

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_genre_3')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:15.0.0
  • 拆分
拆分 样本
'test' 291
'train' 63
'validation' 90
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cross_genre_4

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:guardian_authorship/cross_genre_4')
  • 说明
A dataset cross-topic authorship attribution. The dataset is provided by Stamatatos 2013.
1- The cross-topic scenarios are based on Table-4 in Stamatatos 2017 (Ex. cross_topic_1 => row 1:P S U&W ).
2- The cross-genre scenarios are based on Table-5 in the same paper. (Ex. cross_genre_1 => row 1:B P S&U&W).

3- The same-topic/genre scenario is created by grouping all the datasts as follows.
For ex., to use same_topic and split the data 60-40 use:
train_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[:60%]+validation[:60%]+test[:60%]')
tests_ds = load_dataset('guardian_authorship', name="cross_topic_<<#>>",
                        split='train[-40%:]+validation[-40%:]+test[-40%:]')

Important: train+validation+test[:60%] will generate the wrong splits becasue the data is imbalanced

* See https://huggingface.co/docs/datasets/splits.html for detailed/more examples
  • 许可:无已知许可
  • 版本:16.0.0
  • 拆分
拆分 样本
'test' 264
'train' 63
'validation' 117
  • 特征
{
    "author": {
        "num_classes": 13,
        "names": [
            "catherinebennett",
            "georgemonbiot",
            "hugoyoung",
            "jonathanfreedland",
            "martinkettle",
            "maryriddell",
            "nickcohen",
            "peterpreston",
            "pollytoynbee",
            "royhattersley",
            "simonhoggart",
            "willhutton",
            "zoewilliams"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "topic": {
        "num_classes": 5,
        "names": [
            "Politics",
            "Society",
            "UK",
            "World",
            "Books"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "article": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}