Attend the Women in ML Symposium on December 7 Register now

kd_conv

References:

travel_dialogues

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:kd_conv/travel_dialogues')
  • Description:
KdConv is a Chinese multi-domain Knowledge-driven Conversionsation dataset, grounding the topics in multi-turn conversations to knowledge graphs. KdConv contains 4.5K conversations from three domains (film, music, and travel), and 86K utterances with an average turn number of 19.0. These conversations contain in-depth discussions on related topics and natural transition between multiple topics, while the corpus can also used for exploration of transfer learning and domain adaptation.
  • License: Apache License 2.0
  • Version: 0.0.0
  • Splits:
Split Examples
'test' 150
'train' 1200
'validation' 150
  • Features:
{
    "messages": {
        "feature": {
            "message": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "attrs": {
                "feature": {
                    "attrname": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "attrvalue": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "name": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "name": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "domain": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

travel_knowledge_base

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:kd_conv/travel_knowledge_base')
  • Description:
KdConv is a Chinese multi-domain Knowledge-driven Conversionsation dataset, grounding the topics in multi-turn conversations to knowledge graphs. KdConv contains 4.5K conversations from three domains (film, music, and travel), and 86K utterances with an average turn number of 19.0. These conversations contain in-depth discussions on related topics and natural transition between multiple topics, while the corpus can also used for exploration of transfer learning and domain adaptation.
  • License: Apache License 2.0
  • Version: 0.0.0
  • Splits:
Split Examples
'train' 1154
  • Features:
{
    "head_entity": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "kb_triplets": {
        "feature": {
            "feature": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "length": -1,
            "id": null,
            "_type": "Sequence"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "domain": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

music_dialogues

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:kd_conv/music_dialogues')
  • Description:
KdConv is a Chinese multi-domain Knowledge-driven Conversionsation dataset, grounding the topics in multi-turn conversations to knowledge graphs. KdConv contains 4.5K conversations from three domains (film, music, and travel), and 86K utterances with an average turn number of 19.0. These conversations contain in-depth discussions on related topics and natural transition between multiple topics, while the corpus can also used for exploration of transfer learning and domain adaptation.
  • License: Apache License 2.0
  • Version: 0.0.0
  • Splits:
Split Examples
'test' 150
'train' 1200
'validation' 150
  • Features:
{
    "messages": {
        "feature": {
            "message": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "attrs": {
                "feature": {
                    "attrname": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "attrvalue": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "name": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "name": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "domain": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

music_knowledge_base

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:kd_conv/music_knowledge_base')
  • Description:
KdConv is a Chinese multi-domain Knowledge-driven Conversionsation dataset, grounding the topics in multi-turn conversations to knowledge graphs. KdConv contains 4.5K conversations from three domains (film, music, and travel), and 86K utterances with an average turn number of 19.0. These conversations contain in-depth discussions on related topics and natural transition between multiple topics, while the corpus can also used for exploration of transfer learning and domain adaptation.
  • License: Apache License 2.0
  • Version: 0.0.0
  • Splits:
Split Examples
'train' 4441
  • Features:
{
    "head_entity": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "kb_triplets": {
        "feature": {
            "feature": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "length": -1,
            "id": null,
            "_type": "Sequence"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "domain": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

film_dialogues

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:kd_conv/film_dialogues')
  • Description:
KdConv is a Chinese multi-domain Knowledge-driven Conversionsation dataset, grounding the topics in multi-turn conversations to knowledge graphs. KdConv contains 4.5K conversations from three domains (film, music, and travel), and 86K utterances with an average turn number of 19.0. These conversations contain in-depth discussions on related topics and natural transition between multiple topics, while the corpus can also used for exploration of transfer learning and domain adaptation.
  • License: Apache License 2.0
  • Version: 0.0.0
  • Splits:
Split Examples
'test' 150
'train' 1200
'validation' 150
  • Features:
{
    "messages": {
        "feature": {
            "message": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "attrs": {
                "feature": {
                    "attrname": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "attrvalue": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "name": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "name": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "domain": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

film_knowledge_base

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:kd_conv/film_knowledge_base')
  • Description:
KdConv is a Chinese multi-domain Knowledge-driven Conversionsation dataset, grounding the topics in multi-turn conversations to knowledge graphs. KdConv contains 4.5K conversations from three domains (film, music, and travel), and 86K utterances with an average turn number of 19.0. These conversations contain in-depth discussions on related topics and natural transition between multiple topics, while the corpus can also used for exploration of transfer learning and domain adaptation.
  • License: Apache License 2.0
  • Version: 0.0.0
  • Splits:
Split Examples
'train' 8090
  • Features:
{
    "head_entity": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "kb_triplets": {
        "feature": {
            "feature": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "length": -1,
            "id": null,
            "_type": "Sequence"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "domain": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

all_dialogues

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:kd_conv/all_dialogues')
  • Description:
KdConv is a Chinese multi-domain Knowledge-driven Conversionsation dataset, grounding the topics in multi-turn conversations to knowledge graphs. KdConv contains 4.5K conversations from three domains (film, music, and travel), and 86K utterances with an average turn number of 19.0. These conversations contain in-depth discussions on related topics and natural transition between multiple topics, while the corpus can also used for exploration of transfer learning and domain adaptation.
  • License: Apache License 2.0
  • Version: 0.0.0
  • Splits:
Split Examples
'test' 450
'train' 3600
'validation' 450
  • Features:
{
    "messages": {
        "feature": {
            "message": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "attrs": {
                "feature": {
                    "attrname": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "attrvalue": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "name": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "name": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "domain": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

all_knowledge_base

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:kd_conv/all_knowledge_base')
  • Description:
KdConv is a Chinese multi-domain Knowledge-driven Conversionsation dataset, grounding the topics in multi-turn conversations to knowledge graphs. KdConv contains 4.5K conversations from three domains (film, music, and travel), and 86K utterances with an average turn number of 19.0. These conversations contain in-depth discussions on related topics and natural transition between multiple topics, while the corpus can also used for exploration of transfer learning and domain adaptation.
  • License: Apache License 2.0
  • Version: 0.0.0
  • Splits:
Split Examples
'train' 13685
  • Features:
{
    "head_entity": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "kb_triplets": {
        "feature": {
            "feature": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "length": -1,
            "id": null,
            "_type": "Sequence"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "domain": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}