selfkgdwy100kdbpyg.py

import json
import os
import pickle

import datasets

logger = datasets.logging.get_logger(__name__)

_SUBFIELD = "yg"

_VERSION = "1.0.0"

_DESCRIPTION = """\
    DWY100k-yg is a large-scale monolingual dataset extracted from DBpedia and YAGO3. The suffix yg means DBpedia 
    to YAGO3. And DWY100k-yg has 100,000 reference entity alignments.
"""

_CITATION = """\
    @inproceedings{sun2018bootstrapping,
    title={Bootstrapping Entity Alignment with Knowledge Graph Embedding.},
    author={Sun, Zequn and Hu, Wei and Zhang, Qingheng and Qu, Yuzhong},
    booktitle={IJCAI},
    volume={18},
    pages={4396--4402},
    year={2018}
}
"""

_URL = "https://dl.acm.org/doi/10.1145/3485447.3511945"

_PREFIX = "https://huggingface.co/datasets/matchbench/selfkg-dwy100k-dbpyg"

_URLS = {
    "source": f"{_PREFIX}/resolve/main/selfkg-dwy100k-dbp{_SUBFIELD}-src.zip",
    "target": f"{_PREFIX}/resolve/main/selfkg-dwy100k-dbp{_SUBFIELD}-tgt.zip",
    "pairs": f"{_PREFIX}/resolve/main/selfkg-dwy100k-dbp{_SUBFIELD}-pairs.zip",
}

class SelfkgDwy100kygConfig(datasets.BuilderConfig):
    """BuilderConfig for Selfkg-DWY100k."""

    def __init__(self, features, data_url, citation, url, label_classes=("False", "True"), **kwargs):
        """
        Args:
          **kwargs: keyword arguments forwarded to super.
        """
        super(SelfkgDwy100kygConfig, self).__init__(**kwargs)
        self.features = features
        self.label_classes = label_classes
        self.data_url = data_url
        self.citation = citation
        self.url = url

class DWY100kYg(datasets.GeneratorBasedBuilder):
    """DWY100k-yg: A Entity Alignment Dataset. From DBpedia to YAGO3."""

    BUILDER_CONFIGS = [
        SelfkgDwy100kygConfig(
            name="source",
            features=["column1", "column2", "column3"],
            citation="TODO",
            url="TODO",
            data_url="https://huggingface.co/datasets/matchbench/selfkg-dwy100k-dbpyg/resolve/main/selfkg-dwy100k-dbpyg.zip"
        ),
        SelfkgDwy100kygConfig(
            name="target",
            features=["column1", "column2", "column3"],
            citation="TODO",
            url="TODO",
            data_url="https://huggingface.co/datasets/matchbench/selfkg-dwy100k-dbpyg/resolve/main/selfkg-dwy100k-dbpyg.zip"
        ),
        SelfkgDwy100kygConfig(
            name="pairs",
            features=["left_id","right_id"],
            citation="TODO",
            url="TODO",
            data_url="https://huggingface.co/datasets/matchbench/selfkg-dwy100k-dbpyg/resolve/main/selfkg-dwy100k-dbpyg.zip"
        )
    ]

    def  _info(self) -> datasets.DatasetInfo:
        if self.config.name=="source":
            features = {feature: datasets.Value("string") for feature in self.config.features}
        elif self.config.name=="target":
            features = {feature: datasets.Value("string") for feature in self.config.features}
        elif self.config.name=="pairs":
            features = {feature: datasets.Value("int32") for feature in self.config.features}
        
        return datasets.DatasetInfo(description = _DESCRIPTION, 
                features = datasets.Features(features))


    def _split_generators(self, dl_manager):

        dl_dir = dl_manager.download_and_extract(self.config.data_url) or ""
        if self.config.name == "source":
            return [
            datasets.SplitGenerator(
                name="ent_ids",
                gen_kwargs={
                    "data_file": os.path.join(dl_dir, "id_ent_1"),
                    "split": "ent_ids",
                },
            ),
            datasets.SplitGenerator(
                name="rel_triples_id",
                gen_kwargs={
                    "data_file": os.path.join(dl_dir, "triples_1"),
                    "split": "rel_triples",
                },
            ),
            datasets.SplitGenerator(
                name="LaBSE_emb",
                gen_kwargs={
                    "data_file": os.path.join(dl_dir, "raw_LaBSE_emb_1.pkl"),
                    "split": "embedding",
                },
            ),
            ]
        elif self.config.name == "target":
            return [
            datasets.SplitGenerator(
                name="ent_ids",
                gen_kwargs={
                    "data_file": os.path.join(dl_dir, "id_ent_2"),
                    "split": "ent_ids",
                },
            ),
            datasets.SplitGenerator(
                name="rel_triples_id",
                gen_kwargs={
                    "data_file": os.path.join(dl_dir, "triples_2"),
                    "split": "rel_triples",
                },
            ),
            datasets.SplitGenerator(
                name="LaBSE_emb",
                gen_kwargs={
                    "data_file": os.path.join(dl_dir, "raw_LaBSE_emb_2.pkl"),
                    "split": "embedding",
                },
            ),
            ]
        elif self.config.name == "pairs":
            return [
            datasets.SplitGenerator(
                name="train",
                gen_kwargs={
                    "data_file": os.path.join(dl_dir, "ref_ent_ids"),
                    "split": "train",
                },
            ),
            datasets.SplitGenerator(
                name="valid",
                gen_kwargs={
                    "data_file": os.path.join(dl_dir, "valid.ref"),
                    "split": "valid",
                },
            ),
            datasets.SplitGenerator(
                name="test",
                gen_kwargs={
                    "data_file": os.path.join(dl_dir, "ref_ent_ids"),
                    "split": "test",
                },
            ),
            ]
        
    def _generate_examples(self, data_file, split):
        if split in ["LaBSE_emb"]:
            des = pickle.load(open(data_file,"rb"))
            i = -1
            for ent,ori_emb in des.items():
                i += 1
                yield i, {
                    "column1": ent,
                    "column2": ori_emb,
                    "column3": None
                }
        else:    
            f = open(data_file,"r", encoding='utf-8')
            data = f.readlines()
            for i in range(len(data)):
                if self.config.name in ["source", "target"]:
                    if split in ["ent_ids", "rel_ids"]:
                        row = data[i].strip('\n').split('\t')
                        yield i, {
                                    "id": row[0],
                                    "data": row[1],
                                }
                    elif split in ["rel_triples_id"]:
                        row = data[i].strip('\n').split('\t')
                        yield i, {
                                    "head_ent": row[0],
                                    "relation": row[1],
                                    "tail_ent": row[2]
                                }
                if self.config.name == "pairs":
                    row = data[i].strip('\n').split('\t')
                    yield i, {
                                "left_id": row[0],
                                "right_id": row[1]
                    }