Add dataset

README.md

@@ -0,0 +1,14 @@
+# Wikidata Simplequestions
+
+Huggingface Dataset wrapper for Wikidata-simplequestion dataset
+
+### Usage
+
+```bash
+git clone git@github.com:skoltech-nlp/wikidata-simplequestions-hf.git wikidata_simplequestions
+```
+
+```python3
+from datasets import load_dataset;
+load_dataset('../wikidata_simplequestions', 'answerable_en', cache_dir='/YOUR_PATH_TO_CACHE/', ignore_verifications=True)
+```

wikidata_simplequestions.py

@@ -0,0 +1,129 @@
+import csv
+import json
+import os
+from typing import Any
+
+import datasets
+from datasets.utils import logging
+
+
+_DESCRIPTION = """\
+HuggingFace wrapper for https://github.com/askplatypus/wikidata-simplequestions dataset
+Simplequestions dataset based on Wikidata.
+"""
+
+# TODO: Add a link to an official homepage for the dataset here
+_HOMEPAGE = ""
+
+# TODO: Add the licence for the dataset here if you can find it
+_LICENSE = ""
+
+_LANGS = [
+    "ru",
+    "en",
+]
+
+_URL = "https://raw.githubusercontent.com/askplatypus/wikidata-simplequestions/master/"
+_DATA_DIRECTORY = "./simplequestion"
+VERSION = datasets.Version("0.0.1")
+
+
+class WikidataSimpleQuestionsConfig(datasets.BuilderConfig):
+    """BuilderConfig for WikidataSimpleQuestions."""
+
+    def __init__(self, **kwargs):
+        """BuilderConfig for WikidataSimpleQuestions.
+        Args:
+          **kwargs: keyword arguments forwarded to super.
+        """
+        super(WikidataSimpleQuestionsConfig, self).__init__(**kwargs)
+
+
+class WikidataSimpleQuestions(datasets.GeneratorBasedBuilder):
+    """HuggingFace wrapper for https://github.com/askplatypus/wikidata-simplequestions dataset"""
+
+    BUILDER_CONFIG_CLASS = WikidataSimpleQuestionsConfig
+    BUILDER_CONFIGS = []
+    BUILDER_CONFIGS += [
+        WikidataSimpleQuestionsConfig(
+            name=f"main_{ln}",
+            version=VERSION,
+            description="main version of wikidata simplequestions",
+        )
+        for ln in _LANGS
+    ]
+    BUILDER_CONFIGS += [
+        WikidataSimpleQuestionsConfig(
+            name=f"answerable_{ln}",
+            version=VERSION,
+            description="answerable version of wikidata simplequestions",
+        )
+        for ln in _LANGS
+    ]
+
+    DEFAULT_CONFIG_NAME = "answerable_en"
+
+    def _info(self):
+        features = datasets.Features(
+            {
+                "subject": datasets.Value("string"),
+                "property": datasets.Value("string"),
+                "object": datasets.Value("string"),
+                "question": datasets.Value("string"),
+            }
+        )
+
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=features,
+            homepage=_HOMEPAGE,
+            license=_LICENSE,
+        )
+
+    def _split_generators(self, dl_manager):
+        if self.config.name == "default":
+            version, lang = "main", "en"
+        else:
+            version, lang = self.config.name.split("_")
+
+        if version == "main":
+            version = ""
+        else:
+            version = "_" + version
+
+        data_dir = os.path.join(self.base_path, _DATA_DIRECTORY)
+
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "filepath": os.path.join(data_dir, f"annotated_wd_data_train{version}_{lang}.txt"),
+                },
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.VALIDATION,
+                gen_kwargs={
+                    "filepath": os.path.join(data_dir, f"annotated_wd_data_valid{version}_{lang}.txt"),
+                },
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.TEST,
+                gen_kwargs={
+                    "filepath": os.path.join(data_dir, f"annotated_wd_data_test{version}_{lang}.txt"),
+                },
+            ),
+        ]
+
+    def _generate_examples(self, filepath, vocab_path):
+        with open(filepath, encoding="utf-8") as f:
+            for key, row in enumerate(f):
+                data = row.split("\t")
+                yield (
+                    key,
+                    {
+                        "subject": data[0],
+                        "property": data[1],
+                        "object": data[2],
+                        "question": data[3],
+                    },
+                )