Add dataset

.gitattributes

@@ -35,3 +35,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.mp3 filter=lfs diff=lfs merge=lfs -text
 *.ogg filter=lfs diff=lfs merge=lfs -text
 *.wav filter=lfs diff=lfs merge=lfs -text
+train.csv filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,2 @@
+/.env
+.DS_Store

README.md

@@ -0,0 +1,3 @@
+# BWNS: The Baha'i World News Service dataset.
+
+BWNS articles from 2000 to 2022.

bwns.py

@@ -0,0 +1,98 @@
+import csv
+import datasets
+
+_DESCRIPTION = "BWNS: The Bahai World News Service dataset."
+_HOMEPAGE = "https://news.bahai.org/"
+_URLS = {
+    "train": "./train.csv",
+}
+
+
+class BwnsDataset(datasets.GeneratorBasedBuilder):
+    """BWNS: The Bahai World News Service dataset."""
+
+    BUILDER_CONFIG_CLASS=datasets.BuilderConfig
+
+    VERSION = datasets.Version("1.1.0")
+    NAME = "BWNS"
+
+    BUILDER_CONFIGS = [
+        datasets.BuilderConfig(name="original", version=VERSION, description="Original"),
+        datasets.BuilderConfig(name="concat", version=VERSION, description="Title and content concatenated"),
+    ]
+
+    DEFAULT_CONFIG_NAME = "original"
+
+    def _info(self):
+        # TODO: This method specifies the datasets.DatasetInfo object which contains informations and typings for the dataset
+        if self.config.name == "original":
+            features = datasets.Features(
+                {
+                    "id": datasets.Value("int32"),
+                    "date": datasets.Value("timestamp[s]"),
+                    "city": datasets.Value("string"),
+                    "region": datasets.Value("string"),
+                    "country": datasets.Value("string"),
+                    "related_articles": datasets.Sequence(datasets.Value("int32")),
+                    "title": datasets.Value("string"),
+                    "content": datasets.Value("string"),
+                }
+            )
+        elif self.config.name == "concat":
+            features = datasets.Features(
+                {
+                    "id": datasets.Value("int32"),
+                    "date": datasets.Value("timestamp[s]"),
+                    "city": datasets.Value("string"),
+                    "region": datasets.Value("string"),
+                    "country": datasets.Value("string"),
+                    "related_articles": datasets.Sequence(datasets.Value("int32")),
+                    "title_and_content": datasets.Value("string"),
+                }
+            )
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=features,
+            supervised_keys=None,
+            homepage=_HOMEPAGE,
+        )
+
+    def _split_generators(self, dl_manager):
+        data_dirs = dl_manager.download_and_extract(_URLS)
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                # These kwargs will be passed to _generate_examples
+                gen_kwargs={
+                    "filepath": data_dirs['train'],
+                    "split": "train",
+                },
+            ),
+        ]
+
+    # method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
+    def _generate_examples(self, filepath, split):
+        with open(filepath, encoding="utf-8") as f:
+            reader = csv.DictReader(f)
+            for row in reader:
+                if self.config.name == 'original':
+                    yield row['id'], {
+                        "id": int(row['id']),
+                        "date": row['date'],
+                        "city": row['city'],
+                        "region": row['region'],
+                        "country": row['country'],
+                        "related_articles": [int(x) for x in row['related_articles'][1:-1].split(',') if x],
+                        "title": row['title'],
+                        "content": row['content'],
+                    }
+                elif self.config.name == 'concat':
+                    yield row['id'], {
+                        "id": int(row['id']),
+                        "date": row['date'],
+                        "city": row['city'],
+                        "region": row['region'],
+                        "country": row['country'],
+                        "related_articles": [int(x) for x in row['related_articles'][1:-1].split(',') if x],
+                        "title_and_content": f"{row['title']} {row['content']}",
+                    }

dataset_infos.json

@@ -0,0 +1 @@
+{"original": {"description": "BWNS: The Bahai World News Service dataset.", "citation": "", "homepage": "https://news.bahai.org/", "license": "", "features": {"id": {"dtype": "int32", "id": null, "_type": "Value"}, "date": {"dtype": "timestamp[s]", "id": null, "_type": "Value"}, "city": {"dtype": "string", "id": null, "_type": "Value"}, "region": {"dtype": "string", "id": null, "_type": "Value"}, "country": {"dtype": "string", "id": null, "_type": "Value"}, "related_articles": {"feature": {"dtype": "int32", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "title": {"dtype": "string", "id": null, "_type": "Value"}, "content": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "bwns_dataset", "config_name": "original", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 5642343, "num_examples": 1387, "dataset_name": "bwns_dataset"}}, "download_checksums": {"./train.csv": {"num_bytes": 5648505, "checksum": "59dc4afe0f07ec8a3815de47451ae47b38ac10ea5037166dcb8a125ca1df4f28"}}, "download_size": 5648505, "post_processing_size": null, "dataset_size": 5642343, "size_in_bytes": 11290848}, "concat": {"description": "BWNS: The Bahai World News Service dataset.", "citation": "", "homepage": "https://news.bahai.org/", "license": "", "features": {"id": {"dtype": "int32", "id": null, "_type": "Value"}, "date": {"dtype": "timestamp[s]", "id": null, "_type": "Value"}, "city": {"dtype": "string", "id": null, "_type": "Value"}, "region": {"dtype": "string", "id": null, "_type": "Value"}, "country": {"dtype": "string", "id": null, "_type": "Value"}, "related_articles": {"feature": {"dtype": "int32", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "title_and_content": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "bwns_dataset", "config_name": "concat", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 5638182, "num_examples": 1387, "dataset_name": "bwns_dataset"}}, "download_checksums": {"./train.csv": {"num_bytes": 5648505, "checksum": "59dc4afe0f07ec8a3815de47451ae47b38ac10ea5037166dcb8a125ca1df4f28"}}, "download_size": 5648505, "post_processing_size": null, "dataset_size": 5638182, "size_in_bytes": 11286687}}

train.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59dc4afe0f07ec8a3815de47451ae47b38ac10ea5037166dcb8a125ca1df4f28
+size 5648505