fix: Normalize licenses including casing, uses of "-" etc.

docs/adding_a_dataset.md

@@ -104,7 +104,7 @@ class VGClustering(AbsTaskClustering):
         form="Written",
         domains=["Academic", "Non-fiction"],
         task_subtypes=["Scientific Reranking"],
-        license="cc-by-nc",
+        license="cc-by-nc-4.0",
         annotations_creators="derived",
         dialect=[],
         text_creation="found",

mteb/abstasks/TaskMetadata.py

@@ -60,7 +60,6 @@
     "Web",
     "Written",
     "Programming",
-    None,
 ]
 
 SAMPLE_CREATION_METHOD = Literal[
@@ -135,6 +134,32 @@
     "sql",
 ]
 
+LICENSES = (  # this list can be extended as needed
+    Literal[  # we use lowercase for the licenses similar to the huggingface datasets
+        "not specified",  # or none found
+        "mit",
+        "cc-by-2.0",
+        "cc-by-3.0",
+        "cc-by-4.0",
+        "cc-by-sa-3.0",
+        "cc-by-sa-4.0",
+        "cc-by-nc-4.0",
+        "cc-by-nc-sa-3.0",
+        "cc-by-nc-sa-4.0",
+        "cc-by-nc-nd-4.0",
+        "openrail",
+        "openrail++",
+        "odc-by",
+        "afl-3.0",
+        "apache-2.0",
+        "cc-by-nd-2.1-jp",
+        "cc0-1.0",
+        "bsd-3-clause",
+        "gpl-3.0",
+        "cdla-sharing-1.0",
+        "mpl-2.0",
+    ]
+)
 
 METRIC_NAME = str
 METRIC_VALUE = Union[int, float, Dict[str, Any]]
@@ -163,7 +188,7 @@ class TaskMetadata(BaseModel):
         domains: The domains of the data. These includes "Non-fiction", "Social", "Fiction", "News", "Academic", "Blog", "Encyclopaedic",
             "Government", "Legal", "Medical", "Poetry", "Religious", "Reviews", "Web", "Spoken", "Written". A dataset can belong to multiple domains.
         task_subtypes: The subtypes of the task. E.g. includes "Sentiment/Hate speech", "Thematic Clustering". Feel free to update the list as needed.
-        license: The license of the data.
+        license: The license of the data specified as lowercase, e.g. "cc-by-nc-4.0". If the license is not specified, use "not specified". For custom licenses a URL is used.
         socioeconomic_status: The socioeconomic status of the data. Includes "high", "medium", "low", "mixed".
         annotations_creators: The type of the annotators. Includes "expert-annotated" (annotated by experts), "human-annotated" (annotated e.g. by
             mturkers), "derived" (derived from structure in the data).
@@ -193,7 +218,7 @@ class TaskMetadata(BaseModel):
     date: tuple[STR_DATE, STR_DATE] | None = None
     domains: list[TASK_DOMAIN] | None = None
     task_subtypes: list[TASK_SUBTYPE] | None = None
-    license: str | None = None
+    license: LICENSES | STR_URL | None = None
 
     annotations_creators: ANNOTATOR_TYPE | None = None
     dialect: list[str] | None = None

mteb/tasks/BitextMining/dan/BornholmskBitextMining.py

@@ -22,7 +22,7 @@ class BornholmBitextMining(AbsTaskBitextMining):
         main_score="f1",
         date=("2019-01-01", "2019-12-31"),
         domains=["Web", "Social", "Fiction", "Written"],
-        license="CC-BY-4.0",
+        license="cc-by-4.0",
         task_subtypes=["Dialect pairing"],
         annotations_creators="expert-annotated",
         dialect=["da-dan-bornholm"],

mteb/tasks/BitextMining/kat/TbilisiCityHallBitextMining.py

@@ -37,7 +37,7 @@ class TbilisiCityHallBitextMining(AbsTaskBitextMining, MultilingualTask):
         reference="https://huggingface.co/datasets/jupyterjazz/tbilisi-city-hall-titles",
         date=("2024-05-02", "2024-05-03"),
         task_subtypes=[],
-        license="Not specified",
+        license="not specified",
         annotations_creators="derived",
         dialect=[],
         bibtex_citation="",

mteb/tasks/BitextMining/multilingual/BUCCBitextMining.py

@@ -34,7 +34,7 @@ class BUCCBitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("2017-01-01", "2018-12-31"),
         domains=["Written"],
         task_subtypes=[],
-        license="Unknown",
+        license="not specified",
         annotations_creators="human-annotated",
         dialect=[],
         sample_creation="human-translated",

mteb/tasks/BitextMining/multilingual/BUCCBitextMiningFast.py

@@ -34,7 +34,7 @@ class BUCCBitextMiningFast(AbsTaskBitextMining, MultilingualTask):
         date=("2017-01-01", "2018-12-31"),
         domains=["Written"],
         task_subtypes=[],
-        license="Unknown",
+        license="not specified",
         annotations_creators="human-annotated",
         dialect=[],
         sample_creation="human-translated",

mteb/tasks/BitextMining/multilingual/BibleNLPBitextMining.py

@@ -880,7 +880,7 @@ class BibleNLPBitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("1997-01-01", "2020-12-31"),
         domains=["Religious", "Written"],
         task_subtypes=[],
-        license="CC-BY-SA-4.0",
+        license="cc-by-sa-4.0",
         annotations_creators="expert-annotated",
         dialect=[],
         sample_creation="created",

mteb/tasks/BitextMining/multilingual/DiaBLaBitextMining.py

@@ -29,7 +29,7 @@ class DiaBLaBitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("2016-01-01", "2017-12-31"),
         domains=["Social", "Written"],
         task_subtypes=[],
-        license="CC BY-NC-SA 4.0",
+        license="cc-by-nc-sa-4.0",
         annotations_creators="human-annotated",
         dialect=[],
         sample_creation="created",

mteb/tasks/BitextMining/multilingual/FloresBitextMining.py

@@ -255,7 +255,7 @@ class FloresBitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("2022-01-01", "2022-12-31"),
         domains=["Non-fiction", "Encyclopaedic", "Written"],
         task_subtypes=[],
-        license="CC BY-SA 4.0",
+        license="cc-by-sa-4.0",
         annotations_creators="human-annotated",
         dialect=[],
         sample_creation="created",

mteb/tasks/BitextMining/multilingual/IN22ConvBitextMining.py

@@ -87,7 +87,7 @@ class IN22ConvBitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("2022-10-01", "2023-03-01"),
         domains=["Social", "Spoken", "Fiction", "Spoken"],
         task_subtypes=[],
-        license="CC-BY-4.0",
+        license="cc-by-4.0",
         annotations_creators="expert-annotated",
         dialect=[],
         sample_creation="created",

mteb/tasks/BitextMining/multilingual/IN22GenBitextMining.py

@@ -81,7 +81,7 @@ class IN22GenBitextMining(AbsTaskBitextMining, MultilingualTask):
             "Written",
         ],
         task_subtypes=[],
-        license="CC-BY-4.0",
+        license="cc-by-4.0",
         annotations_creators="expert-annotated",
         dialect=[],
         sample_creation="created",

mteb/tasks/BitextMining/multilingual/IWSLT2017BitextMining.py

@@ -55,7 +55,7 @@ class IWSLT2017BitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("2007-01-01", "2017-12-14"),  # rough estimate
         domains=["Non-fiction", "Fiction", "Written"],
         task_subtypes=[],
-        license="CC-BY-NC-ND-4.0",
+        license="cc-by-nc-nd-4.0",
         annotations_creators="expert-annotated",
         dialect=[],
         sample_creation="found",

mteb/tasks/BitextMining/multilingual/IndicGenBenchFloresBitextMining.py

@@ -116,7 +116,7 @@ class IndicGenBenchFloresBitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("2023-10-01", "2024-05-01"),
         domains=["Web", "News", "Written"],
         task_subtypes=[],
-        license="CC-BY-SA-4.0",
+        license="cc-by-sa-4.0",
         annotations_creators="expert-annotated",
         dialect=[],
         sample_creation="human-translated and localized",

mteb/tasks/BitextMining/multilingual/LinceMTBitextMining.py

@@ -27,7 +27,7 @@ class LinceMTBitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("2019-01-01", "2020-01-01"),
         domains=["Social", "Written"],
         task_subtypes=[],
-        license="Unknown",
+        license="not specified",
         annotations_creators="human-annotated",
         dialect=[],
         sample_creation="found",

mteb/tasks/BitextMining/multilingual/NTREXBitextMining.py

@@ -266,7 +266,7 @@ class NTREXBitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("2019-08-01", "2022-11-01"),
         domains=["News", "Written"],
         task_subtypes=[],
-        license="CC-BY-SA-4.0",
+        license="cc-by-sa-4.0",
         annotations_creators="expert-annotated",
         dialect=[],
         sample_creation="human-translated and localized",

mteb/tasks/BitextMining/multilingual/NollySentiBitextMining.py

@@ -30,7 +30,7 @@ class NollySentiBitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("2022-01-01", "2023-01-01"),
         domains=["Social", "Reviews", "Written"],
         task_subtypes=[],
-        license="CC BY-SA 4.0",
+        license="cc-by-sa-4.0",
         annotations_creators="human-annotated",
         dialect=[],
         sample_creation="found",

mteb/tasks/BitextMining/multilingual/NorwegianCourtsBitextMining.py

@@ -22,7 +22,7 @@ class NorwegianCourtsBitextMining(AbsTaskBitextMining):
         date=("2020-01-01", "2020-12-31"),
         domains=["Legal", "Written"],
         task_subtypes=[],
-        license="CC BY 4.0",
+        license="cc-by-4.0",
         annotations_creators="human-annotated",
         dialect=[],
         sample_creation="found",

mteb/tasks/BitextMining/multilingual/NusaTranslationBitextMining.py

@@ -37,7 +37,7 @@ class NusaTranslationBitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("2021-08-01", "2022-07-01"),
         domains=["Social", "Written"],
         task_subtypes=[],
-        license="CC BY-SA 4.0",
+        license="cc-by-sa-4.0",
         annotations_creators="human-annotated",
         dialect=[],
         sample_creation="created",

mteb/tasks/BitextMining/multilingual/NusaXBitextMining.py

@@ -37,7 +37,7 @@ class NusaXBitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("2021-08-01", "2022-07-01"),
         domains=["Reviews", "Written"],
         task_subtypes=[],
-        license="CC BY-SA 4.0",
+        license="cc-by-sa-4.0",
         annotations_creators="human-annotated",
         dialect=[],
         sample_creation="created",

mteb/tasks/BitextMining/multilingual/PhincBitextMining.py

@@ -27,7 +27,7 @@ class PhincBitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("2019-01-01", "2020-01-01"),
         domains=["Social", "Written"],
         task_subtypes=[],
-        license="CC BY 4.0",
+        license="cc-by-4.0",
         annotations_creators="human-annotated",
         dialect=[],
         sample_creation="found",

mteb/tasks/BitextMining/multilingual/RomaTalesBitextMining.py

@@ -28,7 +28,7 @@ class RomaTalesBitextMining(AbsTaskBitextMining, MultilingualTask):
         ),  # Broad historical range for the creation of folk tales
         domains=["Fiction", "Written"],
         task_subtypes=[],
-        license="Not specified",
+        license="not specified",
         annotations_creators="expert-annotated",
         dialect=["Lovari"],
         sample_creation="created",

mteb/tasks/BitextMining/multilingual/TatoebaBitextMining.py

@@ -141,7 +141,7 @@ class TatoebaBitextMining(AbsTaskBitextMining, MultilingualTask):
             "Written"
         ],  # Tatoeba corpus includes a wide range of topics and domains
         task_subtypes=[],
-        license="CC BY 2.0",
+        license="cc-by-2.0",
         annotations_creators="human-annotated",
         dialect=[],  # No specific dialect mentioned
         sample_creation="found",

mteb/tasks/BitextMining/srn/SRNCorpusBitextMining.py

@@ -42,7 +42,7 @@ class SRNCorpusBitextMining(AbsTaskBitextMining, MultilingualTask):
         date=("2022-04-01", "2022-07-31"),
         domains=["Social", "Web", "Written"],
         task_subtypes=[],
-        license="CC-BY-SA-4.0",
+        license="cc-by-sa-4.0",
         annotations_creators="human-annotated",
         dialect=[],
         sample_creation="found",

mteb/tasks/BitextMining/vie/VieMedEVBitextMining.py

@@ -29,7 +29,7 @@ class VieMedEVBitextMining(AbsTaskBitextMining):
         date=("2024-08-28", "2022-03-28"),
         domains=["Medical", "Written"],
         task_subtypes=[],
-        license="cc-by-nc",
+        license="cc-by-nc-4.0",  # version is assumed, but was previously unspecified
         annotations_creators="expert-annotated",
         dialect=[],
         sample_creation="human-translated and localized",

mteb/tasks/Classification/ara/AJGT.py

@@ -22,7 +22,7 @@ class AJGT(AbsTaskClassification):
         date=("2021-01-01", "2022-01-25"),
         domains=["Social", "Written"],
         task_subtypes=["Sentiment/Hate speech"],
-        license="AFL",
+        license="afl-3.0",
         annotations_creators="human-annotated",
         dialect=["ara-arab-MSA", "ara-arab-JO"],
         sample_creation="found",

mteb/tasks/Classification/ara/HotelReviewSentimentClassification.py

@@ -25,7 +25,7 @@ class HotelReviewSentimentClassification(AbsTaskClassification):
         date=("2016-06-01", "2016-07-31"),
         domains=["Reviews", "Written"],
         task_subtypes=["Sentiment/Hate speech"],
-        license="Not specified",
+        license="not specified",
         annotations_creators="derived",
         dialect=["ara-arab-EG", "ara-arab-JO", "ara-arab-LB", "ara-arab-SA"],
         sample_creation="found",

mteb/tasks/Classification/ara/OnlineStoreReviewSentimentClassification.py

@@ -24,7 +24,7 @@ class OnlineStoreReviewSentimentClassification(AbsTaskClassification):
         date=("2024-05-01", "2024-05-15"),
         domains=["Reviews", "Written"],
         task_subtypes=["Sentiment/Hate speech"],
-        license="Not specified",
+        license="not specified",
         annotations_creators="derived",
         dialect=["ara-Arab-SA"],
         sample_creation="found",

mteb/tasks/Classification/ara/RestaurantReviewSentimentClassification.py

@@ -24,7 +24,7 @@ class RestaurantReviewSentimentClassification(AbsTaskClassification):
         date=("2014-01-01", "2015-01-01"),
         domains=["Reviews", "Written"],
         task_subtypes=["Sentiment/Hate speech"],
-        license="None specified",
+        license="not specified",
         annotations_creators="derived",
         dialect=["ara-arab-EG", "ara-arab-JO", "ara-arab-SA"],
         sample_creation="found",

mteb/tasks/Classification/ara/TweetEmotionClassification.py

@@ -25,7 +25,7 @@ class TweetEmotionClassification(AbsTaskClassification):
         date=("2014-01-01", "2016-08-31"),
         domains=["Social", "Written"],
         task_subtypes=["Sentiment/Hate speech"],
-        license="Not specified",
+        license="not specified",
         annotations_creators="human-annotated",
         dialect=["ara-arab-EG", "ara-arab-LB", "ara-arab-JO", "ara-arab-SA"],
         sample_creation="found",

mteb/tasks/Classification/ara/TweetSarcasmClassification.py

@@ -22,7 +22,7 @@ class TweetSarcasmClassification(AbsTaskClassification):
         date=("2020-01-01", "2021-01-01"),
         domains=["Social", "Written"],
         task_subtypes=["Sentiment/Hate speech"],
-        license="MIT",
+        license="mit",
         annotations_creators="human-annotated",
         dialect=["ara-arab-EG", "ara-arab-LB", "ara-arab-MA", "ara-arab-SA"],
         sample_creation="found",

mteb/tasks/Classification/ben/BengaliDocumentClassification.py

@@ -23,7 +23,7 @@ class BengaliDocumentClassification(AbsTaskClassification):
         dialect=[],
         domains=["News", "Written"],
         task_subtypes=[],
-        license="CC BY-NC-SA 4.0",
+        license="cc-by-nc-sa-4.0",
         annotations_creators="derived",
         sample_creation="found",
         bibtex_citation="""

mteb/tasks/Classification/ben/BengaliHateSpeechClassification.py

@@ -23,7 +23,7 @@ class BengaliHateSpeechClassification(AbsTaskClassification):
         dialect=[],
         domains=["News", "Written"],
         task_subtypes=["Sentiment/Hate speech"],
-        license="MIT",
+        license="mit",
         annotations_creators="expert-annotated",
         sample_creation="found",
         bibtex_citation="""@inproceedings{karim2020BengaliNLP,

mteb/tasks/Classification/ben/BengaliSentimentAnalysis.py

@@ -23,7 +23,7 @@ class BengaliSentimentAnalysis(AbsTaskClassification):
         dialect=[],
         domains=["Reviews", "Written"],
         task_subtypes=["Sentiment/Hate speech"],
-        license="CC BY 4.0",
+        license="cc-by-4.0",
         annotations_creators="human-annotated",
         sample_creation="found",
         bibtex_citation="""@inproceedings{sazzed2020cross,

mteb/tasks/Classification/ces/CSFDCZMovieReviewSentimentClassification.py

@@ -24,7 +24,7 @@ class CSFDCZMovieReviewSentimentClassification(AbsTaskClassification):
         main_score="accuracy",
         domains=["Reviews", "Written"],
         task_subtypes=["Sentiment/Hate speech"],
-        license="CC-BY-SA-4.0",
+        license="cc-by-sa-4.0",
         annotations_creators="derived",
         dialect=[],
         sample_creation="found",

mteb/tasks/Classification/ces/CzechProductReviewSentimentClassification.py

@@ -23,7 +23,7 @@ class CzechProductReviewSentimentClassification(AbsTaskClassification):
         dialect=[],
         domains=["Reviews", "Written"],
         task_subtypes=["Sentiment/Hate speech"],
-        license="CC BY-NC-SA 4.0",
+        license="cc-by-nc-sa-4.0",
         annotations_creators="derived",
         sample_creation="found",
         bibtex_citation="""

mteb/tasks/Classification/ces/CzechSoMeSentimentClassification.py

@@ -23,7 +23,7 @@ class CzechSoMeSentimentClassification(AbsTaskClassification):
         dialect=[],
         domains=["Reviews", "Written"],
         task_subtypes=["Sentiment/Hate speech"],
-        license="CC BY-NC-SA 4.0",
+        license="cc-by-nc-sa-4.0",
         annotations_creators="derived",
         sample_creation="found",
         bibtex_citation="""

mteb/tasks/Classification/ces/CzechSubjectivityClassification.py

@@ -22,7 +22,7 @@ class CzechSubjectivityClassification(AbsTaskClassification):
         main_score="accuracy",
         domains=["Reviews", "Written"],
         task_subtypes=["Sentiment/Hate speech"],
-        license="Not specified",
+        license="not specified",
         annotations_creators="human-annotated",
         dialect=[],
         sample_creation="found",

mteb/tasks/Classification/dan/AngryTweetsClassification.py

@@ -22,7 +22,7 @@ class AngryTweetsClassification(AbsTaskClassification):
         date=("2021-01-01", "2021-12-31"),
         domains=["Social", "Written"],
         task_subtypes=["Sentiment/Hate speech"],
-        license="CC-BY-4.0",
+        license="cc-by-4.0",
         annotations_creators="human-annotated",
         dialect=[],
         sample_creation="found",