Datasets:

Modalities:
Tabular
Text
Formats:
parquet
Languages:
Russian
ArXiv:
Libraries:
Datasets
pandas
License:
evtaktasheva commited on
Commit
1b693c8
1 Parent(s): 682f6fe

updated the dataset card

Browse files
Files changed (2) hide show
  1. README.md +324 -0
  2. rublimp.jpg +3 -0
README.md CHANGED
@@ -1,5 +1,14 @@
1
  ---
2
  license: apache-2.0
 
 
 
 
 
 
 
 
 
3
  dataset_info:
4
  - config_name: add_new_suffix
5
  features:
@@ -1533,3 +1542,318 @@ configs:
1533
  - split: train
1534
  path: verb_ins_object/train-*
1535
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
+ language:
4
+ - ru
5
+ tags:
6
+ - benchmark
7
+ task_ids:
8
+ - acceptability-classification
9
+ pretty_name: RuBLiMP
10
+ size_categories:
11
+ - 10K<n<100K
12
  dataset_info:
13
  - config_name: add_new_suffix
14
  features:
 
1542
  - split: train
1543
  path: verb_ins_object/train-*
1544
  ---
1545
+
1546
+ # RuBLiMP
1547
+
1548
+
1549
+ ## Dataset Description
1550
+
1551
+ **RuBLiMP**, or **Ru**ssian **B**enchmark of **Li**nguistic **M**inimal **P**airs, is the first diverse and large-scale benchmark of minimal pairs in Russian.
1552
+
1553
+ RuBLiMP includes **45k** minimal pairs of sentences that differ in grammaticality and isolate morphological, syntactic, or semantic phenomena. In contrast to existing benchmarks of linguistic minimal pairs, RuBLiMP is created by applying linguistic perturbations to automatically annotated sentences from open text corpora and decontaminating test data, making the minimal pairs more diverse and natural.
1554
+
1555
+
1556
+ ## Dataset Creation
1557
+
1558
+ ![rublimp](rublimp.jpg)
1559
+
1560
+ Example: *Vpervye kosmonavt spal v nevesomosti* "For the first time an astronaut slept in zero gravity".
1561
+
1562
+ The RuBLiMP's creation approach comprises four stages:
1563
+
1564
+ - **(a)** Extract sentences from publicly available corpora of Wikipedia texts, news articles, and books.
1565
+ - **(b)** Annotate each extracted sentence in the Universal Dependencies scheme with a multidomain morphosyntactic parser for Russian.
1566
+ - **(c)** Search the dependency trees for specific lexical units and linguistic structures and apply expert-written perturbation rules to create a pool of minimal pairs for a target paradigm.
1567
+ - **(d)** Compute [Min-K\% Prob](https://swj0419.github.io/detect-pretrain.github.io/) for each grammatical sentence in the pool using a set of LMs. Select *t* (the threshold for the maximum Min-K\% Prob value), which allows to find an intersection of 1k minimal pairs between the LMs. The minimal pairs in the intersection contain grammatical sentences that are not detected as the LMs' pretraining examples.
1568
+
1569
+
1570
+ ## Dataset Structure
1571
+
1572
+ ### Data Instances
1573
+
1574
+ Each instance in the dataset is pair of sentences, where a certain linguistic phenomenon is isolated:
1575
+
1576
+ ```
1577
+ {
1578
+ 'id': 240907,
1579
+ 'source_sentence': 'Это еще вчера все поняли.',
1580
+ 'target_sentence': 'Это еще завтра все поняли.',
1581
+ 'source_word': 'вчера',
1582
+ 'target_word': 'завтра',
1583
+ 'level': 'Semantics',
1584
+ 'phenomenon': 'Tense',
1585
+ 'PID': 'tense_marker',
1586
+ 'subtype': 'tense_marker_simple',
1587
+ 'domain': 'librusec',
1588
+ 'tree_depth': 2
1589
+ }
1590
+ ```
1591
+
1592
+ An example in English for illustration purposes:
1593
+
1594
+ ```
1595
+ {
1596
+ 'id': 240907,
1597
+ 'source_sentence': 'Everyone understood this yesterday already.',
1598
+ 'target_sentence': 'Everyone understood this tomorrow already.',
1599
+ 'source_word': 'yesterday',
1600
+ 'target_word': 'tomorrow',
1601
+ 'level': 'Semantics',
1602
+ 'phenomenon': 'Tense',
1603
+ 'PID': 'tense_marker',
1604
+ 'subtype': 'tense_marker_simple',
1605
+ 'domain': 'librusec',
1606
+ 'tree_depth': 2
1607
+ }
1608
+ ```
1609
+
1610
+ ### Data Fields
1611
+ `id`: minimal pair identificator from the corpus \
1612
+ `source_sentence`: original grammatical sentence \
1613
+ `target_sentence`: perturbed ungrammatical sentence \
1614
+ `source_word`: token in the original, undergoing the perturbation \
1615
+ `target_word`: token in the ungrammatical sentence after the perturbation \
1616
+ `level`: language level, targeted by the minimal pair (Syntax, Semantics or Morphology) \
1617
+ `phenomenon`: targeted phenomenon \
1618
+ `PID`: name of the paradigm \
1619
+ `subtype`: type of the paradigm, if applicable \
1620
+ `domain`: sentence source (librusec, wikipedia or wikinews) \
1621
+ `tree_depth`: depth, of the sentence's syntax tree
1622
+
1623
+
1624
+ ## Phenomena
1625
+
1626
+ ### Morphology
1627
+
1628
+ <details>
1629
+ <summary><b>Word Formation</b></summary>
1630
+
1631
+ - **Addition of Extra Morphemes: Uninterpretable Suffix Combinations** (`add_new_suffix`) \
1632
+ Adding a new suffix to the noun or adjective to create a non-existing word
1633
+
1634
+ - **Addition of Extra Morphemes: Verb Prefixes** (`add_verb_prefix`) \
1635
+ Adding a prefix to a verb to create a violation of prefix stacking rules.
1636
+
1637
+ - **Morpheme Permutation: Verb Prefixes** (`change_verb_prefixes_order`) \
1638
+ Changing the order of the verb's prefixes to create a violation of prefix stacking rules.
1639
+
1640
+ </details>
1641
+
1642
+ <details>
1643
+ <summary><b>Word Inflection</b></summary>
1644
+
1645
+ - **Replacement of Inflectional Affixes: Noun Declensions (Simple)** (`change_declension_ending`) \
1646
+ Changing the inflectional suffixes of a noun to the suffixes of another declension
1647
+
1648
+ - **Replacement of Inflectional Affixes: Declensions of Nouns With Agreeing Dependents** (`change_declension_ending_has_dep`) \
1649
+ Changing the inflectional suffixes of a noun to the suffixes of another declension in the presence of an agreeing noun modifier
1650
+
1651
+ - **Inflectional Affixes: Verbal Conjugation Swap** (`change_verb_conjugation`) \
1652
+ Replacing the verb’s inflection with inflection of the opposite conjugation
1653
+
1654
+
1655
+ </details>
1656
+
1657
+
1658
+ ### Syntax
1659
+ <details>
1660
+ <summary><b>Government</b></summary>
1661
+
1662
+ - **Prepositional Government** (`adp_government_case`) \
1663
+ Changing the case of a noun, governed by a preposition
1664
+
1665
+ - **Verbal Government: Direct Object** (`verb_acc_object`) \
1666
+ Changing the case of a direct verb object
1667
+
1668
+ - **Verbal Government: Genitive Object** (`verb_gen_object`) \
1669
+ Changing the case of an indirect verb object in Genitive case
1670
+
1671
+ - **Verbal Government: Object in Instrumental Case** (`verb_ins_object`) \
1672
+ Changing the case of an indirect verb object in Instrumental case
1673
+
1674
+ - **Verbal Government: Nominalizations** (`nominalization_cas`) \
1675
+ Changing the case of a dependent of a nominalization
1676
+
1677
+ </details>
1678
+
1679
+
1680
+ <details>
1681
+ <summary><b>Subject-Predicate Agreement</b></summary>
1682
+
1683
+ - **Subject-Predicate Agreement (Number)** (`noun_subj_predicate_agreement_number`) \
1684
+ Changing the number of the predicate to be distinct from its subject's (or, sometimes, changing number of the subject to be distinct from its predicate's)
1685
+
1686
+ - **Genitive Subject-Predicate Agreement (Number)** (`genitive_subj_predicate_agreement_number`) \
1687
+ Changing the number of the predicate to plural, when subject is genitive and the agreement must be the default singular neuter
1688
+
1689
+ - **Clausal Subject-Predicate Agreement (Number)** (`clause_subj_predicate_agreement_number`) \
1690
+ Changing the number of the predicate to plural, when subject is a clause and the agreement must be the default singular neuter
1691
+
1692
+ - **Subject-Predicate Agreement in Presence of an Attractor (Number)** (`subj_predicate_agreement_number_attractor`) \
1693
+ Changing the number of the verb to that, which is different from the subject, but the same as subject's dependent, or the attractor
1694
+
1695
+ - **Subject-Predicate Agreement (Gender)** (`noun_subj_predicate_agreement_gender`) \
1696
+ Changing the gender of the predicate to be distinct from its subject's (or, sometimes, changing number of the subject to be distinct from its predicate's)
1697
+
1698
+ - **Genitive Subject-Predicate Agreement (Gender)** (`genitive_subj_predicate_agreement_gender`)
1699
+ Changing the gender of the predicate to feminine or masculine, when subject is genitive and the agreement must be the default singular neuter
1700
+
1701
+ - **Clausal Subject-Predicate Agreement (Gender)** (`clause_subj_predicate_agreement_gender`) \
1702
+ Changing the gender of the predicate to feminine or masculine, when subject is a clause and the agreement must be the default singular neuter
1703
+
1704
+ - **Subject-Predicate Agreement in Presence of an Attractor (Gender)** (`subj_predicate_agreement_gender_attractor`) \
1705
+ Changing the gender of the verb to that, which is different from the subject, but the same as subject's dependent, or the attractor
1706
+
1707
+ - **Subject-Predicate Agreement (Person)** (`noun_subj_predicate_agreement_person`) \
1708
+ Changing the person of the predicate to be distinct from its subject's
1709
+
1710
+ - **Genitive Subject-Predicate Agreement (Person)** (`genitive_subj_predicate_agreement_person`) \
1711
+ Changing the person of the predicate to first or second person, when subject is genitive and the agreement must be the default third person singular
1712
+
1713
+ - **Clausal Subject-Predicate Agreement (Person)** (`clause_subj_predicate_agreement_person`) \
1714
+ Changing the person of the predicate to first or second person, when subject is a clause and the agreement must be the default third person singular
1715
+
1716
+ </details>
1717
+
1718
+
1719
+ <details>
1720
+ <summary><b>Anaphor Agreement</b></summary>
1721
+
1722
+ - **Anaphor Agreement (Number)** (`anaphor_agreement_number`) \
1723
+ Changing the number of the relative pronoun or of its head noun
1724
+
1725
+ - **Anaphor Agreement (Gender)** (`anaphor_agreement_gender`) \
1726
+ Changing the gender of the relative pronoun
1727
+
1728
+ </details>
1729
+
1730
+ <details>
1731
+ <summary><b>Noun Phrase Agreement</b></summary>
1732
+
1733
+ - **Noun Phrase Agreement (Number)** (`np_agreement_number`) \
1734
+ Changing the number of an agreeing adjective
1735
+
1736
+ - **Noun Phrase Agreement (Gender)** (`np_agreement_gender`) \
1737
+ Changing the gender of an agreeing adjective
1738
+
1739
+ - **Noun Phrase Agreement (Case)** (`np_agreement_case`) \
1740
+ Changing the case of an agreeing adjective
1741
+
1742
+ </details>
1743
+
1744
+ <details>
1745
+ <summary><b>Floating Quantifier Agreement</b></summary>
1746
+
1747
+ - **Floating Quantifier Agreement (Number)** (`floating_quantifier_agreement_number`) \
1748
+ Changing the number of the quantifier or of the controller
1749
+
1750
+ - **Floating Quantifier Agreement (Gender)** (`floating_quantifier_agreement_gender`) \
1751
+ Changing the gender of the quantifier or of the controller
1752
+
1753
+ - **Floating Quantifier Agreement (Case)** (`floating_quantifier_agreement_case`) \
1754
+ Changing the case of the quantifier or of the controller
1755
+
1756
+ </details>
1757
+
1758
+
1759
+
1760
+ <details>
1761
+ <summary><b>Reflexives</b></summary>
1762
+
1763
+ - **External Possessor** (`external_possessor`) \
1764
+ Change a noun phrase or a pronoun to a reflexive pronoun sebya ‘self’ in a *u*-phrase inside the existential *be*-possessive construction.
1765
+
1766
+ </details>
1767
+
1768
+ <details>
1769
+ <summary><b>Negation</b></summary>
1770
+
1771
+ - **Negative Concord** (`negative_concord`) \
1772
+ Shifting the negative particle *ne* from a negated verb to another word in the sentence to violate negative concord rules.
1773
+
1774
+ - **Replacement of a Negative Pronoun with an Indefinite One** (`negative_pronoun_to_indefinite`) \
1775
+ Replacing an negative pronoun in the construction without a negated verb to an indefinite pronoun
1776
+
1777
+ - **Replacement of an Indefinite Pronoun with a Negative One** (`indefinite_pronoun_to_negative`) \
1778
+ Replacing an indefinite pronoun in the construction with a negated verb to a negative pronoun
1779
+
1780
+ </details>
1781
+
1782
+
1783
+ ### Semantics
1784
+
1785
+ <details>
1786
+ <summary><b>Argument Structure</b></summary>
1787
+
1788
+ - **Transitivity** (`transitive_verb`) \
1789
+ Replacing a transitive verb with an intransitive one
1790
+
1791
+ - **Animate Subject of a Transitive Verb** (`transitive_verb_subject`) \
1792
+ Swapping the subject and the direct object of a transitive verb or replacing the subject with a random inanimate word
1793
+
1794
+ - **Animate Subject of a Passive Verb** (`transitive_verb_passive`) \
1795
+ Swapping the subject and the direct object of a transitive verb in a passive construction or replacing the subject with a random inanimate word
1796
+
1797
+ - **Animate Direct Object of a Transitive Verb** (`transitive_verb_object`) \
1798
+ Replacing the direct object with a random inanimate word
1799
+
1800
+ - **Animate Indirect Object of a Transitive Verb** (`transitive_verb_iobject`) \
1801
+ Swapping the subject and the indirect object of a transitive verb or replacing the indirect subject of a transitive verb with a random inanimate word
1802
+
1803
+
1804
+ </details>
1805
+
1806
+
1807
+ <details>
1808
+ <summary><b>Aspect</b></summary>
1809
+
1810
+ - **Incompatibility of the Perfective with the Semantics of Duration** (`change_duration_aspect`) \
1811
+ Replacing an imperfective verb with a perfective one in contexts with semantics of duration
1812
+
1813
+ - **Impossibility of the Perfective in Repetitive Situations** (`change_repetition_aspect`) \
1814
+ Replacing an imperfective verb with a perfective one in contexts with semantics of repetition
1815
+
1816
+ - **Impossibility of the Perfective Under Negated Strong Deontic Verbs** (`deontic_imperative_aspect`) \
1817
+ Replacing an imperfective verb with a perfective one in contexts with a negated deontic verb
1818
+
1819
+ </details>
1820
+
1821
+ <details>
1822
+ <summary><b>Tense</b></summary>
1823
+
1824
+ - **Tense** (`single_verb_tense`) \
1825
+ Changing verb tense in the presence of a temporal adverbial
1826
+
1827
+ - **Tense (coordination)** (`conj_verb_tense`) \
1828
+ Changing the tense of a conjoined verb in the presence of a temporal adverbial
1829
+
1830
+ - **Tense Markers** (`tense_marker`) \
1831
+ Changing a temporal adverbial in a sentence with a tense-marked verb
1832
+
1833
+ </details>
1834
+
1835
+
1836
+ ## Additional Information
1837
+
1838
+ ### Dataset Curators
1839
+
1840
+ [Ekaterina Taktasheva](https://github.com/evtaktasheva) ([@evtaktasheva](https://huggingface.co/evtaktasheva)), [Maxim Bazhukov](https://github.com/bamaxi) ([@bamaxi](https://huggingface.co/bamaxi)), [Kirill Koncha](https://github.com/kirillkoncha) ([@midwestcyr](https://huggingface.co/midwestcyr)), [Ekaterina Artemova](https://github.com/artemovae) ([@KatyaA](https://huggingface.co/KatyaA)), [Vladislav Mikhailov](https://github.com/vmkhlv) ([@vmkhlv](https://huggingface.co/vmkhlv))
1841
+
1842
+
1843
+ ### Licensing Information
1844
+
1845
+ Apache 2.0
1846
+
1847
+ ### Citation Information
1848
+
1849
+ ```
1850
+ @misc{taktasheva2024rublimprussianbenchmarklinguistic,
1851
+ title={RuBLiMP: Russian Benchmark of Linguistic Minimal Pairs},
1852
+ author={Ekaterina Taktasheva and Maxim Bazhukov and Kirill Koncha and Alena Fenogenova and Ekaterina Artemova and Vladislav Mikhailov},
1853
+ year={2024},
1854
+ eprint={2406.19232},
1855
+ archivePrefix={arXiv},
1856
+ primaryClass={cs.CL},
1857
+ url={https://arxiv.org/abs/2406.19232},
1858
+ }
1859
+ ```
rublimp.jpg ADDED

Git LFS Details

  • SHA256: ba657d2c5bce313d34afab6f8422c3ca5091c7d1c84fffaff364719b1073b518
  • Pointer size: 131 Bytes
  • Size of remote file: 632 kB