Datasets:
Tasks:
Text Classification
Sub-tasks:
multi-class-classification
Size:
100K<n<1M
ArXiv:
Tags:
relation extraction
License:
Update multitacred.py
Browse files- multitacred.py +17 -13
multitacred.py
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
# coding=utf-8
|
2 |
-
# Copyright
|
3 |
#
|
4 |
# Licensed under the Apache License, Version 2.0 (the "License");
|
5 |
# you may not use this file except in compliance with the License.
|
@@ -81,32 +81,36 @@ _CITATION = """\
|
|
81 |
"""
|
82 |
|
83 |
_DESCRIPTION = """\
|
84 |
-
MultiTACRED is a multilingual version of the large-scale TAC Relation
|
85 |
-
It covers 12 typologically diverse languages from 9 language families, and was created by the
|
86 |
-
Technology group of DFKI by machine-translating the
|
87 |
-
projecting their entity annotations. For details of the original TACRED's data collection and
|
88 |
-
see
|
89 |
-
Any translations with an invalid tag structure, e.g.
|
90 |
-
discarded (on average, 2.3% of the instances).
|
91 |
|
92 |
Languages covered are: Arabic, Chinese, Finnish, French, German, Hindi, Hungarian, Japanese, Polish,
|
93 |
Russian, Spanish, Turkish. Intended use is supervised relation classification. Audience - researchers.
|
94 |
|
95 |
-
Please see our ACL paper for full details.
|
96 |
|
97 |
NOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:
|
98 |
- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid
|
99 |
The motivation for this is that we want to support additional languages, for which these fields were not required
|
100 |
or available. The reader expects the specification of a language-specific configuration specifying the variant
|
101 |
-
(original, revisited or retacred) and the language (as a two-letter iso code).
|
102 |
|
103 |
The DatasetReader changes the offsets of the following fields, to conform with standard Python usage (see
|
104 |
-
|
105 |
- subj_end to subj_end + 1 (make end offset exclusive)
|
106 |
- obj_end to obj_end + 1 (make end offset exclusive)
|
107 |
|
108 |
-
NOTE 2: The MultiTACRED dataset offers an additional 'split', namely the backtranslated test data (translated to
|
109 |
-
language and then back to English). To access this split,
|
|
|
|
|
|
|
|
|
110 |
"""
|
111 |
|
112 |
_HOMEPAGE = "https://github.com/DFKI-NLP/MultiTACRED"
|
|
|
1 |
# coding=utf-8
|
2 |
+
# Copyright 2022 The current dataset script contributor.
|
3 |
#
|
4 |
# Licensed under the Apache License, Version 2.0 (the "License");
|
5 |
# you may not use this file except in compliance with the License.
|
|
|
81 |
"""
|
82 |
|
83 |
_DESCRIPTION = """\
|
84 |
+
MultiTACRED is a multilingual version of the large-scale [https://nlp.stanford.edu/projects/tacred/](TAC Relation
|
85 |
+
Extraction Dataset). It covers 12 typologically diverse languages from 9 language families, and was created by the
|
86 |
+
Speech & Language Technology group of DFKI by machine-translating the instances of the original TACRED dataset and
|
87 |
+
automatically projecting their entity annotations. For details of the original TACRED's data collection and
|
88 |
+
annotation process, see the [https://aclanthology.org/D17-1004/](original paper). Translations are syntactically
|
89 |
+
validated by checking the correctness of the XML tag markup. Any translations with an invalid tag structure, e.g.
|
90 |
+
missing or invalid head or tail tag pairs, are discarded (on average, 2.3% of the instances).
|
91 |
|
92 |
Languages covered are: Arabic, Chinese, Finnish, French, German, Hindi, Hungarian, Japanese, Polish,
|
93 |
Russian, Spanish, Turkish. Intended use is supervised relation classification. Audience - researchers.
|
94 |
|
95 |
+
Please see [https://arxiv.org/abs/2305.04582](our ACL paper) for full details.
|
96 |
|
97 |
NOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:
|
98 |
- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid
|
99 |
The motivation for this is that we want to support additional languages, for which these fields were not required
|
100 |
or available. The reader expects the specification of a language-specific configuration specifying the variant
|
101 |
+
(original, revisited or retacred) and the language (as a two-letter iso code).
|
102 |
|
103 |
The DatasetReader changes the offsets of the following fields, to conform with standard Python usage (see
|
104 |
+
_generate_examples()):
|
105 |
- subj_end to subj_end + 1 (make end offset exclusive)
|
106 |
- obj_end to obj_end + 1 (make end offset exclusive)
|
107 |
|
108 |
+
NOTE 2: The MultiTACRED dataset offers an additional 'split', namely the backtranslated test data (translated to a
|
109 |
+
target language and then back to English). To access this split, use dataset['backtranslated_test'].
|
110 |
+
|
111 |
+
You can find the TACRED dataset reader for the English version of the dataset
|
112 |
+
[here](https://huggingface.co/datasets/DFKI-SLT/tacred).
|
113 |
+
|
114 |
"""
|
115 |
|
116 |
_HOMEPAGE = "https://github.com/DFKI-NLP/MultiTACRED"
|