asahi417 commited on
Commit
285e4d9
1 Parent(s): bb3d557
.gitattributes CHANGED
@@ -61,3 +61,6 @@ data/tweet_ner7/test.jsonl filter=lfs diff=lfs merge=lfs -text
61
  data/tweet_ner7/train.jsonl filter=lfs diff=lfs merge=lfs -text
62
  data/tweet_ner7/validation.jsonl filter=lfs diff=lfs merge=lfs -text
63
  data/tweet_qa/test.jsonl filter=lfs diff=lfs merge=lfs -text
 
 
 
 
61
  data/tweet_ner7/train.jsonl filter=lfs diff=lfs merge=lfs -text
62
  data/tweet_ner7/validation.jsonl filter=lfs diff=lfs merge=lfs -text
63
  data/tweet_qa/test.jsonl filter=lfs diff=lfs merge=lfs -text
64
+ data/tweet_intimacy/validation.jsonl filter=lfs diff=lfs merge=lfs -text
65
+ data/tweet_intimacy/test.jsonl filter=lfs diff=lfs merge=lfs -text
66
+ data/tweet_intimacy/train.jsonl filter=lfs diff=lfs merge=lfs -text
data/tweet_intimacy/test.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65d90ca4aa24f1ce66dee43de896a3cec2d66cc61ef99cfa5fca890843b999cd
3
+ size 39034
data/tweet_intimacy/train.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d298076649d94845628a800b2fb9f801b009518d9cde6b45845d21b06de0f30d
3
+ size 121089
data/tweet_intimacy/validation.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:40a4dc181f8cf3a8e259e510efc533762c7ac7baee7f8058e7b8d23e37f89e85
3
+ size 40483
process/tweet_intimacy.py CHANGED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import os
2
+ import json
3
+ from random import shuffle, seed
4
+ import pandas as pd
5
+
6
+ os.makedirs("data/tweet_intimacy", exist_ok=True)
7
+ df_test = pd.read_csv("misc/multilingual_tweet_intimacy/test.csv")
8
+ df_test = df_test[df_test['language'] == 'English']
9
+ df_test.pop("language")
10
+ test = [i.to_dict() for _, i in df_test.iterrows()]
11
+ for i in test:
12
+ i['label_float'] = i.pop("label")
13
+
14
+ df_train = pd.read_csv("misc/multilingual_tweet_intimacy/train.csv")
15
+ df_train = df_train[df_train['language'] == 'English']
16
+ df_train.pop("language")
17
+ train = [i.to_dict() for _, i in df_train.iterrows()]
18
+ for i in train:
19
+ i['label_float'] = i.pop("label")
20
+ seed(42)
21
+ shuffle(train)
22
+ val = train[:len(test)]
23
+ train = train[len(test):]
24
+
25
+ with open("data/tweet_intimacy/train.jsonl", "w") as f:
26
+ f.write("\n".join([json.dumps(i) for i in train]))
27
+ with open("data/tweet_intimacy/validation.jsonl", "w") as f:
28
+ f.write("\n".join([json.dumps(i) for i in val]))
29
+ with open("data/tweet_intimacy/test.jsonl", "w") as f:
30
+ f.write("\n".join([json.dumps(i) for i in test]))
31
+
32
+
33
+
super_tweet_eval.py CHANGED
@@ -63,6 +63,20 @@ _TWEET_QA_CITATION = """\
63
  year={2019}
64
  }
65
  """
 
 
 
 
 
 
 
 
 
 
 
 
 
 
66
 
67
 
68
  class SuperTweetEvalConfig(datasets.BuilderConfig):
@@ -113,6 +127,13 @@ class SuperTweetEval(datasets.GeneratorBasedBuilder):
113
  citation=_TWEET_QA_CITATION,
114
  features=["text", "label_str", "paragraph", "question"],
115
  data_url="https://huggingface.co/datasets/cardiffnlp/super_tweet_eval/resolve/main/data/tweet_qa",
 
 
 
 
 
 
 
116
  )
117
  ]
118
 
@@ -130,6 +151,8 @@ class SuperTweetEval(datasets.GeneratorBasedBuilder):
130
  'B-corporation', 'B-creative_work', 'B-event', 'B-group', 'B-location', 'B-person', 'B-product',
131
  'I-corporation', 'I-creative_work', 'I-event', 'I-group', 'I-location', 'I-person', 'I-product', 'O']
132
  features["label_sequence"] = datasets.Sequence(datasets.features.ClassLabel(names=names))
 
 
133
 
134
  return datasets.DatasetInfo(
135
  description=_SUPER_TWEET_EVAL_DESCRIPTION + "\n" + self.config.description,
 
63
  year={2019}
64
  }
65
  """
66
+ _TWEET_INTIMACY_DESCRIPTION = """\
67
+ See https://sites.google.com/umich.edu/semeval-2023-tweet-intimacy/home for more detail. We randomly take the same amount of
68
+ test set as the validation set.
69
+ """
70
+ _TWEET_INTIMACY_CITATION = """\
71
+ @misc{pei2023semeval,
72
+ title={SemEval 2023 Task 9: Multilingual Tweet Intimacy Analysis},
73
+ author={Jiaxin Pei and Vítor Silva and Maarten Bos and Yozon Liu and Leonardo Neves and David Jurgens and Francesco Barbieri},
74
+ year={2023},
75
+ eprint={2210.01108},
76
+ archivePrefix={arXiv},
77
+ primaryClass={cs.CL}
78
+ }
79
+ """
80
 
81
 
82
  class SuperTweetEvalConfig(datasets.BuilderConfig):
 
127
  citation=_TWEET_QA_CITATION,
128
  features=["text", "label_str", "paragraph", "question"],
129
  data_url="https://huggingface.co/datasets/cardiffnlp/super_tweet_eval/resolve/main/data/tweet_qa",
130
+ ),
131
+ SuperTweetEvalConfig(
132
+ name="tweet_intimacy",
133
+ description=_TWEET_INTIMACY_DESCRIPTION,
134
+ citation=_TWEET_INTIMACY_CITATION,
135
+ features=["text", "label_float"],
136
+ data_url="https://huggingface.co/datasets/cardiffnlp/super_tweet_eval/resolve/main/data/tweet_intimacy",
137
  )
138
  ]
139
 
 
151
  'B-corporation', 'B-creative_work', 'B-event', 'B-group', 'B-location', 'B-person', 'B-product',
152
  'I-corporation', 'I-creative_work', 'I-event', 'I-group', 'I-location', 'I-person', 'I-product', 'O']
153
  features["label_sequence"] = datasets.Sequence(datasets.features.ClassLabel(names=names))
154
+ if self.config.name == "tweet_intimacy":
155
+ features["label_float"] = datasets.Value("float32")
156
 
157
  return datasets.DatasetInfo(
158
  description=_SUPER_TWEET_EVAL_DESCRIPTION + "\n" + self.config.description,