albertvillanova HF staff commited on
Commit
6e7308d
1 Parent(s): 6f30099

Add dataset loading script

Browse files
Files changed (1) hide show
  1. vietnamese_students_feedback.py +112 -0
vietnamese_students_feedback.py ADDED
@@ -0,0 +1,112 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # coding=utf-8
2
+ # Copyright 2020 The HuggingFace Datasets Authors and 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.
6
+ # You may obtain a copy of the License at
7
+ #
8
+ # http://www.apache.org/licenses/LICENSE-2.0
9
+ #
10
+ # Unless required by applicable law or agreed to in writing, software
11
+ # distributed under the License is distributed on an "AS IS" BASIS,
12
+ # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13
+ # See the License for the specific language governing permissions and
14
+ # limitations under the License.
15
+ """Vietnamese Students’ Feedback Corpus."""
16
+
17
+ import datasets
18
+
19
+
20
+ _CITATION = """\
21
+ @InProceedings{8573337,
22
+ author={Nguyen, Kiet Van and Nguyen, Vu Duc and Nguyen, Phu X. V. and Truong, Tham T. H. and Nguyen, Ngan Luu-Thuy},
23
+ booktitle={2018 10th International Conference on Knowledge and Systems Engineering (KSE)},
24
+ title={UIT-VSFC: Vietnamese Students’ Feedback Corpus for Sentiment Analysis},
25
+ year={2018},
26
+ volume={},
27
+ number={},
28
+ pages={19-24},
29
+ doi={10.1109/KSE.2018.8573337}
30
+ }
31
+ """
32
+
33
+ _DESCRIPTION = """\
34
+ Students’ feedback is a vital resource for the interdisciplinary research involving the combining of two different
35
+ research fields between sentiment analysis and education.
36
+
37
+ Vietnamese Students’ Feedback Corpus (UIT-VSFC) is the resource consists of over 16,000 sentences which are
38
+ human-annotated with two different tasks: sentiment-based and topic-based classifications.
39
+
40
+ To assess the quality of our corpus, we measure the annotator agreements and classification evaluation on the
41
+ UIT-VSFC corpus. As a result, we obtained the inter-annotator agreement of sentiments and topics with more than over
42
+ 91% and 71% respectively. In addition, we built the baseline model with the Maximum Entropy classifier and achieved
43
+ approximately 88% of the sentiment F1-score and over 84% of the topic F1-score.
44
+ """
45
+
46
+ _HOMEPAGE = "https://sites.google.com/uit.edu.vn/uit-nlp/datasets-projects#h.p_4Brw8L-cbfTe"
47
+
48
+ # TODO: Add the licence for the dataset here if you can find it
49
+ # _LICENSE = ""
50
+
51
+ _URLS = {
52
+ datasets.Split.TRAIN: {
53
+ "sentences": "https://drive.google.com/uc?id=1nzak5OkrheRV1ltOGCXkT671bmjODLhP&export=download",
54
+ "sentiments": "https://drive.google.com/uc?id=1ye-gOZIBqXdKOoi_YxvpT6FeRNmViPPv&export=down load",
55
+ "topics": "https://drive.google.com/uc?id=14MuDtwMnNOcr4z_8KdpxprjbwaQ7lJ_C&export=download",
56
+ },
57
+ datasets.Split.VALIDATION: {
58
+ "sentences": "https://drive.google.com/uc?id=1sMJSR3oRfPc3fe1gK-V3W5F24tov_517&export=download",
59
+ "sentiments": "https://drive.google.com/uc?id=1GiY1AOp41dLXIIkgES4422AuDwmbUseL&export=download",
60
+ "topics": "https://drive.google.com/uc?id=1DwLgDEaFWQe8mOd7EpF-xqMEbDLfdT-W&export=download",
61
+ },
62
+ datasets.Split.TEST: {
63
+ "sentences": "https://drive.google.com/uc?id=1aNMOeZZbNwSRkjyCWAGtNCMa3YrshR-n&export=download",
64
+ "sentiments": "https://drive.google.com/uc?id=1vkQS5gI0is4ACU58-AbWusnemw7KZNfO&export=download",
65
+ "topics": "https://drive.google.com/uc?id=1_ArMpDguVsbUGl-xSMkTF_p5KpZrmpSB&export=download",
66
+ },
67
+ }
68
+
69
+
70
+ class VietnameseStudentsFeedback(datasets.GeneratorBasedBuilder):
71
+ """Vietnamese Students’ Feedback Corpus."""
72
+
73
+ VERSION = datasets.Version("1.0.0")
74
+
75
+ def _info(self):
76
+ return datasets.DatasetInfo(
77
+ description=_DESCRIPTION,
78
+ features=datasets.Features(
79
+ {
80
+ "sentence": datasets.Value("string"),
81
+ "sentiment": datasets.ClassLabel(names=["negative", "neutral", "positive"]),
82
+ "topic": datasets.ClassLabel(names=["lecturer", "training_program", "facility", "others"]),
83
+ }
84
+ ),
85
+ homepage=_HOMEPAGE,
86
+ # license=_LICENSE,
87
+ citation=_CITATION,
88
+ )
89
+
90
+ def _split_generators(self, dl_manager):
91
+ data_dir = dl_manager.download(_URLS)
92
+ return [
93
+ datasets.SplitGenerator(
94
+ name=split,
95
+ gen_kwargs={
96
+ "sentences_path": data_dir[split]["sentences"],
97
+ "sentiments_path": data_dir[split]["sentiments"],
98
+ "topics_path": data_dir[split]["topics"],
99
+ },
100
+ ) for split in _URLS
101
+ ]
102
+
103
+ def _generate_examples(self, sentences_path, sentiments_path, topics_path):
104
+ with open(sentences_path, encoding="utf-8") as sentences, open(
105
+ sentiments_path, encoding="utf-8"
106
+ ) as sentiments, open(topics_path, encoding="utf-8") as topics:
107
+ for key, (sentence, sentiment, topic) in enumerate(zip(sentences, sentiments, topics)):
108
+ yield key, {
109
+ "sentence": sentence.strip(),
110
+ "sentiment": int(sentiment.strip()),
111
+ "topic": int(topic.strip()),
112
+ }