Initial commit

.gitattributes

@@ -14,3 +14,6 @@
 *.pb filter=lfs diff=lfs merge=lfs -text
 *.pt filter=lfs diff=lfs merge=lfs -text
 *.pth filter=lfs diff=lfs merge=lfs -text
+data/train.tsv filter=lfs diff=lfs merge=lfs -text
+data/test.tsv filter=lfs diff=lfs merge=lfs -text
+data/valid.tsv filter=lfs diff=lfs merge=lfs -text

data/label.txt

@@ -0,0 +1,6 @@
+매우 긍정
+다소 긍정
+중립
+다소 부정
+매우 부정
+모름

data/test.tsv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b508c2654230f5f945a6da908a939c1aaac4a19c59d589acadf414f904475b6d
+size 256110

data/train.tsv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8722da721f809b4d7cec831fcdc97625c7eceb60c617e7e3ba6cf3c5bb8f7f3a
+size 1601405

data/valid.tsv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbf3cc0adbedde75edf9037889805925944d329fefbfcef8c6ee9ddc5915c826
+size 205479

talktalk-sentiment-210713-multi-singleturn-custom-multiturn.py

@@ -0,0 +1,66 @@
+from __future__ import absolute_import, division, print_function
+import datasets
+import json
+
+_URL = "data/"
+_URLs = {
+    "train": _URL + "train.tsv",
+    "valid": _URL + "valid.tsv",
+    "test": _URL + "test.tsv",
+}
+
+
+class TalktalkSentiment_210713_multi_singleturn_custom_multiturn(
+    datasets.GeneratorBasedBuilder):
+
+    def _info(self):
+        return datasets.DatasetInfo(
+            description="TalkTalk Sentiment Classification Multiturn Dataset (210713 Version)",
+            features=datasets.Features(
+                {
+                    "text": datasets.Value("string"),
+                    "label": datasets.features.ClassLabel(
+                        names=['매우 긍정', '다소 긍정', '중립', '다소 부정', '매우 부정', '모름']),
+                    "text_2": datasets.Value("string"),
+                    "label_2": datasets.Value("string"),
+                }
+            ),
+            supervised_keys=None,
+            license="",
+            homepage="",
+            citation="",
+        )
+
+    def _split_generators(self, dl_manager):
+        downloaded_files = dl_manager.download_and_extract(_URLs)
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "filepath": downloaded_files["train"],
+                }
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.VALIDATION,
+                gen_kwargs={
+                    "filepath": downloaded_files["valid"],
+                }
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.TEST,
+                gen_kwargs={
+                    "filepath": downloaded_files["test"],
+                }
+            ),
+        ]
+
+    def _generate_examples(self, filepath):
+        with open(filepath, "r", encoding='UTF-8') as f:
+            for idx, line in enumerate(f):
+                text, label = line.split("\t")
+                if text[0] == '[' and text[-1] == ']':
+                    yield idx, {"text": '', "label": '모름',
+                                "text_2": str(text), "label_2": str(label)}
+                else:
+                    yield idx, {"text": text.strip(), "label": label.strip(),
+                                "text_2": '', "label_2": ''}