Update ESSAI.py
Browse files
ESSAI.py
CHANGED
@@ -41,28 +41,63 @@ _LICENSE = 'Data User Agreement'
|
|
41 |
|
42 |
class ESSAI(datasets.GeneratorBasedBuilder):
|
43 |
|
44 |
-
DEFAULT_CONFIG_NAME = "
|
45 |
|
46 |
BUILDER_CONFIGS = [
|
47 |
-
datasets.BuilderConfig(name="
|
|
|
|
|
|
|
|
|
|
|
|
|
48 |
]
|
49 |
|
50 |
def _info(self):
|
51 |
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
|
59 |
-
|
60 |
-
|
61 |
-
|
62 |
-
names = ['
|
63 |
-
|
64 |
-
|
65 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
66 |
|
67 |
return datasets.DatasetInfo(
|
68 |
description=_DESCRIPTION,
|
@@ -111,56 +146,178 @@ class ESSAI(datasets.GeneratorBasedBuilder):
|
|
111 |
|
112 |
key = 0
|
113 |
|
114 |
-
|
115 |
|
116 |
-
|
117 |
-
id_docs = []
|
118 |
-
id_words = []
|
119 |
-
words = []
|
120 |
-
lemmas = []
|
121 |
-
POS_tags = []
|
122 |
|
123 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
124 |
|
125 |
-
|
126 |
|
127 |
-
|
128 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
129 |
|
130 |
-
id_doc, id_word, word, lemma, tag = line.split("\t")[0:5]
|
131 |
-
|
132 |
-
id_docs.append(id_doc)
|
133 |
-
id_words.append(id_word)
|
134 |
-
words.append(word)
|
135 |
-
lemmas.append(lemma)
|
136 |
-
POS_tags.append(tag)
|
137 |
-
|
138 |
-
dic = {
|
139 |
-
"id_docs": np.array(list(map(int, id_docs))),
|
140 |
-
"id_words": id_words,
|
141 |
-
"words": words,
|
142 |
-
"lemmas": lemmas,
|
143 |
-
"POS_tags": POS_tags,
|
144 |
-
}
|
145 |
-
|
146 |
-
for doc_id in set(dic["id_docs"]):
|
147 |
-
|
148 |
-
indexes = np.argwhere(dic["id_docs"] == doc_id)[:, 0]
|
149 |
-
tokens = [dic["words"][id] for id in indexes]
|
150 |
-
text_lemmas = [dic["lemmas"][id] for id in indexes]
|
151 |
-
pos_tags = [dic["POS_tags"][id] for id in indexes]
|
152 |
-
|
153 |
-
all_res.append({
|
154 |
-
"id": key,
|
155 |
-
"document_id": doc_id,
|
156 |
-
"tokens": tokens,
|
157 |
-
"lemmas": text_lemmas,
|
158 |
-
"pos_tags": pos_tags,
|
159 |
-
"label": label,
|
160 |
-
})
|
161 |
-
|
162 |
-
key += 1
|
163 |
-
|
164 |
ids = [r["id"] for r in all_res]
|
165 |
|
166 |
random.seed(4)
|
|
|
41 |
|
42 |
class ESSAI(datasets.GeneratorBasedBuilder):
|
43 |
|
44 |
+
DEFAULT_CONFIG_NAME = "pos_spec"
|
45 |
|
46 |
BUILDER_CONFIGS = [
|
47 |
+
datasets.BuilderConfig(name="pos", version="1.0.0", description="The ESSAI corpora - POS Speculation task"),
|
48 |
+
|
49 |
+
datasets.BuilderConfig(name="cls_spec", version="1.0.0", description="The ESSAI corpora - CLS Speculation task"),
|
50 |
+
datasets.BuilderConfig(name="cls_neg", version="1.0.0", description="The ESSAI corpora - CLS Negation task"),
|
51 |
+
|
52 |
+
datasets.BuilderConfig(name="ner_spec", version="1.0.0", description="The ESSAI corpora - NER Speculation task"),
|
53 |
+
datasets.BuilderConfig(name="ner_neg", version="1.0.0", description="The ESSAI corpora - NER Negation task"),
|
54 |
]
|
55 |
|
56 |
def _info(self):
|
57 |
|
58 |
+
if self.config.name.find("pos") != -1:
|
59 |
+
|
60 |
+
features = datasets.Features(
|
61 |
+
{
|
62 |
+
"id": datasets.Value("string"),
|
63 |
+
"document_id": datasets.Value("string"),
|
64 |
+
"tokens": [datasets.Value("string")],
|
65 |
+
"lemmas": [datasets.Value("string")],
|
66 |
+
"pos_tags": [datasets.Value("string")],
|
67 |
+
# "pos_tags": [datasets.features.ClassLabel(
|
68 |
+
# names = ['VER:pper', 'VER:subi', 'VER:cond', 'INT', 'VER:infi', 'PUN:cit', 'ITAC', 'PUN', 'VER:ppre', 'VER:pres', 'PRO:REL', 'ADJ', 'VER:subp', 'NN', 'PREF', 'PRP', 'PRO:IND', 'PRO:POS', 'DET:POS', 'VER:futu', 'PRO:DEM', 'KON', 'DET:ART', 'VER:', 'PRP:det', 'PRO', 'FAG', 'NOM', 'SYM', 'VER:impf', 'CIT02-HM', 'SENT', 'Bayer', 'VER:simp', 'ADV', 'bayer', '@card@', 'PRO:PER', 'NUM', 'ABR', 'NAM'],
|
69 |
+
# )],
|
70 |
+
}
|
71 |
+
)
|
72 |
+
|
73 |
+
elif self.config.name.find("cls") != -1:
|
74 |
+
|
75 |
+
features = datasets.Features(
|
76 |
+
{
|
77 |
+
"id": datasets.Value("string"),
|
78 |
+
"document_id": datasets.Value("string"),
|
79 |
+
"tokens": [datasets.Value("string")],
|
80 |
+
"label": datasets.Value("string"),
|
81 |
+
# "label": datasets.features.ClassLabel(
|
82 |
+
# names = ['VER:pper', 'VER:subi', 'VER:cond', 'INT', 'VER:infi', 'PUN:cit', 'ITAC', 'PUN', 'VER:ppre', 'VER:pres', 'PRO:REL', 'ADJ', 'VER:subp', 'NN', 'PREF', 'PRP', 'PRO:IND', 'PRO:POS', 'DET:POS', 'VER:futu', 'PRO:DEM', 'KON', 'DET:ART', 'VER:', 'PRP:det', 'PRO', 'FAG', 'NOM', 'SYM', 'VER:impf', 'CIT02-HM', 'SENT', 'Bayer', 'VER:simp', 'ADV', 'bayer', '@card@', 'PRO:PER', 'NUM', 'ABR', 'NAM'],
|
83 |
+
# ),
|
84 |
+
}
|
85 |
+
)
|
86 |
+
|
87 |
+
elif self.config.name.find("ner") != -1:
|
88 |
+
|
89 |
+
features = datasets.Features(
|
90 |
+
{
|
91 |
+
"id": datasets.Value("string"),
|
92 |
+
"document_id": datasets.Value("string"),
|
93 |
+
"tokens": [datasets.Value("string")],
|
94 |
+
"lemmas": [datasets.Value("string")],
|
95 |
+
"ner_tags": [datasets.Value("string")],
|
96 |
+
# "ner_tags": [datasets.features.ClassLabel(
|
97 |
+
# names = ['VER:pper', 'VER:subi', 'VER:cond', 'INT', 'VER:infi', 'PUN:cit', 'ITAC', 'PUN', 'VER:ppre', 'VER:pres', 'PRO:REL', 'ADJ', 'VER:subp', 'NN', 'PREF', 'PRP', 'PRO:IND', 'PRO:POS', 'DET:POS', 'VER:futu', 'PRO:DEM', 'KON', 'DET:ART', 'VER:', 'PRP:det', 'PRO', 'FAG', 'NOM', 'SYM', 'VER:impf', 'CIT02-HM', 'SENT', 'Bayer', 'VER:simp', 'ADV', 'bayer', '@card@', 'PRO:PER', 'NUM', 'ABR', 'NAM'],
|
98 |
+
# )],
|
99 |
+
}
|
100 |
+
)
|
101 |
|
102 |
return datasets.DatasetInfo(
|
103 |
description=_DESCRIPTION,
|
|
|
146 |
|
147 |
key = 0
|
148 |
|
149 |
+
subset = self.config.name.split("_")[-1]
|
150 |
|
151 |
+
unique_id_doc = []
|
|
|
|
|
|
|
|
|
|
|
152 |
|
153 |
+
if self.config.name.find("pos") != -1:
|
154 |
+
docs = ["ESSAI_neg.txt", "ESSAI_spec.txt"]
|
155 |
+
else:
|
156 |
+
docs = [f"ESSAI_{subset}.txt"]
|
157 |
+
|
158 |
+
for file in docs:
|
159 |
+
|
160 |
+
filename = os.path.join(datadir, file)
|
161 |
+
|
162 |
+
if self.config.name.find("pos") != -1:
|
163 |
+
|
164 |
+
id_docs = []
|
165 |
+
id_words = []
|
166 |
+
words = []
|
167 |
+
lemmas = []
|
168 |
+
POS_tags = []
|
169 |
+
|
170 |
+
with open(filename) as f:
|
171 |
+
|
172 |
+
for line in f.readlines():
|
173 |
+
|
174 |
+
splitted = line.split("\t")
|
175 |
+
|
176 |
+
if len(splitted) < 5:
|
177 |
+
continue
|
178 |
+
|
179 |
+
id_doc, id_word, word, lemma, tag = splitted[0:5]
|
180 |
+
if len(splitted) >= 8:
|
181 |
+
tag = splitted[6]
|
182 |
+
|
183 |
+
if tag == "@card@":
|
184 |
+
print(splitted)
|
185 |
+
|
186 |
+
if lemma == "000" and tag == "@card@":
|
187 |
+
tag = "NUM"
|
188 |
+
word = "100 000"
|
189 |
+
lemma = "100 000"
|
190 |
+
elif lemma == "45" and tag == "@card@":
|
191 |
+
tag = "NUM"
|
192 |
+
|
193 |
+
# if id_doc in id_docs:
|
194 |
+
# continue
|
195 |
+
|
196 |
+
id_docs.append(id_doc)
|
197 |
+
id_words.append(id_word)
|
198 |
+
words.append(word)
|
199 |
+
lemmas.append(lemma)
|
200 |
+
POS_tags.append(tag)
|
201 |
+
|
202 |
+
dic = {
|
203 |
+
"id_docs": np.array(list(map(int, id_docs))),
|
204 |
+
"id_words": id_words,
|
205 |
+
"words": words,
|
206 |
+
"lemmas": lemmas,
|
207 |
+
"POS_tags": POS_tags,
|
208 |
+
}
|
209 |
+
|
210 |
+
for doc_id in set(dic["id_docs"]):
|
211 |
+
|
212 |
+
indexes = np.argwhere(dic["id_docs"] == doc_id)[:, 0]
|
213 |
+
tokens = [dic["words"][id] for id in indexes]
|
214 |
+
text_lemmas = [dic["lemmas"][id] for id in indexes]
|
215 |
+
pos_tags = [dic["POS_tags"][id] for id in indexes]
|
216 |
+
|
217 |
+
if doc_id not in unique_id_doc:
|
218 |
+
|
219 |
+
all_res.append({
|
220 |
+
"id": str(doc_id),
|
221 |
+
"document_id": doc_id,
|
222 |
+
"tokens": tokens,
|
223 |
+
"lemmas": text_lemmas,
|
224 |
+
"pos_tags": pos_tags,
|
225 |
+
})
|
226 |
+
unique_id_doc.append(doc_id)
|
227 |
+
|
228 |
+
# key += 1
|
229 |
+
|
230 |
+
elif self.config.name.find("ner") != -1:
|
231 |
+
|
232 |
+
id_docs = []
|
233 |
+
id_words = []
|
234 |
+
words = []
|
235 |
+
lemmas = []
|
236 |
+
ner_tags = []
|
237 |
+
|
238 |
+
with open(filename) as f:
|
239 |
+
|
240 |
+
for line in f.readlines():
|
241 |
+
|
242 |
+
if len(line.split("\t")) < 5:
|
243 |
+
continue
|
244 |
+
|
245 |
+
id_doc, id_word, word, lemma, _ = line.split("\t")[0:5]
|
246 |
+
tag = line.replace("\n","").split("\t")[-1]
|
247 |
+
if tag == "***" or tag == "_":
|
248 |
+
tag = "O"
|
249 |
+
elif tag == "v":
|
250 |
+
tag = "I_scope_spec"
|
251 |
+
elif tag == "z":
|
252 |
+
tag = "O"
|
253 |
+
|
254 |
+
id_docs.append(id_doc)
|
255 |
+
id_words.append(id_word)
|
256 |
+
words.append(word)
|
257 |
+
lemmas.append(lemma)
|
258 |
+
ner_tags.append(tag)
|
259 |
+
|
260 |
+
dic = {
|
261 |
+
"id_docs": np.array(list(map(int, id_docs))),
|
262 |
+
"id_words": id_words,
|
263 |
+
"words": words,
|
264 |
+
"lemmas": lemmas,
|
265 |
+
"ner_tags": ner_tags,
|
266 |
+
}
|
267 |
+
|
268 |
+
for doc_id in set(dic["id_docs"]):
|
269 |
+
|
270 |
+
indexes = np.argwhere(dic["id_docs"] == doc_id)[:, 0]
|
271 |
+
tokens = [dic["words"][id] for id in indexes]
|
272 |
+
text_lemmas = [dic["lemmas"][id] for id in indexes]
|
273 |
+
ner_tags = [dic["ner_tags"][id] for id in indexes]
|
274 |
+
|
275 |
+
all_res.append({
|
276 |
+
"id": key,
|
277 |
+
"document_id": doc_id,
|
278 |
+
"tokens": tokens,
|
279 |
+
"lemmas": text_lemmas,
|
280 |
+
"ner_tags": ner_tags,
|
281 |
+
})
|
282 |
+
|
283 |
+
key += 1
|
284 |
+
|
285 |
+
elif self.config.name.find("cls") != -1:
|
286 |
+
|
287 |
+
f_in = open(filename, "r")
|
288 |
+
conll = [
|
289 |
+
[b.split("\t") for b in a.split("\n")]
|
290 |
+
for a in f_in.read().split("\n\n")
|
291 |
+
]
|
292 |
+
f_in.close()
|
293 |
+
|
294 |
+
classe = "negation" if self.config.name.find("neg") != -1 else "speculation"
|
295 |
+
|
296 |
+
all_res = []
|
297 |
+
|
298 |
+
for document in conll:
|
299 |
+
|
300 |
+
if document == [""]:
|
301 |
+
continue
|
302 |
|
303 |
+
identifier = document[0][0]
|
304 |
|
305 |
+
unique = list(set([w[-1] for w in document]))
|
306 |
+
# print(document)
|
307 |
+
tokens = [sent[2] for sent in document if len(sent) > 1]
|
308 |
+
|
309 |
+
if "***" in unique:
|
310 |
+
l = "none"
|
311 |
+
elif "_" in unique:
|
312 |
+
l = classe
|
313 |
+
|
314 |
+
all_res.append({
|
315 |
+
"id": str(identifier),
|
316 |
+
"document_id": identifier,
|
317 |
+
"tokens": tokens,
|
318 |
+
"label": l,
|
319 |
+
})
|
320 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
321 |
ids = [r["id"] for r in all_res]
|
322 |
|
323 |
random.seed(4)
|