Commit
·
20cd80c
1
Parent(s):
c5c45b4
Upload 3 files
Browse files- dataset_info.json +53 -0
- onestop_english.py +135 -0
- state.json +13 -0
dataset_info.json
ADDED
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"builder_name": "onestop_english",
|
3 |
+
"citation": "@inproceedings{vajjala-lucic-2018-onestopenglish,\n title = {OneStopEnglish corpus: A new corpus for automatic readability assessment and text simplification},\n author = {Sowmya Vajjala and Ivana Lu\u010di\u0107},\n booktitle = {Proceedings of the Thirteenth Workshop on Innovative Use of NLP for Building Educational Applications},\n year = {2018}\n}\n",
|
4 |
+
"config_name": "default",
|
5 |
+
"dataset_name": "onestop_english",
|
6 |
+
"dataset_size": 2278039,
|
7 |
+
"description": "This dataset is a compilation of the OneStopEnglish corpus of texts written at three reading levels into one file.\nText documents are classified into three reading levels - ele, int, adv (Elementary, Intermediate and Advance).\nThis dataset demonstrates its usefulness for through two applica-tions - automatic readability assessment and automatic text simplification.\nThe corpus consists of 189 texts, each in three versions/reading levels (567 in total).\n",
|
8 |
+
"download_checksums": {
|
9 |
+
"https://github.com/purvimisal/OneStopCorpus-Compiled/raw/main/Texts-SeparatedByReadingLevel.zip": {
|
10 |
+
"num_bytes": 1228804,
|
11 |
+
"checksum": null
|
12 |
+
}
|
13 |
+
},
|
14 |
+
"download_size": 1228804,
|
15 |
+
"features": {
|
16 |
+
"text": {
|
17 |
+
"dtype": "string",
|
18 |
+
"_type": "Value"
|
19 |
+
},
|
20 |
+
"label": {
|
21 |
+
"names": [
|
22 |
+
"ele",
|
23 |
+
"int",
|
24 |
+
"adv"
|
25 |
+
],
|
26 |
+
"_type": "ClassLabel"
|
27 |
+
}
|
28 |
+
},
|
29 |
+
"homepage": "https://github.com/nishkalavallabhi/OneStopEnglishCorpus",
|
30 |
+
"license": "Creative Commons Attribution-ShareAlike 4.0 International License",
|
31 |
+
"size_in_bytes": 3506843,
|
32 |
+
"splits": {
|
33 |
+
"train": {
|
34 |
+
"name": "train",
|
35 |
+
"num_bytes": 2278039,
|
36 |
+
"num_examples": 567,
|
37 |
+
"dataset_name": "onestop_english"
|
38 |
+
}
|
39 |
+
},
|
40 |
+
"supervised_keys": {},
|
41 |
+
"task_templates": [
|
42 |
+
{
|
43 |
+
"task": "text-classification",
|
44 |
+
"label_column": "label"
|
45 |
+
}
|
46 |
+
],
|
47 |
+
"version": {
|
48 |
+
"version_str": "1.1.0",
|
49 |
+
"major": 1,
|
50 |
+
"minor": 1,
|
51 |
+
"patch": 0
|
52 |
+
}
|
53 |
+
}
|
onestop_english.py
ADDED
@@ -0,0 +1,135 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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 |
+
"""OneStopEnglish Corpus: Dataset of texts classified into reading levels/text complexities."""
|
16 |
+
|
17 |
+
|
18 |
+
import os
|
19 |
+
|
20 |
+
import datasets
|
21 |
+
from datasets.tasks import TextClassification
|
22 |
+
|
23 |
+
|
24 |
+
logger = datasets.logging.get_logger(__name__)
|
25 |
+
|
26 |
+
|
27 |
+
_CITATION = """\
|
28 |
+
@inproceedings{vajjala-lucic-2018-onestopenglish,
|
29 |
+
title = {OneStopEnglish corpus: A new corpus for automatic readability assessment and text simplification},
|
30 |
+
author = {Sowmya Vajjala and Ivana Lučić},
|
31 |
+
booktitle = {Proceedings of the Thirteenth Workshop on Innovative Use of NLP for Building Educational Applications},
|
32 |
+
year = {2018}
|
33 |
+
}
|
34 |
+
"""
|
35 |
+
|
36 |
+
_DESCRIPTION = """\
|
37 |
+
This dataset is a compilation of the OneStopEnglish corpus of texts written at three reading levels into one file.
|
38 |
+
Text documents are classified into three reading levels - ele, int, adv (Elementary, Intermediate and Advance).
|
39 |
+
This dataset demonstrates its usefulness for through two applica-tions - automatic readability assessment and automatic text simplification.
|
40 |
+
The corpus consists of 189 texts, each in three versions/reading levels (567 in total).
|
41 |
+
"""
|
42 |
+
|
43 |
+
_HOMEPAGE = "https://github.com/nishkalavallabhi/OneStopEnglishCorpus"
|
44 |
+
|
45 |
+
_LICENSE = "Creative Commons Attribution-ShareAlike 4.0 International License"
|
46 |
+
|
47 |
+
_URL = "https://github.com/purvimisal/OneStopCorpus-Compiled/raw/main/Texts-SeparatedByReadingLevel.zip"
|
48 |
+
|
49 |
+
|
50 |
+
# TODO: Name of the dataset usually match the script name with CamelCase instead of snake_case
|
51 |
+
class OnestopEnglish(datasets.GeneratorBasedBuilder):
|
52 |
+
"""OneStopEnglish Corpus: Dataset of texts classified into reading levels"""
|
53 |
+
|
54 |
+
VERSION = datasets.Version("1.1.0")
|
55 |
+
|
56 |
+
def _info(self):
|
57 |
+
# TODO: This method specifies the datasets.DatasetInfo object which contains informations and typings for the dataset
|
58 |
+
return datasets.DatasetInfo(
|
59 |
+
description=_DESCRIPTION,
|
60 |
+
features=datasets.Features(
|
61 |
+
{"text": datasets.Value("string"), "label": datasets.features.ClassLabel(names=["ele", "int", "adv"])}
|
62 |
+
),
|
63 |
+
supervised_keys=[""],
|
64 |
+
homepage=_HOMEPAGE,
|
65 |
+
license=_LICENSE,
|
66 |
+
citation=_CITATION,
|
67 |
+
task_templates=[TextClassification(text_column="text", label_column="label")],
|
68 |
+
)
|
69 |
+
|
70 |
+
def _vocab_text_gen(self, train_file):
|
71 |
+
for _, ex in self._generate_examples(train_file):
|
72 |
+
yield ex["text"]
|
73 |
+
|
74 |
+
def _split_generators(self, dl_manager):
|
75 |
+
"""Downloads OneStopEnglish corpus"""
|
76 |
+
extracted_folder_path = dl_manager.download_and_extract(_URL)
|
77 |
+
return [
|
78 |
+
datasets.SplitGenerator(
|
79 |
+
name=datasets.Split.TRAIN,
|
80 |
+
gen_kwargs={"split_key": "train", "data_dir": extracted_folder_path},
|
81 |
+
)
|
82 |
+
]
|
83 |
+
|
84 |
+
def _get_examples_from_split(self, split_key, data_dir):
|
85 |
+
"""Reads the downloaded and extracted files and combines the individual text files to one dataset."""
|
86 |
+
|
87 |
+
data_dir = os.path.join(data_dir, "Texts-SeparatedByReadingLevel")
|
88 |
+
|
89 |
+
ele_samples = []
|
90 |
+
dir_path = os.path.join(data_dir, "Ele-Txt")
|
91 |
+
files = os.listdir(dir_path)
|
92 |
+
for f in sorted(files):
|
93 |
+
try:
|
94 |
+
with open(os.path.join(dir_path, f), encoding="utf-8-sig") as myfile:
|
95 |
+
text = myfile.read().strip()
|
96 |
+
ele_samples.append(text)
|
97 |
+
except Exception as e:
|
98 |
+
logger.info("Error with:", os.path.join(dir_path, f), e)
|
99 |
+
|
100 |
+
int_samples = []
|
101 |
+
dir_path = os.path.join(data_dir, "Int-Txt")
|
102 |
+
files = os.listdir(dir_path)
|
103 |
+
for f in sorted(files):
|
104 |
+
try:
|
105 |
+
with open(os.path.join(dir_path, f), encoding="utf-8-sig") as myfile:
|
106 |
+
text = myfile.read().strip()
|
107 |
+
int_samples.append(text)
|
108 |
+
except Exception as e:
|
109 |
+
logger.info("Error with:", os.path.join(dir_path, f), e)
|
110 |
+
|
111 |
+
adv_samples = []
|
112 |
+
dir_path = os.path.join(data_dir, "Adv-Txt")
|
113 |
+
files = os.listdir(dir_path)
|
114 |
+
for f in sorted(files):
|
115 |
+
try:
|
116 |
+
with open(os.path.join(dir_path, f), encoding="utf-8-sig") as myfile:
|
117 |
+
text = myfile.read().strip()
|
118 |
+
adv_samples.append(text)
|
119 |
+
except Exception as e:
|
120 |
+
logger.info("Error with:", os.path.join(dir_path, f), e)
|
121 |
+
|
122 |
+
train_samples = ele_samples + int_samples + adv_samples
|
123 |
+
train_labels = (["ele"] * len(ele_samples)) + (["int"] * len(int_samples)) + (["adv"] * len(adv_samples))
|
124 |
+
|
125 |
+
if split_key == "train":
|
126 |
+
return (train_samples, train_labels)
|
127 |
+
else:
|
128 |
+
raise ValueError(f"Invalid split key {split_key}")
|
129 |
+
|
130 |
+
def _generate_examples(self, split_key, data_dir):
|
131 |
+
"""Yields examples for a given split of dataset."""
|
132 |
+
split_text, split_labels = self._get_examples_from_split(split_key, data_dir)
|
133 |
+
for id_, (text, label) in enumerate(zip(split_text, split_labels)):
|
134 |
+
feature_dict = {"text": text, "label": label}
|
135 |
+
yield id_, feature_dict
|
state.json
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_data_files": [
|
3 |
+
{
|
4 |
+
"filename": "data-00000-of-00001.arrow"
|
5 |
+
}
|
6 |
+
],
|
7 |
+
"_fingerprint": "c6f001bca857b8c0",
|
8 |
+
"_format_columns": null,
|
9 |
+
"_format_kwargs": {},
|
10 |
+
"_format_type": null,
|
11 |
+
"_output_all_columns": false,
|
12 |
+
"_split": "train"
|
13 |
+
}
|