Datasets:
Tasks:
Text2Text Generation
Sub-tasks:
abstractive-qa
Languages:
English
Size:
10K<n<100K
ArXiv:
License:
Commit
•
a82dfbc
0
Parent(s):
Update files from the datasets library (from 1.3.0)
Browse filesRelease notes: https://github.com/huggingface/datasets/releases/tag/1.3.0
- .gitattributes +27 -0
- README.md +199 -0
- dataset_infos.json +1 -0
- dummy/1.0.0/dummy_data.zip +3 -0
- narrativeqa_manual.py +220 -0
.gitattributes
ADDED
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
*.7z filter=lfs diff=lfs merge=lfs -text
|
2 |
+
*.arrow filter=lfs diff=lfs merge=lfs -text
|
3 |
+
*.bin filter=lfs diff=lfs merge=lfs -text
|
4 |
+
*.bin.* filter=lfs diff=lfs merge=lfs -text
|
5 |
+
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
6 |
+
*.ftz filter=lfs diff=lfs merge=lfs -text
|
7 |
+
*.gz filter=lfs diff=lfs merge=lfs -text
|
8 |
+
*.h5 filter=lfs diff=lfs merge=lfs -text
|
9 |
+
*.joblib filter=lfs diff=lfs merge=lfs -text
|
10 |
+
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
11 |
+
*.model filter=lfs diff=lfs merge=lfs -text
|
12 |
+
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
13 |
+
*.onnx filter=lfs diff=lfs merge=lfs -text
|
14 |
+
*.ot filter=lfs diff=lfs merge=lfs -text
|
15 |
+
*.parquet filter=lfs diff=lfs merge=lfs -text
|
16 |
+
*.pb filter=lfs diff=lfs merge=lfs -text
|
17 |
+
*.pt filter=lfs diff=lfs merge=lfs -text
|
18 |
+
*.pth filter=lfs diff=lfs merge=lfs -text
|
19 |
+
*.rar filter=lfs diff=lfs merge=lfs -text
|
20 |
+
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
21 |
+
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
22 |
+
*.tflite filter=lfs diff=lfs merge=lfs -text
|
23 |
+
*.tgz filter=lfs diff=lfs merge=lfs -text
|
24 |
+
*.xz filter=lfs diff=lfs merge=lfs -text
|
25 |
+
*.zip filter=lfs diff=lfs merge=lfs -text
|
26 |
+
*.zstandard filter=lfs diff=lfs merge=lfs -text
|
27 |
+
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
@@ -0,0 +1,199 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
annotations_creators:
|
3 |
+
- crowdsourced
|
4 |
+
language_creators:
|
5 |
+
- found
|
6 |
+
languages:
|
7 |
+
- en
|
8 |
+
licenses:
|
9 |
+
- apache-2-0
|
10 |
+
multilinguality:
|
11 |
+
- monolingual
|
12 |
+
size_categories:
|
13 |
+
- 10K<n<100K
|
14 |
+
source_datasets:
|
15 |
+
- original
|
16 |
+
task_categories:
|
17 |
+
- question-answering
|
18 |
+
task_ids:
|
19 |
+
- abstractive-qa
|
20 |
+
---
|
21 |
+
|
22 |
+
# Dataset Card for Narrative QA Manual
|
23 |
+
|
24 |
+
## Table of Contents
|
25 |
+
- [Dataset Description](#dataset-description)
|
26 |
+
- [Dataset Summary](#dataset-summary)
|
27 |
+
- [Supported Tasks](#supported-tasks-and-leaderboards)
|
28 |
+
- [Languages](#languages)
|
29 |
+
- [Dataset Structure](#dataset-structure)
|
30 |
+
- [Data Instances](#data-instances)
|
31 |
+
- [Data Fields](#data-fields)
|
32 |
+
- [Data Splits](#data-splits)
|
33 |
+
- [Dataset Creation](#dataset-creation)
|
34 |
+
- [Curation Rationale](#curation-rationale)
|
35 |
+
- [Source Data](#source-data)
|
36 |
+
- [Annotations](#annotations)
|
37 |
+
- [Personal and Sensitive Information](#personal-and-sensitive-information)
|
38 |
+
- [Considerations for Using the Data](#considerations-for-using-the-data)
|
39 |
+
- [Social Impact of Dataset](#social-impact-of-dataset)
|
40 |
+
- [Discussion of Biases](#discussion-of-biases)
|
41 |
+
- [Other Known Limitations](#other-known-limitations)
|
42 |
+
- [Additional Information](#additional-information)
|
43 |
+
- [Dataset Curators](#dataset-curators)
|
44 |
+
- [Licensing Information](#licensing-information)
|
45 |
+
- [Citation Information](#citation-information)
|
46 |
+
- [Contributions](#contributions)
|
47 |
+
|
48 |
+
## Dataset Description
|
49 |
+
|
50 |
+
- **Homepage:** [NarrativeQA Homepage](https://deepmind.com/research/open-source/narrativeqa)
|
51 |
+
- **Repository:** [NarrativeQA Repo](https://github.com/deepmind/narrativeqa)
|
52 |
+
- **Paper:** [The NarrativeQA Reading Comprehension Challenge](https://arxiv.org/pdf/1712.07040.pdf)
|
53 |
+
- **Leaderboard:**
|
54 |
+
- **Point of Contact:** [Tomáš Kočiský](mailto:tkocisky@google.com) [Jonathan Schwarz](mailto:schwarzjn@google.com) [Phil Blunsom](pblunsom@google.com) [Chris Dyer](cdyer@google.com) [Karl Moritz Hermann](mailto:kmh@google.com) [Gábor Melis](mailto:melisgl@google.com) [Edward Grefenstette](mailto:etg@google.com)
|
55 |
+
|
56 |
+
### Dataset Summary
|
57 |
+
|
58 |
+
NarrativeQA Manual is an English-language dataset of stories and corresponding questions designed to test reading comprehension, especially on long documents. THIS DATASET REQUIRES A MANUALLY DOWNLOADED FILE! Because of a script in the original repository which downloads the stories from original URLs everytime, the links are sometimes broken or invalid. Therefore, you need to manually download the stories for this dataset using the script provided by the authors (https://github.com/deepmind/narrativeqa/blob/master/download_stories.sh). Running the shell script creates a folder named "tmp" in the root directory and downloads the stories there. This folder containing the stories can be used to load the dataset via `datasets.load_dataset("narrativeqa_manual", data_dir="<path/to/folder>")`.
|
59 |
+
|
60 |
+
### Supported Tasks and Leaderboards
|
61 |
+
|
62 |
+
The dataset is used to test reading comprehension. There are 2 tasks proposed in the paper: "summaries only" and "stories only", depending on whether the human-generated summary or the full story text is used to answer the question.
|
63 |
+
|
64 |
+
### Languages
|
65 |
+
|
66 |
+
English
|
67 |
+
|
68 |
+
## Dataset Structure
|
69 |
+
|
70 |
+
### Data Instances
|
71 |
+
|
72 |
+
A typical data point consists of a question and answer pair along with a summary/story which can be used to answer the question. Additional information such as the url, word count, wikipedia page, are also provided.
|
73 |
+
|
74 |
+
A typical example looks like this:
|
75 |
+
```
|
76 |
+
{
|
77 |
+
"document": {
|
78 |
+
"id": "23jncj2n3534563110",
|
79 |
+
"kind": "movie",
|
80 |
+
"url": "https://www.imsdb.com/Movie%20Scripts/Name%20of%20Movie.html",
|
81 |
+
"file_size": 80473,
|
82 |
+
"word_count": 41000,
|
83 |
+
"start": "MOVIE screenplay by",
|
84 |
+
"end": ". THE END",
|
85 |
+
"summary": {
|
86 |
+
"text": "Joe Bloggs begins his journey exploring...",
|
87 |
+
"tokens": ["Joe", "Bloggs", "begins", "his", "journey", "exploring",...],
|
88 |
+
"url": "http://en.wikipedia.org/wiki/Name_of_Movie",
|
89 |
+
"title": "Name of Movie (film)"
|
90 |
+
},
|
91 |
+
"text": "MOVIE screenplay by John Doe\nSCENE 1..."
|
92 |
+
},
|
93 |
+
"question": {
|
94 |
+
"text": "Where does Joe Bloggs live?",
|
95 |
+
"tokens": ["Where", "does", "Joe", "Bloggs", "live", "?"],
|
96 |
+
},
|
97 |
+
"answers": [
|
98 |
+
{"text": "At home", "tokens": ["At", "home"]},
|
99 |
+
{"text": "His house", "tokens": ["His", "house"]}
|
100 |
+
]
|
101 |
+
}
|
102 |
+
```
|
103 |
+
|
104 |
+
### Data Fields
|
105 |
+
|
106 |
+
- `document.id` - Unique ID for the story.
|
107 |
+
- `document.kind` - "movie" or "gutenberg" depending on the source of the story.
|
108 |
+
- `document.url` - The URL where the story was downloaded from.
|
109 |
+
- `document.file_size` - File size (in bytes) of the story.
|
110 |
+
- `document.word_count` - Number of tokens in the story.
|
111 |
+
- `document.start` - First 3 tokens of the story. Used for verifying the story hasn't been modified.
|
112 |
+
- `document.end` - Last 3 tokens of the story. Used for verifying the story hasn't been modified.
|
113 |
+
- `document.summary.text` - Text of the wikipedia summary of the story.
|
114 |
+
- `document.summary.tokens` - Tokenized version of `document.summary.text`.
|
115 |
+
- `document.summary.url` - Wikipedia URL of the summary.
|
116 |
+
- `document.summary.title` - Wikipedia Title of the summary.
|
117 |
+
- `question` - `{"text":"...", "tokens":[...]}` for the question about the story.
|
118 |
+
- `answers` - List of `{"text":"...", "tokens":[...]}` for valid answers for the question.
|
119 |
+
|
120 |
+
### Data Splits
|
121 |
+
|
122 |
+
The data is split into training, valiudation, and test sets based on story (i.e. the same story cannot appear in more than one split):
|
123 |
+
|
124 |
+
| Train | Valid | Test |
|
125 |
+
| ------ | ----- | ----- |
|
126 |
+
| 32747 | 3461 | 10557 |
|
127 |
+
|
128 |
+
## Dataset Creation
|
129 |
+
|
130 |
+
### Curation Rationale
|
131 |
+
|
132 |
+
[More Information Needed]
|
133 |
+
|
134 |
+
### Source Data
|
135 |
+
|
136 |
+
#### Initial Data Collection and Normalization
|
137 |
+
Stories and movies scripts were downloaded from [Project Gutenburg](https://www.gutenberg.org) and a range of movie script repositories (mainly [imsdb](http://www.imsdb.com)).
|
138 |
+
|
139 |
+
#### Who are the source language producers?
|
140 |
+
|
141 |
+
The language producers are authors of the stories and scripts as well as Amazon Turk workers for the questions.
|
142 |
+
|
143 |
+
### Annotations
|
144 |
+
|
145 |
+
#### Annotation process
|
146 |
+
|
147 |
+
Amazon Turk Workers were provided with human written summaries of the stories (To make the annotation tractable and to lead annotators towards asking non-localized questions). Stories were matched with plot summaries from Wikipedia using titles and verified the matching with help from human annotators. The annotators were asked to determine if both the story and the summary refer to a movie or a book (as some books are made into movies), or if they are the same part in a series produced in the same year. Annotators on Amazon Mechanical Turk were instructed to write 10 question–answer pairs each based solely on a given summary. Annotators were instructed to imagine that they are writing questions to test students who have read the full stories but not the summaries. We required questions that are specific enough, given the length and complexity of the narratives, and to provide adiverse set of questions about characters, events, why this happened, and so on. Annotators were encouraged to use their own words and we prevented them from copying. We asked for answers that are grammatical, complete sentences, and explicitly allowed short answers (one word, or a few-word phrase, or ashort sentence) as we think that answering with a full sentence is frequently perceived as artificial when asking about factual information. Annotators were asked to avoid extra, unnecessary information in the question or the answer, and to avoid yes/no questions or questions about the author or the actors.
|
148 |
+
|
149 |
+
#### Who are the annotators?
|
150 |
+
|
151 |
+
Amazon Mechanical Turk workers.
|
152 |
+
|
153 |
+
### Personal and Sensitive Information
|
154 |
+
|
155 |
+
None
|
156 |
+
|
157 |
+
## Considerations for Using the Data
|
158 |
+
|
159 |
+
### Social Impact of Dataset
|
160 |
+
|
161 |
+
[More Information Needed]
|
162 |
+
|
163 |
+
### Discussion of Biases
|
164 |
+
|
165 |
+
[More Information Needed]
|
166 |
+
|
167 |
+
### Other Known Limitations
|
168 |
+
|
169 |
+
[More Information Needed]
|
170 |
+
|
171 |
+
## Additional Information
|
172 |
+
|
173 |
+
### Dataset Curators
|
174 |
+
|
175 |
+
[More Information Needed]
|
176 |
+
|
177 |
+
### Licensing Information
|
178 |
+
|
179 |
+
The dataset is released under a [Apache-2.0 License](https://github.com/deepmind/narrativeqa/blob/master/LICENSE).
|
180 |
+
|
181 |
+
### Citation Information
|
182 |
+
|
183 |
+
```
|
184 |
+
@article{narrativeqa,
|
185 |
+
author = {Tom\'a\v s Ko\v cisk\'y and Jonathan Schwarz and Phil Blunsom and
|
186 |
+
Chris Dyer and Karl Moritz Hermann and G\'abor Melis and
|
187 |
+
Edward Grefenstette},
|
188 |
+
title = {The {NarrativeQA} Reading Comprehension Challenge},
|
189 |
+
journal = {Transactions of the Association for Computational Linguistics},
|
190 |
+
url = {https://TBD},
|
191 |
+
volume = {TBD},
|
192 |
+
year = {2018},
|
193 |
+
pages = {TBD},
|
194 |
+
}
|
195 |
+
```
|
196 |
+
|
197 |
+
### Contributions
|
198 |
+
|
199 |
+
Thanks to [@rsanjaykamath](https://github.com/rsanjaykamath) for adding this dataset.
|
dataset_infos.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"default": {"description": "The Narrative QA Manual dataset is a reading comprehension dataset, in which the reader must answer questions about stories by reading entire books or movie scripts. The QA tasks are designed so that successfully answering their questions requires understanding the underlying narrative rather than relying on shallow pattern matching or salience.\\\nTHIS DATASET REQUIRES A MANUALLY DOWNLOADED FILE! Because of a script in the original repository which downloads the stories from original URLs everytime, The links are sometimes broken or invalid. Therefore, you need to manually download the stories for this dataset using the script provided by the authors (https://github.com/deepmind/narrativeqa/blob/master/download_stories.sh). Running the shell script creates a folder named \"tmp\" in the root directory and downloads the stories there. This folder containing the storiescan be used to load the dataset via `datasets.load_dataset(\"narrativeqa_manual\", data_dir=\"<path/to/folder>\")`. ", "citation": "@article{kovcisky2018narrativeqa,\n title={The narrativeqa reading comprehension challenge},\n author={Ko{\u000b{c}}isk{'y}, Tom{'a}{\u000b{s}} and Schwarz, Jonathan and Blunsom, Phil and Dyer, Chris and Hermann, Karl Moritz and Melis, G{'a}bor and Grefenstette, Edward},\n journal={Transactions of the Association for Computational Linguistics},\n volume={6},\n pages={317--328},\n year={2018},\n publisher={MIT Press}\n}\n", "homepage": "https://deepmind.com/research/publications/narrativeqa-reading-comprehension-challenge", "license": "https://github.com/deepmind/narrativeqa/blob/master/LICENSE", "features": {"document": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "kind": {"dtype": "string", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "file_size": {"dtype": "int32", "id": null, "_type": "Value"}, "word_count": {"dtype": "int32", "id": null, "_type": "Value"}, "start": {"dtype": "string", "id": null, "_type": "Value"}, "end": {"dtype": "string", "id": null, "_type": "Value"}, "summary": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "tokens": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "title": {"dtype": "string", "id": null, "_type": "Value"}}, "text": {"dtype": "string", "id": null, "_type": "Value"}}, "question": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "tokens": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}}, "answers": [{"text": {"dtype": "string", "id": null, "_type": "Value"}, "tokens": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}}]}, "post_processed": null, "supervised_keys": null, "builder_name": "narrativeqa_manual", "config_name": "default", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 9115940054, "num_examples": 32747, "dataset_name": "narrativeqa_manual"}, "test": {"name": "test", "num_bytes": 2911702563, "num_examples": 10557, "dataset_name": "narrativeqa_manual"}, "validation": {"name": "validation", "num_bytes": 968994186, "num_examples": 3461, "dataset_name": "narrativeqa_manual"}}, "download_checksums": {"https://raw.githubusercontent.com/deepmind/narrativeqa/master/documents.csv": {"num_bytes": 341683, "checksum": "6dffa4cc0b5c9963fe3a097c87f04aa7767da36627500cc7a0d69e2405e1144a"}, "https://raw.githubusercontent.com/deepmind/narrativeqa/master/third_party/wikipedia/summaries.csv": {"num_bytes": 10821085, "checksum": "87d12b849219015c4fe717e95c707671d802c181145e8ba2acc11dff23ea7c75"}, "https://raw.githubusercontent.com/deepmind/narrativeqa/master/qaps.csv": {"num_bytes": 11475505, "checksum": "990b02af0b5280de210f0e6b80f43b3fab80dc6de630c4d5059a1b7131c26e38"}}, "download_size": 22638273, "post_processing_size": null, "dataset_size": 12996636803, "size_in_bytes": 13019275076}}
|
dummy/1.0.0/dummy_data.zip
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:23516fde7bc6e3cdba094b1e0d339d1836818f8adec6335a27c8e0ee4f498167
|
3 |
+
size 4698
|
narrativeqa_manual.py
ADDED
@@ -0,0 +1,220 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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 |
+
"""NarrativeQA Reading Comprehension Challenge"""
|
16 |
+
|
17 |
+
from __future__ import absolute_import, division, print_function
|
18 |
+
|
19 |
+
import csv
|
20 |
+
import os
|
21 |
+
from os import listdir
|
22 |
+
from os.path import isfile, join
|
23 |
+
|
24 |
+
import datasets
|
25 |
+
|
26 |
+
|
27 |
+
_CITATION = """\
|
28 |
+
@article{kovcisky2018narrativeqa,
|
29 |
+
title={The narrativeqa reading comprehension challenge},
|
30 |
+
author={Ko{\v{c}}isk{\'y}, Tom{\'a}{\v{s}} and Schwarz, Jonathan and Blunsom, Phil and Dyer, Chris and Hermann, Karl Moritz and Melis, G{\'a}bor and Grefenstette, Edward},
|
31 |
+
journal={Transactions of the Association for Computational Linguistics},
|
32 |
+
volume={6},
|
33 |
+
pages={317--328},
|
34 |
+
year={2018},
|
35 |
+
publisher={MIT Press}
|
36 |
+
}
|
37 |
+
"""
|
38 |
+
|
39 |
+
|
40 |
+
_DESCRIPTION = """\
|
41 |
+
The Narrative QA Manual dataset is a reading comprehension \
|
42 |
+
dataset, in which the reader must answer questions about stories \
|
43 |
+
by reading entire books or movie scripts. \
|
44 |
+
The QA tasks are designed so that successfully answering their questions \
|
45 |
+
requires understanding the underlying narrative rather than \
|
46 |
+
relying on shallow pattern matching or salience.\\
|
47 |
+
THIS DATASET REQUIRES A MANUALLY DOWNLOADED FILE! \
|
48 |
+
Because of a script in the original repository which downloads the stories from original URLs everytime, \
|
49 |
+
The links are sometimes broken or invalid. \
|
50 |
+
Therefore, you need to manually download the stories for this dataset using the script provided by the authors \
|
51 |
+
(https://github.com/deepmind/narrativeqa/blob/master/download_stories.sh). Running the shell script creates a folder named "tmp" \
|
52 |
+
in the root directory and downloads the stories there. This folder containing the stories\
|
53 |
+
can be used to load the dataset via `datasets.load_dataset("narrativeqa_manual", data_dir="<path/to/folder>")`. """
|
54 |
+
|
55 |
+
|
56 |
+
_HOMEPAGE = "https://deepmind.com/research/publications/narrativeqa-reading-comprehension-challenge"
|
57 |
+
_LICENSE = "https://github.com/deepmind/narrativeqa/blob/master/LICENSE"
|
58 |
+
|
59 |
+
|
60 |
+
# The HuggingFace dataset library don't host the datasets but only point to the original files
|
61 |
+
# This can be an arbitrary nested dict/list of URLs (see below in `_split_generators` method)
|
62 |
+
_URL = "https://github.com/deepmind/narrativeqa"
|
63 |
+
_URLS = {
|
64 |
+
"documents": "https://raw.githubusercontent.com/deepmind/narrativeqa/master/documents.csv",
|
65 |
+
"summaries": "https://raw.githubusercontent.com/deepmind/narrativeqa/master/third_party/wikipedia/summaries.csv",
|
66 |
+
"qaps": "https://raw.githubusercontent.com/deepmind/narrativeqa/master/qaps.csv",
|
67 |
+
}
|
68 |
+
|
69 |
+
|
70 |
+
class NarrativeqaManual(datasets.GeneratorBasedBuilder):
|
71 |
+
"""The NarrativeQA Manual dataset"""
|
72 |
+
|
73 |
+
VERSION = datasets.Version("1.0.0")
|
74 |
+
|
75 |
+
@property
|
76 |
+
def manual_download_instructions(self):
|
77 |
+
return """ You need to manually download the stories for this dataset using the script provided by the authors \
|
78 |
+
(https://github.com/deepmind/narrativeqa/blob/master/download_stories.sh). Running the shell script creates a folder named "tmp"\
|
79 |
+
in the root directory and downloads the stories there. This folder containing the stories\
|
80 |
+
can be used to load the dataset via `datasets.load_dataset("narrativeqa_manual", data_dir="<path/to/folder>")."""
|
81 |
+
|
82 |
+
def _info(self):
|
83 |
+
|
84 |
+
return datasets.DatasetInfo(
|
85 |
+
description=_DESCRIPTION,
|
86 |
+
features=datasets.Features(
|
87 |
+
{
|
88 |
+
"document": {
|
89 |
+
"id": datasets.Value("string"),
|
90 |
+
"kind": datasets.Value("string"),
|
91 |
+
"url": datasets.Value("string"),
|
92 |
+
"file_size": datasets.Value("int32"),
|
93 |
+
"word_count": datasets.Value("int32"),
|
94 |
+
"start": datasets.Value("string"),
|
95 |
+
"end": datasets.Value("string"),
|
96 |
+
"summary": {
|
97 |
+
"text": datasets.Value("string"),
|
98 |
+
"tokens": datasets.features.Sequence(datasets.Value("string")),
|
99 |
+
"url": datasets.Value("string"),
|
100 |
+
"title": datasets.Value("string"),
|
101 |
+
},
|
102 |
+
"text": datasets.Value("string"),
|
103 |
+
},
|
104 |
+
"question": {
|
105 |
+
"text": datasets.Value("string"),
|
106 |
+
"tokens": datasets.features.Sequence(datasets.Value("string")),
|
107 |
+
},
|
108 |
+
"answers": [
|
109 |
+
{
|
110 |
+
"text": datasets.Value("string"),
|
111 |
+
"tokens": datasets.features.Sequence(datasets.Value("string")),
|
112 |
+
}
|
113 |
+
],
|
114 |
+
}
|
115 |
+
),
|
116 |
+
supervised_keys=None,
|
117 |
+
homepage=_HOMEPAGE,
|
118 |
+
license=_LICENSE,
|
119 |
+
citation=_CITATION,
|
120 |
+
)
|
121 |
+
|
122 |
+
def _split_generators(self, dl_manager):
|
123 |
+
"""Returns SplitGenerators."""
|
124 |
+
data_dir = dl_manager.download_and_extract(_URLS)
|
125 |
+
manual_dir = os.path.abspath(os.path.expanduser(dl_manager.manual_dir))
|
126 |
+
|
127 |
+
if not os.path.exists(manual_dir):
|
128 |
+
raise FileNotFoundError(
|
129 |
+
"{} does not exist. Make sure you insert a manual dir via `datasets.load_dataset('narrativeqa_manual', data_dir=...)` that includes the stories downloaded from the original repository. Manual download instructions: {}".format(
|
130 |
+
manual_dir, self.manual_download_instructions
|
131 |
+
)
|
132 |
+
)
|
133 |
+
|
134 |
+
return [
|
135 |
+
datasets.SplitGenerator(
|
136 |
+
name=datasets.Split.TRAIN,
|
137 |
+
gen_kwargs={
|
138 |
+
"data_dir": data_dir,
|
139 |
+
"manual_dir": manual_dir,
|
140 |
+
"split": "train",
|
141 |
+
},
|
142 |
+
),
|
143 |
+
datasets.SplitGenerator(
|
144 |
+
name=datasets.Split.TEST,
|
145 |
+
gen_kwargs={
|
146 |
+
"data_dir": data_dir,
|
147 |
+
"manual_dir": manual_dir,
|
148 |
+
"split": "test",
|
149 |
+
},
|
150 |
+
),
|
151 |
+
datasets.SplitGenerator(
|
152 |
+
name=datasets.Split.VALIDATION,
|
153 |
+
gen_kwargs={
|
154 |
+
"data_dir": data_dir,
|
155 |
+
"manual_dir": manual_dir,
|
156 |
+
"split": "valid",
|
157 |
+
},
|
158 |
+
),
|
159 |
+
]
|
160 |
+
|
161 |
+
def _generate_examples(self, data_dir, manual_dir, split):
|
162 |
+
""" Yields examples. """
|
163 |
+
|
164 |
+
documents = {}
|
165 |
+
with open(data_dir["documents"], encoding="utf-8") as f:
|
166 |
+
reader = csv.DictReader(f)
|
167 |
+
for row in reader:
|
168 |
+
if row["set"] != split:
|
169 |
+
continue
|
170 |
+
documents[row["document_id"]] = row
|
171 |
+
|
172 |
+
summaries = {}
|
173 |
+
with open(data_dir["summaries"], encoding="utf-8") as f:
|
174 |
+
reader = csv.DictReader(f)
|
175 |
+
for row in reader:
|
176 |
+
if row["set"] != split:
|
177 |
+
continue
|
178 |
+
summaries[row["document_id"]] = row
|
179 |
+
|
180 |
+
onlyfiles = [f for f in listdir(manual_dir) if isfile(join(manual_dir, f))]
|
181 |
+
story_texts = {}
|
182 |
+
for i in onlyfiles:
|
183 |
+
if "content" in i:
|
184 |
+
with open(os.path.join(manual_dir, i), "r", encoding="utf-8", errors="ignore") as f:
|
185 |
+
text = f.read()
|
186 |
+
story_texts[i.split(".")[0]] = text
|
187 |
+
|
188 |
+
with open(data_dir["qaps"], encoding="utf-8") as f:
|
189 |
+
reader = csv.DictReader(f)
|
190 |
+
for id_, row in enumerate(reader):
|
191 |
+
if row["set"] != split:
|
192 |
+
continue
|
193 |
+
document_id = row["document_id"]
|
194 |
+
document = documents[document_id]
|
195 |
+
summary = summaries[document_id]
|
196 |
+
full_text = story_texts[document_id]
|
197 |
+
res = {
|
198 |
+
"document": {
|
199 |
+
"id": document["document_id"],
|
200 |
+
"kind": document["kind"],
|
201 |
+
"url": document["story_url"],
|
202 |
+
"file_size": document["story_file_size"],
|
203 |
+
"word_count": document["story_word_count"],
|
204 |
+
"start": document["story_start"],
|
205 |
+
"end": document["story_end"],
|
206 |
+
"summary": {
|
207 |
+
"text": summary["summary"],
|
208 |
+
"tokens": summary["summary_tokenized"].split(),
|
209 |
+
"url": document["wiki_url"],
|
210 |
+
"title": document["wiki_title"],
|
211 |
+
},
|
212 |
+
"text": full_text,
|
213 |
+
},
|
214 |
+
"question": {"text": row["question"], "tokens": row["question_tokenized"].split()},
|
215 |
+
"answers": [
|
216 |
+
{"text": row["answer1"], "tokens": row["answer1_tokenized"].split()},
|
217 |
+
{"text": row["answer2"], "tokens": row["answer2_tokenized"].split()},
|
218 |
+
],
|
219 |
+
}
|
220 |
+
yield id_, res
|