Datasets:
michal-stefanik
commited on
Update README.md
Browse files
README.md
CHANGED
@@ -8,7 +8,7 @@ dataset_info:
|
|
8 |
dtype: string
|
9 |
- name: sentence
|
10 |
dtype: string
|
11 |
-
- name: y
|
12 |
dtype: string
|
13 |
- name: confidence
|
14 |
dtype: string
|
@@ -66,7 +66,7 @@ dataset_info:
|
|
66 |
dtype: string
|
67 |
- name: sentence
|
68 |
dtype: string
|
69 |
-
- name: y
|
70 |
dtype: string
|
71 |
- name: confidence
|
72 |
dtype: string
|
@@ -105,4 +105,83 @@ configs:
|
|
105 |
path: en/validation-*
|
106 |
- split: test
|
107 |
path: en/test-*
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
108 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
8 |
dtype: string
|
9 |
- name: sentence
|
10 |
dtype: string
|
11 |
+
- name: 'y'
|
12 |
dtype: string
|
13 |
- name: confidence
|
14 |
dtype: string
|
|
|
66 |
dtype: string
|
67 |
- name: sentence
|
68 |
dtype: string
|
69 |
+
- name: 'y'
|
70 |
dtype: string
|
71 |
- name: confidence
|
72 |
dtype: string
|
|
|
105 |
path: en/validation-*
|
106 |
- split: test
|
107 |
path: en/test-*
|
108 |
+
license: apache-2.0
|
109 |
+
task_categories:
|
110 |
+
- text-classification
|
111 |
+
language:
|
112 |
+
- en
|
113 |
+
- cs
|
114 |
+
tags:
|
115 |
+
- education
|
116 |
+
pretty_name: Czech-English Reflective Dataset (CEReD)
|
117 |
---
|
118 |
+
|
119 |
+
# Dataset Card for Czech-English Reflective Dataset (CEReD)
|
120 |
+
|
121 |
+
This directory contains an anonymized data set of separated sentences and original reflective journals collected within the Reflection Classification project: https://github.com/EduMUNI/reflection-classification
|
122 |
+
See the project repository for more details and the [corresponding paper](https://rdcu.be/cUWGY) for more details on data curation methodology.
|
123 |
+
|
124 |
+
The data is available in in two types of subsets:
|
125 |
+
|
126 |
+
1. The `cs-orig-diaries` contains the full texts of the original reflection journals together with the authors' responses to our questionnaire.
|
127 |
+
Entries in this split contain the following attributes:
|
128 |
+
* `id`: unique reflective diary id
|
129 |
+
* `person_id`: synthetic id of a creator of the diary
|
130 |
+
* `subject`: subject that the reflective diary concern
|
131 |
+
* `ordering`: relative rank of the diary relative to other diaries of the same author
|
132 |
+
* `Q1`: Teacher evaluation: "Student treated the leading teacher with respect."
|
133 |
+
* `Q2`: Teacher evaluation: "Student took responsibility in a preparation for practice."
|
134 |
+
* `Q3`: Teacher evaluation: "Student discussed specific means of their further development."
|
135 |
+
* `Q4`: Teacher evaluation: "Student actively asked me for a support, feedback, reflection."
|
136 |
+
* `Q5`: Teacher evaluation: "Student actively reflected on their activity on practice."
|
137 |
+
* `Q6`: Teacher evaluation: "Student recognized the situation of the class and reacted to it with selected stragegy."
|
138 |
+
* `Q7`: Teacher evaluation: "Student shown interest in a situation in school, in general."
|
139 |
+
* `diary`: Text of the reflective diary
|
140 |
+
|
141 |
+
All questions `Q[1-7]` are part of the questionnaire
|
142 |
+
filled by the supervising teacher on the relevant practice.
|
143 |
+
The questionnaire concerned the performance evaluation of
|
144 |
+
the candidate teacher student, that authored the reflective diary.
|
145 |
+
|
146 |
+
3. Subsets `cs` and `en` contain separate sentences that can be used for training a classifier, in
|
147 |
+
selected language: original: Czech (`cs`) or translated: English (`en`).
|
148 |
+
Sentences are divided into train, validation (val) and test set.
|
149 |
+
This split can be used to evaluate the classifier on the same
|
150 |
+
data, as we did, hence it allows for comparability of
|
151 |
+
the results.
|
152 |
+
Again, the tab-separated `sentences.tsv` files contain following
|
153 |
+
attributes:
|
154 |
+
* `idx`: unique sentence id
|
155 |
+
* `context`: textual context surrounding the classified sentence
|
156 |
+
* `sentence`: text of the classified sentence
|
157 |
+
* `y`: target category of the sentence, that annotators agreed upon
|
158 |
+
* `confidence`: confidence, or typicality of the sentence in its assigned category. Annotators were asked: "How typical is this sentence for the picked category?"
|
159 |
+
* `y_requires_context`: whether annotators needed to look at the context, when selecting a category.
|
160 |
+
|
161 |
+
# Citation
|
162 |
+
|
163 |
+
For the data collection methodology:
|
164 |
+
```bibtex
|
165 |
+
@Article{Nehyba2022applications,
|
166 |
+
author={Nehyba, Jan and {\v{S}}tef{\'a}nik, Michal},
|
167 |
+
title={Applications of deep language models for reflective writings},
|
168 |
+
journal={Education and Information Technologies},
|
169 |
+
year={2022},
|
170 |
+
month={Sep},
|
171 |
+
day={05},
|
172 |
+
issn={1573-7608},
|
173 |
+
doi={10.1007/s10639-022-11254-7},
|
174 |
+
url={https://doi.org/10.1007/s10639-022-11254-7}
|
175 |
+
}
|
176 |
+
```
|
177 |
+
|
178 |
+
For the dataset itself:
|
179 |
+
```bibtex
|
180 |
+
@misc{Stefanik2021CEReD,
|
181 |
+
title = {Czech and English Reflective Dataset ({CEReD})},
|
182 |
+
author = {{\v S}tef{\'a}nik, Michal and Nehyba, Jan},
|
183 |
+
url = {http://hdl.handle.net/11372/LRT-3573},
|
184 |
+
copyright = {Creative Commons - Attribution 4.0 International ({CC} {BY} 4.0)},
|
185 |
+
year = {2021}
|
186 |
+
}
|
187 |
+
```
|