Update README.md
Browse files
README.md
CHANGED
@@ -54,7 +54,7 @@ pretty_name: Quranic Audio Dataset - Crowdsourced and Labeled Recitation from No
|
|
54 |
|
55 |
# Dataset Card for Quranic Audio Dataset : Crowdsourced and Labeled Recitation from Non-Arabic Speakers
|
56 |
|
57 |
-
|
58 |
|
59 |
We explore the possibility of crowdsourcing a carefully annotated Quranic dataset, on top of which AI models can be built to simplify the learning process.
|
60 |
In particular, we use the volunteer-based crowdsourcing genre and implement a crowdsourcing API to gather audio assets.
|
@@ -63,17 +63,71 @@ As a result, we have collected around 7000 Quranic recitations from a pool of 12
|
|
63 |
We have achieved a crowd accuracy of 0.77, an inter-rater agreement of 0.63 between the annotators, and 0.89 between the labels assigned by the algorithm and the expert judgments.
|
64 |
|
65 |
## How to use
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
66 |
|
67 |
## Dataset Structure
|
68 |
|
69 |
### Data Instances
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
70 |
|
71 |
### Data Fields
|
72 |
|
73 |
### Citation Information
|
74 |
|
75 |
```
|
76 |
-
@inproceedings{
|
77 |
author = {Salameh, R., Mdfaa, M. A., Askarbekuly, N., & Mazzara, M.},
|
78 |
title = {Quranic Audio Dataset: Crowdsourced and Labeled Recitation from Non-Arabic Speakers},
|
79 |
year = 2024,
|
@@ -81,8 +135,6 @@ We have achieved a crowd accuracy of 0.77, an inter-rater agreement of 0.63 betw
|
|
81 |
eprinttype = {arxiv},
|
82 |
eprintclass = {cs.SD},
|
83 |
url = {https://arxiv.org/abs/2405.02675},
|
84 |
-
language = {english}
|
85 |
-
booktitle = {},
|
86 |
-
pages = {}
|
87 |
}
|
88 |
```
|
|
|
54 |
|
55 |
# Dataset Card for Quranic Audio Dataset : Crowdsourced and Labeled Recitation from Non-Arabic Speakers
|
56 |
|
57 |
+
## Dataset Summary
|
58 |
|
59 |
We explore the possibility of crowdsourcing a carefully annotated Quranic dataset, on top of which AI models can be built to simplify the learning process.
|
60 |
In particular, we use the volunteer-based crowdsourcing genre and implement a crowdsourcing API to gather audio assets.
|
|
|
63 |
We have achieved a crowd accuracy of 0.77, an inter-rater agreement of 0.63 between the annotators, and 0.89 between the labels assigned by the algorithm and the expert judgments.
|
64 |
|
65 |
## How to use
|
66 |
+
The dataset can be downloaded using the `load_dataset` function from `datasets` library.
|
67 |
+
|
68 |
+
```python
|
69 |
+
!pip install datasets
|
70 |
+
```
|
71 |
+
|
72 |
+
```python
|
73 |
+
from datasets import load_dataset
|
74 |
+
|
75 |
+
quranic_dataset = load_dataset("RetaSy/quranic_audio_dataset")
|
76 |
+
|
77 |
+
print(quranic_dataset)
|
78 |
+
```
|
79 |
|
80 |
## Dataset Structure
|
81 |
|
82 |
### Data Instances
|
83 |
+
```python
|
84 |
+
{
|
85 |
+
'audio': {
|
86 |
+
'path': '0058a4f7-6a3a-4665-b43b-d6f67fd14dbf.wav',
|
87 |
+
'array': array([0.00000000e+00, 0.00000000e+00, 3.05175781e-05, ...,9.15527344e-05, 0.00000000e+00, 1.83105469e-04]),
|
88 |
+
'sampling_rate': 16000
|
89 |
+
},
|
90 |
+
'Surah': 'Al-Faatihah',
|
91 |
+
'Aya': 'أَعُوذُ بِاللَّهِ مِنَ الشَّيْطَانِ الرَّجِيْمِ',
|
92 |
+
'duration_ms': 3520,
|
93 |
+
'create_date': '2023-03-15T19:57:35.027430+03:00',
|
94 |
+
'golden': False,
|
95 |
+
'final_label': 'in_correct',
|
96 |
+
'reciter_id': 'ef1ada15-e225-4155-a81c-fc461d940a6d',
|
97 |
+
'reciter_country': 'AT',
|
98 |
+
'reciter_gender': 'female',
|
99 |
+
'reciter_age': 'Unknown',
|
100 |
+
'reciter_qiraah': 'hafs',
|
101 |
+
'judgments_num': 3,
|
102 |
+
'annotation_metadata': '{
|
103 |
+
"label_1": "in_correct",
|
104 |
+
"annotator1_id": "1",
|
105 |
+
"annotator1_SCT": "257",
|
106 |
+
"annotator1_MCC": "0.87",
|
107 |
+
"annotator1_ACC": "0.92",
|
108 |
+
"annotator1_F1": "0.91",
|
109 |
+
"label_2": "correct",
|
110 |
+
"annotator2_id": "10",
|
111 |
+
"annotator2_SCT": "21",
|
112 |
+
"annotator2_MCC": "0.52",
|
113 |
+
"annotator2_ACC": "0.57",
|
114 |
+
"annotator2_F1": "0.55",
|
115 |
+
"label_3": "in_correct",
|
116 |
+
"annotator3_id": "19",
|
117 |
+
"annotator3_SCT": "12",
|
118 |
+
"annotator3_MCC": "0.75",
|
119 |
+
"annotator3_ACC": "0.83",
|
120 |
+
"annotator3_F1": "0.78"
|
121 |
+
}'
|
122 |
+
}
|
123 |
+
```
|
124 |
|
125 |
### Data Fields
|
126 |
|
127 |
### Citation Information
|
128 |
|
129 |
```
|
130 |
+
@inproceedings{quran_audio_dataset:2024,
|
131 |
author = {Salameh, R., Mdfaa, M. A., Askarbekuly, N., & Mazzara, M.},
|
132 |
title = {Quranic Audio Dataset: Crowdsourced and Labeled Recitation from Non-Arabic Speakers},
|
133 |
year = 2024,
|
|
|
135 |
eprinttype = {arxiv},
|
136 |
eprintclass = {cs.SD},
|
137 |
url = {https://arxiv.org/abs/2405.02675},
|
138 |
+
language = {english}
|
|
|
|
|
139 |
}
|
140 |
```
|