Datasets:
Tasks:
Automatic Speech Recognition
Modalities:
Audio
Formats:
soundfolder
Languages:
Russian
Size:
< 1K
License:
Update README.md
Browse files
README.md
CHANGED
|
@@ -11,81 +11,176 @@ tags:
|
|
| 11 |
- speech
|
| 12 |
- call-center
|
| 13 |
- russian
|
| 14 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
---
|
| 16 |
|
| 17 |
-
-
|
| 18 |
-
|
| 19 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
|
| 21 |
-
## Sample
|
| 22 |
-
[▶️ Listen to sample (RU)](https://huggingface.co/datasets/MaratDV/russian-call-center-speech-ru/resolve/main/sample_ru.ogg)
|
| 23 |
-
<!-- Встроенный плеер -->
|
| 24 |
<audio controls preload="none" style="width:100%">
|
| 25 |
<source src="https://huggingface.co/datasets/MaratDV/russian-call-center-speech-ru/resolve/main/sample_ru.ogg" type="audio/ogg">
|
| 26 |
-
<!-- опционально: если добавишь MP3, будет запасной источник -->
|
| 27 |
-
<source src="https://huggingface.co/datasets/MaratDV/russian-call-center-speech-ru/resolve/main/sample_ru.mp3" type="audio/mpeg">
|
| 28 |
Your browser does not support the audio element.
|
| 29 |
</audio>
|
| 30 |
|
|
|
|
| 31 |
|
|
|
|
| 32 |
|
| 33 |
-
|
| 34 |
|
| 35 |
-
|
| 36 |
-
A large-scale dataset of real Russian call center recordings.
|
| 37 |
-
Telephone-quality conversations between customers and agents.
|
| 38 |
-
Useful for ASR, NLP, speech analytics, and training voicebots.
|
| 39 |
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
- Total duration: ~832 hours
|
| 43 |
-
- Format: MP3
|
| 44 |
-
- Channels: Mono (1 channel, mixed client and agent)
|
| 45 |
-
- Sample rate: 8000 Hz
|
| 46 |
-
- Bitrate: 32 kbps
|
| 47 |
-
- Metadata: Not available
|
| 48 |
|
| 49 |
-
|
| 50 |
-
Commercial license only. Redistribution not allowed.
|
| 51 |
|
| 52 |
-
|
| 53 |
-
📧 chinzad@gmail.com
|
| 54 |
-
💬 Telegram [@Marat_DV](https://t.me/Marat_DV)
|
| 55 |
|
| 56 |
-
-
|
| 57 |
|
| 58 |
-
|
| 59 |
|
| 60 |
-
|
| 61 |
-
Крупный датасет реальных записей колл-центров на русском языке.
|
| 62 |
-
Телефонное качество, разговоры «клиент–оператор».
|
| 63 |
-
Подходит для обучения систем ASR (распознавание речи), NLP, голосовых ассистентов и анализа диалогов.
|
| 64 |
|
| 65 |
-
|
| 66 |
-
- Язык: русский
|
| 67 |
-
- Общая продолжительность: ~832 часа
|
| 68 |
-
- Формат: MP3
|
| 69 |
-
- Каналы: моно (клиент и оператор в одном канале)
|
| 70 |
-
- Частота дискретизации: 8000 Гц
|
| 71 |
-
- Битрейт: 32 кбит/с
|
| 72 |
-
- Метаданные: отсутствуют
|
| 73 |
|
| 74 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 75 |
|
| 76 |
-
|
| 77 |
|
| 78 |
-
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 87 |
|
|
|
|
| 88 |
|
| 89 |
-
|
| 90 |
-
Только коммерческое использование. Перераспространение запрещено.
|
| 91 |
|
|
|
|
|
|
|
|
|
| 11 |
- speech
|
| 12 |
- call-center
|
| 13 |
- russian
|
| 14 |
+
- conversational-ai
|
| 15 |
+
- customer-service
|
| 16 |
+
pretty_name: Russian Call Center Speech Dataset - 832 Hours
|
| 17 |
+
size_categories:
|
| 18 |
+
- 1K<n<10K
|
| 19 |
---
|
| 20 |
|
| 21 |
+
# Russian Call Center Speech Dataset - 832 Hours
|
| 22 |
+
|
| 23 |
+
## Description
|
| 24 |
+
|
| 25 |
+
A large-scale dataset of real Russian call center and business-call recordings.
|
| 26 |
+
|
| 27 |
+
The dataset contains 9,215 unique audio recordings with a total duration of approximately 832.02 hours.
|
| 28 |
+
|
| 29 |
+
The corpus covers 22 business domains, including restaurant and food delivery services, e-commerce, recruitment, IT and software services, digital marketing, logistics, automotive sales, banking, healthcare, real estate, telecom, insurance, energy and other sectors.
|
| 30 |
+
|
| 31 |
+
It can be used for speech recognition research, speech analytics, conversational AI, dialogue systems, domain classification, and call center analytics.
|
| 32 |
+
|
| 33 |
+
## Dataset Summary
|
| 34 |
+
|
| 35 |
+
- Language: Russian
|
| 36 |
+
- Number of audio files: 9,215
|
| 37 |
+
- Total duration: ~832.02 hours
|
| 38 |
+
- Average recording duration: ~5.42 minutes
|
| 39 |
+
- Median recording duration: ~4.45 minutes
|
| 40 |
+
- Format: WAV
|
| 41 |
+
- Codec: PCM 16-bit (pcm_s16le)
|
| 42 |
+
- Channels: Mono (1 channel)
|
| 43 |
+
- Sample rate: 8,000 Hz
|
| 44 |
+
- Bit depth: 16-bit
|
| 45 |
+
- Bitrate: ~128 kbps
|
| 46 |
+
- Unique files by MD5: 9,215 / 9,215
|
| 47 |
+
|
| 48 |
+
## Domain Distribution
|
| 49 |
+
|
| 50 |
+
| Domain | Files | Hours | Share |
|
| 51 |
+
|---|---:|---:|---:|
|
| 52 |
+
| Restaurant / Food Delivery Services | 2,358 | 215.69 | 25.92% |
|
| 53 |
+
| E-commerce / Retail | 1,439 | 134.23 | 16.13% |
|
| 54 |
+
| Recruitment / Job Interviews | 1,342 | 126.18 | 15.17% |
|
| 55 |
+
| IT / Software Services | 938 | 86.18 | 10.36% |
|
| 56 |
+
| Online Directory / Digital Marketing | 945 | 84.54 | 10.16% |
|
| 57 |
+
| Logistics / Delivery | 361 | 28.49 | 3.43% |
|
| 58 |
+
| Automotive Sales | 257 | 21.44 | 2.58% |
|
| 59 |
+
| Certification / Business Services | 262 | 20.75 | 2.49% |
|
| 60 |
+
| Legal Services | 226 | 20.20 | 2.43% |
|
| 61 |
+
| Banking / Credit | 191 | 17.11 | 2.06% |
|
| 62 |
+
| General Business Services | 139 | 15.50 | 1.86% |
|
| 63 |
+
| Construction / Home Services | 139 | 12.97 | 1.56% |
|
| 64 |
+
| Healthcare | 136 | 10.72 | 1.29% |
|
| 65 |
+
| Real Estate | 115 | 9.35 | 1.12% |
|
| 66 |
+
| Manufacturing / Industrial | 83 | 6.42 | 0.77% |
|
| 67 |
+
| Travel / Hospitality | 83 | 6.27 | 0.75% |
|
| 68 |
+
| Education | 78 | 6.12 | 0.74% |
|
| 69 |
+
| Telecom | 58 | 4.73 | 0.57% |
|
| 70 |
+
| Insurance | 20 | 1.95 | 0.23% |
|
| 71 |
+
| Energy / Utilities | 27 | 1.67 | 0.20% |
|
| 72 |
+
| Government / Public Services | 9 | 0.93 | 0.11% |
|
| 73 |
+
| Security Services | 9 | 0.59 | 0.07% |
|
| 74 |
+
|
| 75 |
+
## Metadata
|
| 76 |
+
|
| 77 |
+
Metadata is available for the recordings and includes:
|
| 78 |
+
|
| 79 |
+
- recording ID
|
| 80 |
+
- domain
|
| 81 |
+
- duration
|
| 82 |
+
- filename
|
| 83 |
+
- MD5 hash
|
| 84 |
+
|
| 85 |
+
All 9,215 files are unique by MD5 hash.
|
| 86 |
+
|
| 87 |
+
## Potential Use Cases
|
| 88 |
+
|
| 89 |
+
- Automatic Speech Recognition research
|
| 90 |
+
- Speech-to-text systems
|
| 91 |
+
- Call center speech analytics
|
| 92 |
+
- Conversational AI
|
| 93 |
+
- Dialogue systems
|
| 94 |
+
- Domain and intent classification
|
| 95 |
+
- Customer-service analytics
|
| 96 |
+
- Business-call speech research
|
| 97 |
+
|
| 98 |
+
## Sample
|
| 99 |
+
|
| 100 |
+
[Listen to sample (RU)](https://huggingface.co/datasets/MaratDV/russian-call-center-speech-ru/resolve/main/sample_ru.ogg)
|
| 101 |
|
|
|
|
|
|
|
|
|
|
| 102 |
<audio controls preload="none" style="width:100%">
|
| 103 |
<source src="https://huggingface.co/datasets/MaratDV/russian-call-center-speech-ru/resolve/main/sample_ru.ogg" type="audio/ogg">
|
|
|
|
|
|
|
| 104 |
Your browser does not support the audio element.
|
| 105 |
</audio>
|
| 106 |
|
| 107 |
+
## License
|
| 108 |
|
| 109 |
+
Commercial license only.
|
| 110 |
|
| 111 |
+
Redistribution, resale, sublicensing, or public hosting of the full dataset is not permitted.
|
| 112 |
|
| 113 |
+
## Contact
|
|
|
|
|
|
|
|
|
|
| 114 |
|
| 115 |
+
Email: marat@menkroft.com
|
| 116 |
+
Telegram: [@Marat_DV](https://t.me/Marat_DV)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 117 |
|
| 118 |
+
---
|
|
|
|
| 119 |
|
| 120 |
+
# Описание на русском
|
|
|
|
|
|
|
| 121 |
|
| 122 |
+
Этот датасет содержит реальные записи разговоров колл-центров и деловых телефонных разговоров на русском языке.
|
| 123 |
|
| 124 |
+
Набор включает 9 215 уникальных аудиозаписей общей продолжительностью около 832,02 часа.
|
| 125 |
|
| 126 |
+
Датасет охватывает 22 бизнес-домена, включая рестораны и доставку еды, e-commerce, рекрутинг, IT, digital-маркетинг, логистику, продажу автомобилей, банковские услуги, здравоохранение, недвижимость, телекоммуникации, страхование, энергетику и другие направления.
|
|
|
|
|
|
|
|
|
|
| 127 |
|
| 128 |
+
## Основные характеристики
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 129 |
|
| 130 |
+
- Язык: русский
|
| 131 |
+
- Количество аудиофайлов: 9 215
|
| 132 |
+
- Общая продолжительность: ~832,02 часа
|
| 133 |
+
- Средняя длительность записи: ~5,42 минуты
|
| 134 |
+
- Медианная длительность: ~4,45 минуты
|
| 135 |
+
- Формат: WAV
|
| 136 |
+
- Кодек: PCM 16-bit (pcm_s16le)
|
| 137 |
+
- Каналы: 1 (моно)
|
| 138 |
+
- Частота дискретизации: 8 000 Гц
|
| 139 |
+
- Разрядность: 16 бит
|
| 140 |
+
- Битрейт: ~128 кбит/с
|
| 141 |
+
- Уникальные файлы по MD5: 9 215 / 9 215
|
| 142 |
|
| 143 |
+
## Основные домены
|
| 144 |
|
| 145 |
+
- Рестораны / доставка еды
|
| 146 |
+
- E-commerce / розничная торговля
|
| 147 |
+
- Рекрутинг / собеседования
|
| 148 |
+
- IT / программное обеспечение
|
| 149 |
+
- Онлайн-каталоги / digital-маркетинг
|
| 150 |
+
- Логистика / доставка
|
| 151 |
+
- Продажа автомобилей
|
| 152 |
+
- Сертификация / бизнес-услуги
|
| 153 |
+
- Юридические услуги
|
| 154 |
+
- Банковские услуги / кредитование
|
| 155 |
+
- Здравоохранение
|
| 156 |
+
- Недвижимость
|
| 157 |
+
- Телекоммуникации
|
| 158 |
+
- Страхование
|
| 159 |
+
- Энергетика / коммунальные услуги
|
| 160 |
+
- Другие бизнес-домены
|
| 161 |
+
|
| 162 |
+
## Возможные применения
|
| 163 |
+
|
| 164 |
+
- Распознавание речи
|
| 165 |
+
- Speech-to-text
|
| 166 |
+
- Аналитика разговоров колл-центров
|
| 167 |
+
- Разговорный ИИ
|
| 168 |
+
- Диалоговые системы
|
| 169 |
+
- Классификация доменов и намерений
|
| 170 |
+
- Аналитика клиентского обслуживания
|
| 171 |
+
- Исследования деловой телефонной речи
|
| 172 |
+
|
| 173 |
+
## Пример
|
| 174 |
+
|
| 175 |
+
[Прослушать sample (RU)](https://huggingface.co/datasets/MaratDV/russian-call-center-speech-ru/resolve/main/sample_ru.ogg)
|
| 176 |
+
|
| 177 |
+
## Лицензия
|
| 178 |
+
|
| 179 |
+
Только коммерческое использование.
|
| 180 |
|
| 181 |
+
Перераспространение, перепродажа, сублицензирование или публичное размещение полного датасета запрещены.
|
| 182 |
|
| 183 |
+
## Контакты
|
|
|
|
| 184 |
|
| 185 |
+
Email: marat@menkroft.com
|
| 186 |
+
Telegram: [@Marat_DV](https://t.me/Marat_DV)
|