Amharic-ASR-Dataset

speech recognition for Amharic language

Text data is adobted from Contemporary Amharic Corpus (CACO)

Length

---

• 2.3 hours

unsplitted

---

unsplitted row data can be found in unsplitted branch of this repository.

Contents

---

Amharic-ASR-Dataset
|
├── data
│   └── records
│       ├── train
│       │   └── *.wav
│       ├── val
│       │   └── *.wav
│       ├── test
│           └── *.wav
│       
└── README.md
└── chars.txt
└── charset.json
└── linker.txt
└── raw_text_file.txt

Data collection App

---

Dataset was recorded using LIG-Aikuma app

Dataset

---

@inproceedings{gezmu-etal-2018-contemporary,
    title = "Contemporary {A}mharic Corpus: Automatically Morpho-Syntactically Tagged {A}mharic Corpus",
    author = {Gezmu, Andargachew Mekonnen  and
      Seyoum, Binyam Ephrem  and
      Gasser, Michael  and
      N{\"u}rnberger, Andreas},
    booktitle = "Proceedings of the First Workshop on Linguistic Resources for Natural Language Processing",
    month = aug,
    year = "2018",
    address = "Santa Fe, New Mexico, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/W18-3809",
    pages = "65--70",
    abstract = "We introduced the contemporary Amharic corpus, which is automatically tagged for morpho-syntactic information. Texts are collected from 25,199 documents from different domains and about 24 million orthographic words are tokenized. Since it is partly a web corpus, we made some automatic spelling error correction. We have also modified the existing morphological analyzer, HornMorpho, to use it for the automatic tagging.",
}