You need to agree to share your Hugging Face information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

English-Kalenjin Agricultural Parallel Corpus (MT-Eng-Kln)

Dataset Summary

MT-Eng-Kln is a parallel corpus of 50,887 English-Kalenjin sentence pairs in the agricultural domain. It is published by AI4Agric to support machine translation and agricultural advisory services for Kalenjin-speaking farmers in Kenya. The English side consists of farmer questions and extension content covering crops, livestock, soil fertility, pest management, post-harvest handling and agribusiness. Each English sentence has a human translation into Kalenjin that has been reviewed and approved.

Kalenjin has very little parallel text available for machine translation. This corpus is among the larger English-Kalenjin resources released openly.

The corpus is one of three parallel datasets released together. The other two are MT-Eng-Kik (Kikuyu) and MT-Eng-Swa (Swahili). Around 46,500 serial numbers are shared across all three datasets. This allows multi-way alignment through English.

Languages

Role Language ISO 639 code
Source English en
Target Kalenjin kln

Kalenjin is a Southern Nilotic language cluster spoken mainly in the Rift Valley region of Kenya. It includes several closely related varieties such as Kipsigis and Nandi. TODO: state which variety or varieties the translators used.

Dataset Structure

Data Fields

Field Type Description
source_text string English sentence
target_text string Kalenjin translation
language string Target language (always kalenjin)
serial_number string Unique identifier for the sentence pair. Prefixes are Hq or KARLO
import_source string Origin of the English sentence (Huggingface, KARLO or Farmerchat)

Data Instance

{
  "source_text": "Alternative feed sources for smallholder dairy farmers",
  "target_text": "Koorookik che bo amitwogik che bo nekoto bo kobotiik che kimen che bo baetab tuga chekikee",
  "language": "kalenjin",
  "serial_number": "Hq406",
  "import_source": "Huggingface"
}

Data Splits

The dataset is released as a single train split of 50,887 pairs. No official validation or test split is defined. Users who report results should state how they split the data and should fix a random seed.

Dataset Statistics

Source Composition

import_source Pairs Share
Huggingface 45,424 89.3%
KARLO 3,113 6.1%
Farmerchat 2,350 4.6%
Total 50,887 100%

Sentence Length (words)

Mean Median Min Max Total tokens Vocabulary (lower-cased)
English 10.9 11 4 96 556,019 16,119
Kalenjin 10.7 10 3 78 543,281 34,968

The Kalenjin vocabulary is more than twice the size of the English vocabulary on a similar number of tokens. This reflects the rich verbal and nominal morphology of Kalenjin as well as spelling variation, since Kalenjin orthography is not fully standardised. Subword or byte-level tokenisation is recommended.

Usage

from datasets import load_dataset

ds = load_dataset("AI4Agric/MT-Eng-Kln", split="train")
print(ds[0])

# Create a reproducible 80/10/10 split
tmp = ds.train_test_split(test_size=0.2, seed=42)
vt = tmp["test"].train_test_split(test_size=0.5, seed=42)
train, validation, test = tmp["train"], vt["train"], vt["test"]

Dataset Creation

Source Data

The English sentences come from three sources:

  • Huggingface (Hq serial numbers): farmer questions from an existing agricultural question dataset on Hugging Face. TODO: name and link the dataset.
  • KARLO (KARLO serial numbers): sentences from agricultural extension and research material. TODO: give the full name of the source organisation.
  • Farmerchat (Hq serial numbers): farmer queries from the FarmerChat advisory service. TODO: confirm and link.

Many of the Huggingface and Farmerchat sentences are real farmer questions with informal spelling and phrasing. These have been left as they are because they reflect how farmers actually write.

Translation and Review

Native Kalenjin speakers translated the English sentences. Each translation was then reviewed and approved before it was included. TODO: describe the number of translators, how they were recruited, the review protocol and the platform used.

Cleaning Applied Before Release

Leading, trailing and repeated whitespace was removed. Pairs with an empty source or target were dropped. Exact duplicate pairs were removed, which left 50,887 pairs out of 50,898.

Considerations and Limitations

  • Domain. The corpus is specific to agriculture. Models trained only on it will not generalise well to other domains.
  • Dialect and spelling variation. Kalenjin covers several varieties and its orthography is not fully standardised. The same word may be spelled in more than one way across the corpus.
  • Repeated source sentences. 40 English sentences occur more than once with different Kalenjin translations. These should be kept in the same split to avoid leakage.
  • Untranslated terms. Some technical terms and pest or product names (for example, "mealy bug") are kept in English within the Kalenjin text.
  • Noisy source text. Farmer questions contain spelling and grammar errors on the English side.

Related Datasets

Citation

@misc{ai4agric_mt_eng_kln_2026,
  author       = {AI4Agric},
  title        = {{MT-Eng-Kln}: English-Kalenjin Agricultural Parallel Corpus},
  year         = {2026},
  publisher    = {Hugging Face},
  howpublished = {\url{https://huggingface.co/datasets/AI4Agric/MT-Eng-Kln}}
}
Downloads last month
32