Logo Lanfrica

McNoblesse/Language_Identification

Domain:

natural language processing

Record type:

dataset
Creator:
McN
Host:
South African Language Identification Prediction # South African Language Identification Hack 2023 EA language classification hackathon ## Description ExploreAI Academy Classification Hackathon ## Overview South Africa is a multicultural society that is characterised by its rich linguistic diversity. Language is an indispensable tool that can be used to deepen democracy and also contribute to the social, cultural, intellectual, economic and political life of the South African society. The country is multilingual with 11 official languages, each of which is guaranteed equal status. Most South Africans are multilingual and able to speak at least two or more of the official languages. From South African Government With such a multilingual population, it is only obvious that our systems and devices also communicate in multi-languages. In this challenge, you will take text which is in any of South Africa's 11 Official languages and identify which language the text is in. This is an example of NLP's Language Identification, the task of determining the natural language that a piece of text is written in. ## Evaluation The evaluation metric for this competition is Mean F1-Score. The F1 score, commonly used in information retrieval, measures accuracy using the statistics precision p and recall r. Precision is the ratio of true positives (tp) to all predicted positives (tp + fp). Recall is the ratio of true positives to all actual positives (tp + fn). The F1 score is given by: The F1 metric weights recall and precision equally, and a good retrieval algorithm will maximize both precision and recall simultaneously. Thus, moderately good performance on both will be favored over extremely good performance on one and poor performance on the other. ## Submission Format For every index in the dataset, submission files should contain one column: lang_id, the language tag of the predicted language. ## Dataset Description The dataset used for this challenge is the NCHLT Text Corpora collected by the South African Department o …