IDIOMATIC EXPRESSION IDENTIFICATION MODEL FROM AFAAN OROMOO TEXT USING DEEP LEARNING

This study aims at exploring the identification of idioms in Afaan Oromoo whose challenges emanates from the use of models of different languages. To address this issue, we proposed a deep learning-based model to implement this Python framework. CNN, LSTM and Bi-LSTM models were tested on a dataset of 4000 Afaan Oromoo sentences and proved that the accuracy of Bi-LSTM is very high 99%. This outperformed other deep learning models and the conventional, generally applicable machine learning algorithms such as Support Vector Machines and k-Nearest Neighbouring algorithms, thereby affirming the usability of Bidirectional Long Short Term Memories in the identification of the idioms in Afaan Oromoo