Automatic Idiom Identification Model for Amharic Language

Idiomatic expressions are important natural parts of all languages and prominent our daily speech. Idioms cannot be interpreted from the word that they formed directly people may not understand meaning. From past literature, it was noted idiom affects Natural Language Processing research like machine translation, semantic analysis, sentiment analysis. Other English, Chinese, Indian idioms recognized through different methods in research. As there is no standard method to identify Amharic idioms, this study aimed build a model for language using supervised learning approach. The used eight hundred labeled training two testing books “የአማ ረኛ ፈሊጦች” documents. To measure performance model, we accuracy, precision, recall, F-score. Finally, 97.5% accuracy result achieved dataset showing promising result. contributes information systems discourse about improving awareness, knowledge researchers on idioms.