Semantic Role Labeler for Amharic Text Using Memory Based Learning

No Thumbnail Available



Journal Title

Journal ISSN

Volume Title


Addis Ababa University


Human knowledge is recoded in natural language. The records are kept in computers or on paper to be manipulated and reserved for use in the future. Then natural language processing plays an important role to accomplish computer domain-independent understanding of natural language. A useful step towards that role is assigning semantic roles to the constituents of a sentence, which refers to semantic role labeling. It allows one to recognize semantic arguments of a situation, even when expressed in different syntactic configurations. Still now there is no automatic semantic role labeler for Amharic adopting semantic role labeler of other languages for Amharic language is difficult since Amharic language is morphologically complex and the general role of language specific characteristics in the extraction of semantic content is different. Therefore, the objective of this study was to develop semantic role labeler for Amharic text using memory based learning algorithm. The system was trained on 551 instances. To evaluate the result leave one out cross-validation technique was employed. The evaluation result showed that the accuracy in classifying semantic role achieved 82.51% accuracy and 78.19% F-Score with default parameter and 89.29% accuracy and 79.77% F-Score with optimize parameter setting.



Semantic Role Labeler, Memory Based Learning, Treebank, Cross Validation, Feature Extraction