Amharic Open Information Extraction with Syntactic Sentence Simplification

Open Information Extraction (OIE) is the process of discovering domain-independent relations from natural language text. It has recently received increased attention and been applied extensively to various downstream applications, such as text summarization, question answering, and informational retrieval. In this paper, we propose a method of OIE for Amharic language. To improve the performance of relation extraction, the proposed OIE method implements a sentence simplification technique that breaks down complex and compound sentences into simple sentences. Linguistic rules are utilized to extract domain-independent and unanticipated relation instances with their arguments from simple sentences. The proposed method and algorithms are implemented and evaluated with a dataset from different domains. Test results show that the system achieved an overall precision of 0.88.