Honorio Apaza
Computer Science Master • AI Researcher • NLP Specialist
My research interests lie in Artificial Intelligence, specializing in Natural Language Processing (NLP). This includes neural machine translation for low-resource languages like Aymara, text mining for recommendation systems, and the classification, summarization, and clustering of textual information across various domains, such as legal.
Master's Thesis
This thesis develops and evaluates neural machine translation models for translating from Spanish to Aymara, focusing on Seq2Seq with attention and Transformer architectures. Results show that the Transformer outperforms Seq2Seq across key metrics, with higher masked accuracy (0.87 vs. 0.68), better BLEU (0.042 vs. 0.031), and improved ROUGE and METEOR scores, highlighting its superior accuracy and efficiency for low-resource languages like Aymara.