Honorio Apaza

Honorio Apaza

AI/NLP Researcher specializing in Low-resource Languages

Ilo, Moquegua, Perú
•
[email protected]
AiMara Lab Founder IEEE Senior Member Professor at UNAM

I'm the Founder of AiMara Lab and a Professor at Universidad Nacional de Moquegua (Perú). I also serve as a Senior Member of IEEE and Co-Founder of the AiMARA Research Group.

Previously, I worked as an AI Developer at ITACA LEGAL and INDECOPI, developing NLP models for legal tech and recommendation systems. I hold a Master's degree in Computer Science from Universidad Nacional de San Agustín (UNSA), where I specialized in Neural Machine Translation for low-resource languages.

I earned my Systems and Computer Engineering degree from UNAM, where I won the university's research project contest funded by mining royalties. My research lies at the intersection of Deep Learning, Natural Language Processing, and the preservation of native languages like Aymara and Quechua.

Available for consulting and collaborations. Contact me for more information.

Master's Thesis Highlight

Key Research Contribution
Neural Machine Translator for Native Language Aymara

Developed and evaluated neural machine translation models (Spanish ↔ Aymara), focusing on Seq2Seq with attention versus Transformer architectures. Results demonstrated that the Transformer model significantly outperforms Seq2Seq, achieving higher masked accuracy (0.87 vs 0.68) and better BLEU scores (0.042 vs 0.031). primary contribution was demonstrating the efficiency of Transformers for agglutinative, low-resource languages.

Professional Experience

Professor
2023 - Present
Universidad Nacional de Moquegua, Ilo
Teaching: Artificial Intelligence, Algorithms Design & Analysis, Databases, Programming Fundamentals, Data Structures, and Thesis Seminar.
AI Developer (NLP)
2022
ITACA LEGAL S.A.C, Lima
Developed NLP models for legal tech: Automatic text classification, document summarization, clustering legal precedents, and semantic text retrieval.
AI Model Developer (NLP)
2021
INDECOPI, Lima
Improved behavioral dictionaries and recommender models for detecting infringing behaviors.
Tech Stack: BERT, Word2vec, TF-IDF, Logistic Regression, VSM, Naive Bayes.
Professor Assistant & Researcher
2018 - 2022
Universidad Nacional de Moquegua
Principal researcher: "Text mining to build a data mining model for job offer recommendations". Research grant winner (UNAM-2016).

Education

Master's in Computer Science
2021 - 2023
Universidad Nacional de San Agustín, Arequipa
Master of Computer Science candidate. Focused on Advanced AI and NLP.
Systems and Computer Engineering
2017 - 2018
Universidad Nacional de Moquegua
Engineer degree. Winner of the UNAM-2016 research project contest with financing from mining royalties.
Bachelor in Systems and Computer Engineering
2012 - 2016
Universidad Nacional de Moquegua
Graduated with Scholarship from "Beca 18 Perú" (National Scholarship and Educational Credit Program).

Recent Publications

Development of a Chatbot Prototype to Serve University Graduates
Intelligent Systems and Applications, Springer Nature (2024)
NLP Models to Clarify Degree Regulations in Peruvian Universities
IEEE ANDESCON (2024)
Peruvian Sign Recognition (LSP) to Native Quechua Language Using LSTM
IEEE ANDESCON (2024)
Neural Machine Translation for Aymara to Spanish
Intelligent Systems and Applications, Springer (2023)
Neural Machine Translation for Native Language Aymara to English
Future Technologies Conference (FTC), Springer (2023)
Job Recommendation Based on Curriculum Vitae Using Text Mining
Advances in Information and Communication, Springer (2021)

Skills & Expertise

Research Areas

NLP NMT Deep Learning Text Mining Low-resource Langs

Programming

Python Java C++ R

AI Technologies

BERT Transformers LSTMs Word2Vec HuggingFace

Frameworks

PyTorch TensorFlow scikit-learn Keras