Bustio-Martínez, Lázaro - PhD

Software applications

Presentations

Quimiometrix-I: Chemometrics toolbox for chemical data analysis.
Quimiometrix-II: Chemometrics toolbox for chemical data analysis.
PAS: Plataforma para el análisis de sustancias.

Projects developed & directed

2026-2027

Explainable Fake News Detection using Psychological and Linguistic Pattern Analysis

Team members: Lázaro Bustio-Martínez (Technical Leader), Felipe A. Trujillo Fernández, Vitali Herrera-Semenets, Miguel Ángel Álvarez-Carmona, Jorge Ángel González-Ordiano, Pedro Santander Molina, Jan van den Berg, Leovildo Diago Cisneros, Viviana Inés Fuentes Fuentes.
Digital disinformation exploits psychological vulnerabilities, emotional triggers, and persuasive linguistic strategies to manipulate perception and influence decision-making. Its effectiveness lies in targeting people rather than technical systems, making automated detection a complex multidisciplinary challenge. Although recent advances in Artificial Intelligence have improved fake news detection, most existing approaches rely on opaque models, are strongly biased toward English datasets, and largely ignore psychological and sociocultural dimensions.
This project aims to develop explainable and lightweight Artificial Intelligence methods for the automatic detection of disinformation in textual content, with a specific focus on Spanish-language contexts. The approach integrates Natural Language Processing, Machine Learning, discourse analysis, and psychological models of persuasion to identify recurrent linguistic and emotional patterns used in deceptive narratives. Beyond classification accuracy, the project emphasizes model interpretability, enabling users to understand why a piece of content is flagged as disinformation. The final outcome is a transparent, privacy-respecting, and platform-independent prototype that contributes both to scientific research and to media literacy and social resilience against digital manipulation.

2022-2023

Phishing attacks detection in electronic messages using Artificial Intelligence techniques

Team members: Lázaro Bustio-Martínez (Tecnical Leader), Miguel A. Álvarez-Carmona, Vitali Herrera-Semenets, Jorga A. González-Ordiano , Luis Zúñiga-Morales, Juan Luis García-Mendoza.
Phishing is the cybercrime that use electronic media to obtain sensitive information from their victims. To do so, it is common to induce feelings of urgency, necessity, vulnerability, doubt, curiosity, etc. The success of these attacks is remarkable as they attack people rather than systems. Recently there has been a worldwide increase in phishing attacks and they are being targeted at socio-demographic groups of interest. In view of this situation, it is necessary to propose new tools for detecting such attacks. To deal with these cases, it is necessary to employ methods that "learn" from identified attacks and apply "what has been learned" in unknown situations. In this sense, Machine Learning groups the necessary methods to detect phishing and zero-day attacks. For this reason, this project aims to develop methods capable of identifying phishing attacks in electronic media by using Machine Learning techniques.

More about this project can be found here.

2020-2022

A Lightweight Data Representation for Phishing URLs Detection in IoT Environments

Team members: Lázaro Bustio-Martínez (Tecnical Leader), Miguel A. Álvarez-Carmona, Vitali Herrera-Semenets, Claudia Feregrino-Uribe , René Cumplido
Phishing is a cyber-attack that exploits victims' technical ignorance or naivety and commonly involves a Uniform Resources Locator (URL). Hence, s. With the raising of the Internet of Things (IoT), phishing attacks are moving to this field due to the number of IoT devices and the amount of personal information they handle. Although several approaches were proposed for phishing attack detection, the URL-based Machine Learning approaches obtain better performance results, but all depend on the feature set. Contradictorily, only a few works on selecting the best-suited feature set for improving the phishing detection process have been published. The present research explores obtaining a feature set that substantially enhances the phishing detection rate in IoT environments. Hence, a feature selection algorithm was adopted and extended to get the most representative feature set. When Random Forest is used with the proposed data representation, the phishing URL attacks discovery rate is 99.57%.
More about this project can be found in: Information Sciences journal and here.

2013-2015

Quimiometrix II

Team members: Lázaro Bustio-Martínez (Principal Senior Software Engineer), Isneri Talavera-Bustamante, Katherine González-Vargas, Yenice Coma-Peña.
The project "Quimiometrix II" encompasses a set of tools specifically crafted to facilitate various stages of chemical and biochemical data analysis, including preprocessing, exploration, classification, and calibration. It serves as a comprehensive system that leverages cutting-edge statistical pattern recognition techniques. Notably, this initiative aligns with Cuba's continuous advancement in laboratory instrumentation within the chemical domain. "Quimiometrix II" stands as an automated platform dedicated to executing intricate multivariate processing tasks concerning chemical and biochemical information. Its utilization empowers researchers and analysts to handle complex data sets efficiently, ultimately contributing to the enhancement of analytical outcomes and insights in the realm of chemistry.

2010-2012

PAS: Plataforma computacional para el análisis de sustancias

Team members: Lázaro Bustio-Martínez (Principal Senior Software Engineer), Isneri Talavera-Bustamante, Fidel A. Castro-Puebla, Félix González, Yoana Martínez-Díaz.
The project "Plataforma computacional para el análisis de sustancias" streamlines diverse substance analysis by integrating advanced tools and techniques. It provides an accessible environment for researchers, offering tailored features for data input, preprocessing, and result generation. Raw data from sources like spectroscopy and chromatography can be refined through preprocessing, enhancing subsequent analysis. The platform's plugin-based architecture ensures continuous relevance by seamlessly integrating new algorithms, fostering innovation. Interactive visualization tools enable intuitive data exploration, while statistical analysis aids pattern recognition. The adaptable platform serves chemistry, biology, and medicine, encouraging interdisciplinary collaboration. In essence, the "Computational Platform for Substance Analysis" merges technology and expertise, fueling research, innovation, and discovery while maintaining scientific currency through its innovative architecture.

2010-2012

Data Mining in high-volume datasets

Team members: Lázaro Bustio-Martínez (Researcher), René Cumplido, Alejandro Mesa-Rodríguez, José Hernández-Palancar.
"Data Mining in High-Volume Datasets" is an innovative project focused on developing novel data mining algorithms, specifically for data classification. Leveraging reconfigurable hardware and Field-Programmable Gate Arrays (FPGAs), this initiative aims to enhance classification processes in large datasets. By harnessing FPGA's parallel processing capabilities, the project seeks to accelerate and optimize classification tasks, unlocking insights from vast data volumes efficiently. This fusion of advanced algorithms and reconfigurable hardware promises to revolutionize data mining, catering to the demands of modern big data scenarios while enabling precise and rapid classification outcomes.

2008-2010

Quimiometrix I

Team members: Lázaro Bustio-Martínez (Principal Senior Software Engineer), Isneri Talavera-Bustamante, Noslén Hernández-González, Diana Porro-Muñoz, Oneisys Cuadra, Rainer Larín-Fonseca, Malena Bermúdez.
"Quimiometrix" refers to a project that gave rise to the software known as "Quimiometrix 1." Developed by researchers at the Center for Advanced Technology Applications (CENATAV). Quimiometrix 1 is designed for the multivariate processing of chemical and biochemical data. This powerful software amalgamates chemical, mathematical, and computational methods to extract and analyze concealed insights from numerical matrices generated during research endeavors. With its capability to undertake multivariate data analysis, Chemometrics presents substantial potential and streamlined tools for the research community at a national level.

Datasets

SpaPhish: A Spanish Dataset for Phishing and Psychological Pattern Detection.

Created: 2025.12.30Author: Lazaro Bustio-Martinez, Viviana Fuentes-Fuentes, Luisa Fernanda Agudelo-Fuentes, Vitali Herrera-Semenets, Darian Llanes-Guilarte, Felipe Antonio Trujillo-Fernández, Carlos Antonio Cardeña-Matamoros, Carlos Francisco Betancourt-Moreno, Joshua Ismael Haase-Hernández, Andrés Guillermo Molano-Jiménez, Jan van den Berg.Download here.

Dataset description:

SpaPhish is a curated dataset of Spanish-language phishing and legitimate emails annotated at both technical and psychological levels. It includes structural metadata, linguistic features, and continuous labels of persuasion principles (POP), enabling analysis beyond binary classification. The dataset is designed for reproducible research and fine-grained evaluation of persuasive intensity in real-world social engineering scenarios.
Dataset description URL: TBA

Phishing URL dataset.

Created: 2020.05.02Author: Lázaro Bustio Martínez (lazaro.bustio@ibero.mx)Used in: A Lightweight Data Representation for URL Phishing Detection in IoT Environments.Download here.

Dataset description:

A URL dataset composed of 52,000 URLs were collected: 26,000 were taken from the Alexa Top website and labelled as Legitimate (class 0). The other 26,000 URLs were obtained from PhishTank and were labelled as Phishing (class 1).
Dataset description URL: A Lightweight Data Representation for URLs Phishing Detection in IoT Environments

Electrical demand/consumption and climatological data of a small Mexican community.

Created: 2023.09.05Author: Santos-Moreno, Milagros ; González-Ordiano, Jorge Ángel; Quiroz, Emilio; Perez-De La Mora, D.A; Mizrahi-Cojab, Jaime ; Román-Sánchez, Emilio; Montero-Cantú, José Pablo ; Bustio-Martinez, Lazaro.Used in: Santos-Moreno, M., González-Ordiano, J. Á., Quiroz-Ibarra, J. E., Perez-De La Mora, D. A., Mizrahi-Cojab, J., Román-Sánchez, E., Montero-Cantú, J. P., & Bustio-Martinez, L. (2023). Weather and electrical demand and consumption data of a small Mexican community. Data in Brief.Download here.

Dataset description:

This dataset includes electricity demand/consumption and weather data from a small community in Puebla, Mexico. The electricity data, measured in 15-minute intervals, covers five households from May 2022 to May 2023. The weather data, containing 24 variables, was collected from May 2022 to March 2023. Missing values are denoted with -1.

Page updated

Google Sites

Report abuse