Gabriel Alejandro León Paredes

Research Interests

In general, I am working in high-performance computing using CUDA architecture of NVIDIA graphics cards to accelerate processes. My studies are based on the application of CUDA with natural languages processing techniques in order to process, analyze and retrieve information from large-scale datasets.

Areas of interesting:

Smart Cities, High Performance Computing, Scientific Programming, Natural Language processing, Semantic Analysis, Information Retrieval.

PhD. Thesis

Título: “A Parallel Computing Approach for Latent Semantic Analysis using CUDA”
Director: Ph.D. Liliana Ibeth Barbosa Santillán.

 
Abstract: With a large amount of information avail- able on the Internet through Social Networks, Blogs, and Websites, among others sources, we need appropriate methods to analyze text information. Latent Semantic Analysis (LSA) can be used to index and reduce large Term- Document matrices using Singular Value Decomposition (SVD) technique. However, to indexing large amounts of information, we need quicker implementations. The General- Purpose computing on Graphics Processing Units (GPGPU) can solve big problems faster through the thousands of concurrently threads on the many-core multiprocessors of GPUs. In this paper, we present a GPU-LSA system, using the NVIDIA Compute Unified Device Architecture (CUDA), and python programming language. The acceleration reached by our system implementation for large matrices with one hundred and fifty thousand millions of values is five to six times faster than the LSA- CPU version.
 

Education:

• Máster Universitario en Ingeniería Web, Universidad Politécnica de Madrid, España (2013).
• Ingeniero de Sistemas, Universidad del Azuay, Ecuador (2010).

Doctoral Research  stays:

• School of Computer Science and Technology, Nantong University, China (2017)
• Facultad de Informática, Universidad Complutense de Madrid, España (2017).

Professional experience:

• Assistant Professor
  Universidad Politécnica Salesiana
  septiembre de 2013 – actualidad (4 años 1 mes)
• Research Group Coordinator.
  Cloud Computing, Smart Cities & High Performance Computing
  Universidad Politécnica Salesiana
  agosto de 2014 – actualidad (3 años 2 meses)
• Member Scientific Staff
  Revista Ingenius
  septiembre de 2016 – actualidad (1 año 1 mes)
• Member Scientific Staff
  XII Congreso Iberoamericano de Inclusión Educativa con tecnologías Emergentes CIIEE 2017
• Member Scientific Staff
  ATICA 2016
• Junior Programmer
  Universidad Politécnica Salesiana
  febrero de 2011 – septiembre de 2013 (2 años 8 meses)

Awards

·       Becario SENESCYT 2012
Secretaría de Educación Superior, Ciencia, Tecnología e Innovación

·       Becario UPS 2015
Universidad Politécnica Salesiana

Publications

1. Gabriel A. León-Paredes and Liliana I. Barbosa-Santillán, “A Heterogeneous System Based on Latent Semantic Analysis (hLSA)”, in the GPU Technology Conference (GTC), 2017
2. Pablo L. Gallegos-Segovia, Jack F. Bravo-Torres, Marco A. López-Reinoso, Gabriel A. León-Paredes, Juan D. Jara-Saltos and Victor M. Larios-Rosillo, “Unified Communications as a Service in HybridNetworks”, in the Journal of Advanced Research in Dynamical and Control Systems, Volume 11, Special Edition, 2017.
3. Juan J. Sánchez-Escobar, Liliana I. Barbosa-Santillán, Luis F. Barbosa-Santillán, Fatemeh G. Nasri-Shandiz and Gabriel A. León-Paredes, “Onto-Traffic: A Semantic Traffic Analysis Tool based on GPS Data for Smart-Cities”, in The Sixth International Conference on Advances in Information Mining and Management, 2010.

Links with additional info:

Linkedin:                                https://www.linkedin.com/in/gabrielleonp/
Google Scholar:                    https://scholar.google.es/citations?user=fIoURQwAAAAJ&hl=es
Reseacrh group GIHP4C:      http://gihp4c.blog.ups.edu.ec/
Research group data mining: http://www.dataminingengineeringgroup.net