Keynote: Investigation opportunities in Latin America

Speaker: Laura Montoya – Laura Montoya, Founder & Executive Director Accel.ai

Abstract: In this presentation, we delve into the burgeoning landscape of AI research and startup opportunities in Latin America, presented from the perspective of LatinX in AI (LXAI). We explore the unique challenges and vast potential of the region in harnessing artificial intelligence for sustainable growth and innovation. The talk will highlight key areas of AI research gaining momentum in Latin America, the role of local communities in shaping AI’s future, and how startups in the region are uniquely positioned to address both local and global problems. Attendees will gain an understanding of the current state of AI in Latin America, the untapped opportunities, and the importance of inclusive and diverse perspectives in driving AI research and entrepreneurship.  Learn how to connect and collaborate with the vibrant LatinX in AI network. Join us to discover how Latin America is not just following but also contributing to the AI revolution, creating a landscape rich with opportunities for growth, collaboration, and innovation.

Keynote: The long trail: Natural Language Processing for the Indigenous Languages of the Americas

Speaker: Manuel Mager – Applied Scientist at Amazon AWS

Abstract:  The indigenous languages of the Americas have historically been relegated to the study of NLP, with the exception of specialized workshops for endangered and low-resource languages. Luckily this has changed. In recent years a strong interest in these languages and other endangered and low resourced languages has emerged. In this framework, we present our contribution to the study of these languages, focusing on the polysynthetic languages of Mexico. We present our work on morphological segmentation with neural networks in environments with extreme low resource settings. Later we will analyze the impact of this segmentation on the machine translation task, and our experience from the AmericasNLP workshop. Since these languages also have constant contact with the dominant languages of their countries, the phenomenon of code-switching is also present. Therefore, we conclude the talk with the presentation of this phenomenon, we introduce the task of subwork-language identification, and present our study of the effect of code-switching in machine translation environments, where polysynthetic languages are present.

Panel:  LATAM Data Horizons: Pioneering Paths in Data Mining and Web Search

Panelists: Saiph Savage; Ricardo Baeza-Yates; Victor Fragoso

Moderator: Diego Carmona

Abstract: This panel provides an in-depth exploration of the evolving realms of data mining and web search technologies in Latin America (LATAM). It emphasizes not only the latest trends and challenges but also the innovative solutions and career-building strategies employed in the region. Distinguished professionals from across LATAM will share their personal insights and experiences, shedding light on how they have built successful careers in data mining and web search fields. Attendees will gain valuable perspectives on navigating the unique opportunities and obstacles in LATAM’s digital landscape, understand the impact of regional regulations on data practices, and learn about the role of local tech communities in shaping the future of these technologies. This panel aims to inspire and inform, offering a comprehensive view of the current state and future potential in these critical areas of technology.

Panel: Mayan speakers meets Natural Language Processing (NLP) scientists

Panelists: Manuel Mager; Samuel Canul; Fidencio Briceño; María Elisa Chavarrea; Silvia Fernández

Moderator: Silvia Leirana

Abstract: The development of technologies for the indigenous languages of America has had a slow development where native speakers have had little or nothing to do with it. This panel brings together NLP specialists with native Mayan speakers to discuss how to take into account the actual needs of the communities when developing NLP technologies

Demo: Building a Search Engine to Preserve Mayan Language

Speaker: Alejando Molina

Abstract: taantsil.com.mx is the first search engine designed to preserve and promote the Mayan language. During the development of this software system, several challenges have been encountered due to the technological gap between Maya and other languages, which have been cleverly overcome. * Throughout the day there will be a permanent stand with the Information Retrieval System for the Mayan Language.

Accept WSDM Day Talks

Automated Topic Generation for the Mexican Platform for Access to Government Public Information during the Period 2003-2020
Hermelando Cruz-Pérez and Alejandro Molina-Villegas

Automatic Extraction of Patterns in Digital News Articles of Femicides occurred in Mexico by text mining techniques
Jonathan Zárate-Cartas and Alejandro Molina-Villegas

Profiling urban mobility patterns with high spatial and temporal resolution: A deep dive into cellphone geo-position data
Luisa Fernanda Chaparro Sierra and Jose Huertas

Genomic-World Fungi Data: Synteny part
Pedro Escobar-Turriza, Luis Muñoz-Miranda and Alejandro Pereira-Santana

AI-Assisted Intelligent Systems to Enhanced Surgical Robotics
Alberto Munoz

Integrating Knowledge Graph Data with Large Language Models for Explainable Inference
Carlos Efraín Quintero-Narváez and Raul Monroy

Preserving Heritage: Developing a Translation Tool for Indigenous Dialects
Melissa Robles, Cristian Adrián Martínez, Juan Camilo Prieto, Sara Palacios Chavarro and Rubén Manrique