top of page

Organized

Criminal

Violence

Event

Data.

(OCVED)

JAVIER OSORIO.
UNIVERSITY OF ARIZONA.

01 map.

MY KNOWLEDGE LEVEL IN SOFTWARE

This visualization presents an interactive heat map of the territorial violent presence of 

Organized Criminal Groups in Mexico between 2000 and 2019.

 

The map includes the following Organized Criminal Groups:

  • Cartel de Juárez

  • Cartel de Tijuana

  • Cartel de Sinaloa

  • Cartel del Golfo

  • La Familia Michoacana

  • Los Zetas

  • Cartel de Jalisco Nueva Generación

  • Organización de los Beltrán Leyva

  • Organización de La Barbie

  • Huachicoleros (oil thieves)

  • Other

 

Each of these Organized Criminal Groups is further disaggregated in a multitude of sub-groups or criminal cells.   

 

This map is the first delivery of Organized Criminal Violence Event Data (OCVED), a database generated using computerized text annotation with Eventus ID, a protocol for supervised event coding capable of geo-referencing actors mentioned in text written in Spanish (Osorio & Reyes 2017).

 

The information came from 105 sources including national and local newspapers, as well as government press releases between 1/1/2000 and 12/31/2019.

 

 

To open the map, click on the image below.

MAP

02 CRIMINAL GROUPS.

The database contains geo-referenced information at the municipality-day level of the following Organized Criminal Groups:

Cartel de Juárez

Cartel de Juarez

La Linea

Los Aztecas

Grupo Sombra

Cartel de Tijuana

Cartel de Tijuana
Arellano Félix
Facción de El Teo
Los Incorregibles

Cartel de Sinaloa

Cartel de Sinaloa
El Chapo
Zambada
Nacho Coronel
Cartel del Pacifico
Los Jaguares
Artistas Asesinos
Los Salazar

Cartel del Golfo

Cartel del Golfo 
Los Corsarios
Los Cuervos
Los Escorpiones
Los Halcones
Los Linces
Los Metros

La Familia Michoacana

La Familia Michoacana
Caballeros Templarios
Cartel de Tepalcatepec
Cartel de Zicurian
Champis 
El Charro
La Empresa
La Nueva Empresa 
La Nueva Familia Michoacana
Los Gordos 
Los Jaguares
Los Perez 
Los Pumas
Los Troyanos 
Los Viagras 
Brown Side Family 

Organización Beltrán Leyva

Beltrán Leyva
Limpia Mazateca
La Oficina
Los Pelones
La Nueva Administración
Los Maquina
Nuevo Cartel de la Sierra
Cartel Independiente de Acapulco

Los Zetas

Los Zetas
Cartel del Noreste
Los Broncos
Los Cotorros
Los Enfermeros
Los Enrique 
Los Guerreros
Los Hijos del Diablo
Los Lancheros
Los Legendarios
Los Numeros
Banda del Chaparro

Cartel de Jalisco Nueva Generacion

Cartel de Jalisco Nueva Generacion
Cartel de Colima
Cartel del Milenio
Gente Nueva
La Resistencia
Los Mata Zetas 
Los Valencia

Organización de la Barbie

La Barbie
Cartel del Pacifico Sur
Guerreros Unidos
La Mano con Ojos
Los Rojos
Los Ardillos
La Barredora

Huachicoleros
(Oil Thievs)

Huachicoleros

Cartel de Santa Rosa de Lima

Other

Cartel de Neza
Cartel de Tlahuac
Individual
Los Antrax 
Los Ayutla
Los Babicoras 
Los Benitez
Los Brasilennos de la Reynosa 
Los Burros
Los Cachos
Los Campesinos 
Los Caraveo
Los Carcachos
Los Cazo
Los Chachines
Los Charros 
Los Chatos 
Los Chavez
Los Chavez 
Los Chibuyar 
Los Chinconcuac
Los Chinos
Los Cholos
Los Chutas 
Los Ciruelos 
Los Colin
Los Colmenos 
Los Costenos
Los Coyotes
Los Coyotes 
Los Danieles 
Los Dannys

Los Dedos 

Los Dianelillos 
Los Dragones 
Los Flacos
Los Fuerenos 
Los Furcios 
Los Garcias 
Los Garibay 
Los Gaseros
Los Gatos 
Los Gigies 
Los Indios 
Los Japos
Los Jarochos 
Los Jarquin 
Los Jeremia 
Los Juniors 
Los Kinkones 
Los Lagartus 
Los Ledesma 
Los Liborio
Los Limones 
Los Marin
Los Maximos
Los Mellizos
Los Moncaus 
Los Mudos
Los Nacizos 
Los Negros
Los Netos 
Los Ninnos de Oro
Los Nortenos

Los Nuevos Pelones 

Los Ortiz 
Los Pacheco 
Los Pajaros 
Los Palafox 
Los Panchos
Los Patilla 
Los Pedraza
Los Pedraza 
Los Perros 
Los Petriciolet 
Los Pipas
Los Pipos 
Los Punta Norte
Los Purina
Los Rancheros
Los Rapidos 
Los Ratoncitos
Los Rodolfos
Los Santeros
Los Simpson
Los Smith
Los Tablajeros
Los Temixco
Los Thunder
Los Tuneros
Los Vectra
Los Veneros
Los Yonqueros
Los Zampayo
Los Zodiaco
M60
Union Tepito

The data presented in the map indicates the violent presence of Organized Criminal Groups at the municipality-day level. However, it does not necessarily reflect cases in which Criminal Groups are present in a territory but refrain from exercising violence (Arjona 2011). Such cases would correspond to areas of hegemonic or monopolistic control in which the use of violence may not be necessary.

 

Despite the effort of systematically gathering thousands of news reports from dozens of national and local newspapers, and relevant government agencies, the media does not cover all the actions of all Organized Criminal Groups. As such, the data provides a conservative measure of Organized Criminal Groups violent presence. Inferences drawn from this data should be interpreted carefully.

​

The grouping of different criminal cells into larger Organized Criminal Groups is based on my understanding of these groups based on multiple academic and journalistic documentation. However, these Organized Criminal Groups are fluid organizations and they frequently emerge, collapse, split, ally, and mutate. Therefore, these groupings should not be considered as fixed categories.

ARMD ACTORS
TRENDS
METHODOLOGY

03 METHODOLOGY.

Methodology

Methodological Summary

 

This visualization presents an interactive heat map of Organized Criminal Groups in Mexico geo-referenced at the municipality-day level between 2000 and 2019. 

 

To generate the data, the project relies on the computerized identification of actors and locations using Eventus ID (Osorio & Reyes 2017).  The raw information comes from a large collection of news stories from local and national Mexican newspapers and press releases from government agencies.

​

To select the relevant stories, the methodology first relied on a team of trained human coders who identified news stories by strictly following the codebook guidelines. The team of research assistants has an intercoder agreement of  90.4% and a Fleiss’ Kappa of 0.704. In order to automatically update the selection of relevant news stories, the methodology used the manually gathered data to train a Machine Learning (ML) model. After comparing the performance of different ML models, the Linear Regression model performed with the highest accuracy with an F1 of 0.949. The final collection of selected news stories became the input corpus for computerized event coding.

 

Based on a dictionary of Organized Criminal Groups containing more than 7,000 names of criminal groups and individuals, Eventus ID reads the content of the corpus in order to identify the actors and places mentioned in the news stories. The coding protocol implements additional routines of geographic disambiguation and deduplication of coded actors. The result of the coding process is a database of Organized Criminal Groups geo-located at the municipality-year level.

 

Each data point in the database represents one observation of Organized Criminal Groups in a given municipality-day as detected in the corpus The exact location of the incidents at the XY coordinate level is not determined in the newspaper narratives, so the methodology assigned the coordinates of the municipality center to their respective observation. The map presents a heat map displaying a higher intensity color range in areas concentrating a higher density of Organized Criminal Groups. 

 

Funding

OCVED is possible thanks to the support of a variety of funding sources

that have supported this research effort through different stages:

 

The University of Arizona, Research, Discovery & Innovation, Technology & Research Initiative Fund, 2018-2020.

Cornell University, Mario Einaudi Center for International Studies Postdoctoral Fellowship, 2013-2014.

Yale University, Order, Conflict, and Violence Predoctoral Fellowship, 2012-2013. 

The Harry Frank Guggenheim Foundation, Dissertation Fellowship, 2012-2013.

The Kellogg Institute for International Studies - University of Notre Dame, Dissertation Year Fellowship, 2012-2013.

The United States Institute for Peace, Jennings Randolph Peace Scholar Dissertation Fellowship, 2011-2012.

The National Science Foundation, Doctoral Dissertation Research Improvement Grant, 2011-2012.

The Social Science Research Council - Open Society Foundation, Drugs, Security, and Democracy Fellowship. 2011-2012.
The Kellogg Institute for International Studies - University of Notre Dame, Graduate Research Grant, 2011.

 

Research Team

Javier Osorio

Principal Investigator​

Graduate Research Assistants

Alejandro Beltrán

Luis Enrique Medina

External Consultant

Alejandro Reyes

Undergraduate Research Assistants

(current and past)

Ingrid Ibarra

Sofia Revilak

Luna Ruiz

Jose Carlos Rivera

Sara Torres

Erick Alonso

Eréndira González

Frank Orta

Mauricio Ochoa

04 data access.

OCVED is publicly available for free. Please enter your information to request access to the data. You will receive an email with the data and related documentation.

 

Please cite as:

Osorio, Javier, and Alejandro Beltrán. 2020. "Enhancing the Detection of Criminal Organizations in Mexico using ML and NLP," 2020 International Joint Conference on Neural Networks (IJCNN),  pp. 1-7, doi: 10.1109/IJCNN48605.2020.9207039

Intended data use:

Thanks for submitting your data request.

DATA ACCESS
bottom of page