Eagle Eye, a machine learning model to fight against feminicide

0
Eagle Eye, a machine learning model to fight against feminicide

An initiative presented in Santo Domingo has been launched in an attempt to prevent femicide in the Dominican Republic. The platform, called Eagle Eye , uses artificial intelligence to analyse data and generate predictive scenarios to prevent such tragedies. The tool, developed by Plan International and Asesoría Integral a la Medida a Empresas (AIME), with the participation of the local Public Prosecutor’s Office, hopes to reduce the rate of femicide in the country, the fifth highest in Latin America.

Artificial intelligence tool to fight feminicide

The Dominican Republic is the fifth country with the highest feminicide rate in Latin America (1.9 per 100,000 inhabitants according to data from the Economic Commission for Latin America and the Caribbean). In 2020, 60 women died in the country at the hands of their partners or former partners.

To combat these tragedies, Plan International working in the Dominican Republic has partnered with the technology company AIME to develop an AI solution. Virginia Saiz, director of Plan International in the Dominican Republic, spoke around the use of AI to design these kinds of solutions:

“Until now, artificial intelligence was mainly used by the private sector. Social problems are extremely complex and must also rely on artificial intelligence. This type of platform gives us more information, predictive ability and to provide more effective solutions.”

A model generating predictive situations through machine learning

The Attorney General’s office provided the data set needed to developEagle Eye and its machine learning algorithm. With this data, the system is able to simulate models that could be used to develop public policies to reduce the rate of femicide. To do this, the system has been trained with many different data sets in an attempt to increase accuracy.

Gregorio Tapia, Marketing, Innovation and Communication Manager for Plan International in the Dominican Republic explains:

“Using artificial intelligence to respond to such a complex social problem can put public policy makers in a proactive mode and generate new ways of understanding and dealing with the problem.”

The program was first tested and found to be successful in predicting femicide situations by taking into account variables such as location, baseball games, salary payment dates, vacation days, as well as other economic factors the researchers believe have influenced previous cases. Rainier Mallol, co-founder and president of AIME, said the tool was able to identify the months and municipalities where the most femicides occurred. He also said that the quality of the data provided by the authorities should be improved to increase accuracy. According to him, one of the major problems with the data available so far is that there is little information on the victims.

During the remainder of its test phase, Eagle Eye will generate predictions up to 2023. From these scenarios, solutions should be put in place to help follow up on complaints of assault and serve as a support to prevent further possible feminicides. Plan International has invited the government of the Dominican Republic and the private sector to use this tool to find new and innovative solutions to other problems.

Translated from Eagle Eye, un modèle de machine learning pour lutter contre les féminicides