Using Analytics, AI to Fight Violence Against Women

In Mexico, gender-based violence and feminicide are on the rise and appear far from abating. It is a problem that requires urgent attention. However, responses from the authorities are limited and citizens, particularly women, are demanding more effective action to reduce these aggressions.

According to data from the Executive Secretariat of the National Public Security System (SESNSP) and the National Institute of Statistics and Geography (INEGI), 10 women lose their lives in the country on a daily basis. The Network for the Rights of Children Mexico (REDIM), for its part, points out that one in 10 femicides is committed against girls and adolescents under the age of 17.

In Latin America, according to information from the Economic Commission for Latin America and the Caribbean (ECLAC), in 2020, more than 4,000 women were victims of femicide. This organization states that violence against women is constant throughout their lives and the intensity increases during reproductive ages, between 30 and 44 years old, followed by adolescent girls and women between the ages of 15 and 29.

Violence against women inside and outside their homes is a problem that is not exclusive to a specific country or region; it is present practically everywhere in the world. To address it, there are projects that leverage innovations like advanced analytics, artificial intelligence (AI) and machine learning. Through these technologies, project leaders are taking the data at their disposal and turning them into insights to create strategies aimed at reducing the aggressions suffered by women and responding in a quick and effective manner.

Focused Attention

An example is the Gender Violence Unit of the Secretariat of State for Security of the Ministry of the Interior in Spain. Since 2017, it has leveraged advanced analytics and AI to assess and make an accurate diagnosis of reported cases of gender-based violence.

Law enforcement agencies analyze large volumes of data and automate processes, so they can provide immediate protection and more reliably classify and determine cases of gender-based violence. Through analytics, reported cases can be proactively managed, as can repeat offenses, so that support resources can be allocated according to the level of risk to which the victim is exposed.

A project of this magnitude posed several obstacles. One of the main challenges was to be able to coordinate the different public institutions dedicated to addressing the problem of gender-based violence in order to gather all the information available and, at the same time, add value to the process. The objective was, essentially, to unify this historical information so that, by applying advanced analytics, AI and machine learning, an accurate diagnosis was obtained on each particular case.

This process results in timely and appropriate actions through which government and public organizations seek to prevent further acts of violence against women and to closely monitor cases where the risk is latent. At the end of the day, it's about protecting women.

The results have been encouraging. Of the 50,000 new cases that the Gender-based Violence System (VOIGEN) handles each year, it has been possible to reduce the calculation times of new models from 30 to one day, and improve high-risk prediction from 30 percent to 80 percent. This means that women in imminent danger are assured that they receive care, and that protection resources will be allocated in a timely manner.  

Also noteworthy is the decrease in recurrence of cases that have been classified in an area of non-appreciable risk: from 14 to less than 10.

Therefore, analytics, AI and machine learning make an important contribution to combating feminicide and gender-based violence, processing available information and making an optimized modeling of the probabilities of risk and incidence, thus improving prediction to evaluate cases and make effective and timely decisions.

This initiative is a comprehensive social responsibility project that was based on demonstrating and raising awareness of how data and advanced analytics combine to help combat gender-based violence and feminicide. The results go beyond a region or country, and can be used by other governments to apply them to combat a problem that affects millions of women and girls on the planet.

In addition to the administration of justice and protection, there is also the possibility of establishing more efficient emotional support networks for victims and providing a greater sense of security for women.

Companies like SAS provide the technological foundation to incorporate advanced analytics and AI to automate the analysis of huge volumes of criminal data, helping to better examine algorithms, identify new risk indicators and to do so in a shorter time.  

For more information on how analytics are being leveraged to combat gender-based violence and feminicide, as well as solving other problems afflicting humanity, visit Data 4 Good or contact Héctor Cobo, Regional Vice President, SAS Mexico, Caribbean and Central America, mail: HectorJavier.Cobo@sas.com