First-ever global DataJam in the anti-trafficking field

Oct132020

FIRST-EVER GLOBAL TECHNOLOGICAL COMPETITION TO DEVELOP DATA-BASED SOLUTIONS AGAINST HUMAN TRAFFICKING

A total of 158 participants and 29 experts from 13 countries as well as 24 data-based solutions to disrupt human trafficking are part of the results of the successful third version of DataJam Pasos Libres run by Pasos Libres and IBM. This year the emblematic competition was held online, globally and multilingual for the first time with the support of the Traffik Analysis Hub, Western Union, The Future Society, Walk Free, Platzi, STOP THE TRAFFIK, Movistar, la Fiscalía General de la Nación, One Young World, Think Tank de Ideas Disruptivas contra el crimen Organizado and the United Nations Office on Drugs and Crime.

With 34 teams from Colombia, Mexico, Peru, Chile, Bolivia, the United States, Nigeria, Tunisia, Guinea and Ecuador, representing 42 universities, international organizations, government institutions, companies, and law enforcement agencies, the DataJam Pasos Libres Online 2020 became the first-ever global technological competition in the anti-trafficking field. Some of the organizations represented in the contest included the Colombian National Police, the Anti-trafficking Division of the Bolivian Police, the Peruvian Attorney General’s Office, the International Organization for Migration in Tunisia and the Ministry of Government of Ecuador.

During the eight days of the competition, participants had access to IBM technologies and specialized lectures, talks of experts, workshops, training material and an exclusive DataJam’s Participant Guide. Judges from Pasos Libres, IBM, Traffik Analysis Hub, Movistar and UNODC selected 4 winning and 5 outstanding teams based on the solution’s potential to tackle the proposed challenges, positively impact the counter-trafficking field and be developed in the Traffik Analysis Hub.

MEET THE WINNERS

LATIN AMERICAN WINNERS

First place: Aleph Mind, Colombia
Solution: The Witness
The Witness is a technological tool to investigate the connections between different internet pages where human trafficking networks could operate. It uses web scrapping supported by natural language processing techniques to extract data such as phone numbers, organizations’ names, people’s names, social networks, locations, images, emails, and bank accounts. The set of web pages explored by The Witness is represented by a mathematical object known as a graph, which allows studying this phenomenon from a holistic perspective, providing valuable information for decision-makers based on different network analyzes, such as vertices centrality, communities identification, perpetrators and victims relations, among others.

Second place: DaTTa, Colombia
Solution: Job Satisfaction Index
DaTTa created Job Satisfaction Index a sentiment analysis model trained with a Twitter dataset to tagged tweets based on specific characteristics related to forced labor and trafficking dynamics. This neural network was pre-trained with millions of texts from the internet, which guarantees exceptional precision in the texts’ numerical representation. After trained, the model was tested with tweets from specific cities containing specific words related to labor issues to build the job satisfaction index and visualize the positive or negative work environment in different Latin American countries and several cities in Colombia.

Third place: Law-IT, Perú
Solution: Trata Brújula
Law-IT developed Trata Brújula, a machine learning model based on neural networks that allows predicting the number of human trafficking cases in each department of Peru to identify the most vulnerable areas. The predictive model was trained with socioeconomic indicators related to human trafficking risk factors such as poverty, labor informality, population density from 25 departments of Peru from 2016 to 2018

REST OF THE WORLD WINNER

Techtavism, United States.
Solution: Prevention as Intervention
Prevention as Intervention calculates a person’s risk as a target for human trafficking by considering the specific risk associated with their location and factors taken from social media such as gender, mental status, and friends. By partnering with companies like Facebook and Google, targeted ads can be displayed to high-risk individuals educating them and encouraging them to report suspicious behavior. Additionally, individual risk scores can be aggregated by region to help police departments create accurate risk assessments of their communities. Departments in higher-risk areas can allocate more resources towards education against trafficking.

WHAT IS NEXT?

Pasos Libres and the IBM Service Corps are developing a post-competition plan and an expansion strategy of the DataJam Pasos Libres. On the one hand, the DataJam organizers will support the winning teams in the further development of the proposed solutions in the Traffik Analysis Hub, the first-ever global data hub to facilitate the sharing of information about human trafficking across all industries and sectors. Participants will partake in a three-phases program during which they will work to improve their proposals along with experts of the industry.

On the another, due to the results of the DataJam Pasos Libres Online 2020, next year, the initiative will be expanded to many other countries to reach more participants, experts and partners. The DataJam’s expansion strategy includes creating “regional chapters” in countries such as Mexico, Ecuador, Peru, the United States, the United Kingdom and Tunisia to promote the collection and analysis of regional trafficking data and strengthening an open innovation community against human trafficking based on IBM technologies and the Pasos Libres’ DataJam format.