Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Desaparecidxs: characterizing the population of missing children using Twitter

Carolina Coimbra Vieira, Diego Alburez-Gutierrez, Marília R. Nepomuceno, Tom Theile

TL;DR

This study leverages real-time Twitter data from Guatemala's Alerta Alba-Keneth to characterize the demographic and geographic profile of missing children during 2018-2020, supplementing official police records. By extracting age, sex, and place of disappearance from tweet images via OCR and geocoding locations, the authors reveal a pronounced overrepresentation of adolescent girls (especially 13-17) among missing children and urban–border regional patterns that may relate to trafficking and violence. The work demonstrates the potential of social media data to illuminate hard-to-reach populations and informs policy by identifying high-risk groups and spatial concentrations. It also discusses limitations like reporting biases and data-matching challenges, suggesting replication in other settings and methodological enhancements.

Abstract

Missing children, i.e., children reported to a relevant authority as having "disappeared," constitute an important but often overlooked population. From a research perspective, missing children constitute a hard-to-reach population about which little is known. This is a particular problem in regions of the Global South that lack robust or centralized data collection systems. In this study, we analyze the composition of the population of missing children in Guatemala, a country with high levels of violence. We contrast the official aggregated-level data from the Guatemalan National Police during the 2018-2020 period with real-time individual-level data on missing children from the official Twitter account of the Alerta Alba-Keneth, a governmental warning system tasked with disseminating information about missing children. Using the Twitter data, we characterize the population of missing children in Guatemala by single-year age, sex, and place of disappearance. Our results show that women are more likely to be reported as missing, particularly those aged 13-17. We discuss the findings in light of the known links between missing people, violence, and human trafficking. Finally, the study highlights the potential of web data to contribute to society by improving our understanding of this and similar hard-to-reach populations.

Desaparecidxs: characterizing the population of missing children using Twitter

TL;DR

This study leverages real-time Twitter data from Guatemala's Alerta Alba-Keneth to characterize the demographic and geographic profile of missing children during 2018-2020, supplementing official police records. By extracting age, sex, and place of disappearance from tweet images via OCR and geocoding locations, the authors reveal a pronounced overrepresentation of adolescent girls (especially 13-17) among missing children and urban–border regional patterns that may relate to trafficking and violence. The work demonstrates the potential of social media data to illuminate hard-to-reach populations and informs policy by identifying high-risk groups and spatial concentrations. It also discusses limitations like reporting biases and data-matching challenges, suggesting replication in other settings and methodological enhancements.

Abstract

Missing children, i.e., children reported to a relevant authority as having "disappeared," constitute an important but often overlooked population. From a research perspective, missing children constitute a hard-to-reach population about which little is known. This is a particular problem in regions of the Global South that lack robust or centralized data collection systems. In this study, we analyze the composition of the population of missing children in Guatemala, a country with high levels of violence. We contrast the official aggregated-level data from the Guatemalan National Police during the 2018-2020 period with real-time individual-level data on missing children from the official Twitter account of the Alerta Alba-Keneth, a governmental warning system tasked with disseminating information about missing children. Using the Twitter data, we characterize the population of missing children in Guatemala by single-year age, sex, and place of disappearance. Our results show that women are more likely to be reported as missing, particularly those aged 13-17. We discuss the findings in light of the known links between missing people, violence, and human trafficking. Finally, the study highlights the potential of web data to contribute to society by improving our understanding of this and similar hard-to-reach populations.
Paper Structure (9 sections, 5 figures)

This paper contains 9 sections, 5 figures.

Table of Contents

  1. Introduction
  2. Related Work
  3. Data
  4. Guatemalan National Police data
  5. Twitter data
  6. Data protection
  7. Results
  8. Discussion
  9. Conclusion

Figures (5)

  • Figure 1: Example of a tweet from the Alerta Alba-Keneth Twitter account (@alba_keneth).
  • Figure 2: Number of missing children by month: a comparative overview of data from the Guatemalan National Police and Alerta Alba-Keneth Twitter data. Note that data for May-December 2019 were not provided by the National Police.
  • Figure 3: Age and sex distribution of the number of missing children by month of reported disappearance (2018 - 2020) according to the Alerta Alba-Keneth Twitter data.
  • Figure 4: Stock of missing children: cumulative number of disappearances (2018 - 2020) according to the Alerta Alba-Keneth Twitter data.
  • Figure 5: Geographic distribution of the number of reported missing children (2018 - 2020) according to the Alerta Alba-Keneth Twitter data. Location data were extracted from the "Place of disappearance" field of the tweet.