An analysis of violence against women based on victims' reports

• Isabella Tannús Corrêa UFU
• Elaine Ribeiro Faria UFU

Resumo

Every two seconds, a woman is a victim of physical or verbal violence in Brazil, and every 1.4 seconds, a woman is a victim of harassment. Due to the growth of anthropology and notions of feminism, violence against women has been seen as a public problem. This work proposes a method to analyse reports of violence against women using text mining techniques to characterize the aggression or the aggressor. We collected data from electronic newspapers, sites, and social networks, pre-processed them, extracted topics from the data using the Latent Dirichlet Allocation (LDA), and classified the text according to the frequency of occurrence (constant or sporadic) using the Naive Bayes algorithm. Among the analyzed cases, most of them indicate that the violence was sporadic; it happened once. The results suggest that most reports of constant violence – meaning that they occur frequently – are related to someone close to the victim, such as a family member, spouse or friend. Words such as ”speak”, ”say” and ”report” are frequent, indicating the victim’s willingness to express the aggression. When categorizing documents into topics, it is possible to find scenarios of family abuse executed by the father or a brother, and also the presence of sexual violence as one of the aggressions suffered.

Palavras-chave: violence against women, text mining, sentiment analysis

Referências

Christian Aranha and Emmanuel Passos. 2006. A Tecnologia de Mineração de Textos. RESI - Revista Eletrônica de Sistemas de Informação 5, 2(2006), 1–6. https://doi.org/10.21529/RESI.2006.0502001

Ricardo Baeza-Yates and Berthier Ribeiro-Neto. 2013. Recuperação de Informação: Conceitos e Tecnologia das Máquinas de Busca(2 ed.). Bookman Editora, Porto Alegre. 590 pages.

Alice Bianchini. 2011. Por que as mulheres não denunciam seus agressores? Com a palavra, a vítima. [link]. Acesso em 03 abr. 2019.

David M. Blei, Andrew Y. Ng, and Michael I. Jordan. 2003. Latent Dirichlet Allocation. J. Mach. Learn. Res. 3 (mar 2003), 993–1022. 

Joshua Charles Campbell, Abram Hindle, and Eleni Stroulia. 2015. Chapter 6 - Latent Dirichlet Allocation: Extracting Topics from Software Engineering Data. In The Art and Science of Analyzing Software Data, Christian Bird, Tim Menzies, and Thomas Zimmermann(Eds.). Morgan Kaufmann, Boston, 139–159. https://doi.org/10.1016/B978-0-12-411519-4.00006-9

CLAUDIA. 2019. Eva Luana: Jovem denuncia padrasto pelo Instagram: “Me fazia comer o próprio vômito”. [link]. Acesso em 03 abr. 2019.

Igor T. Correa, Daniel D. Abdala, Rodrigo S. Miani, and Elaine R. Faria. 2018. Sentiment Analysis of Twitter Posts About the 2017 Academy Awards. In Anais do XV Encontro Nacional de Inteligência Artificial e Computacional (São Paulo). SBC, Porto Alegre, RS, Brasil, 320–331. https://doi.org/10.5753/eniac.2018.4427

Instituto Maria da Penha. 2019. Relógios da violência. https://www.relogiosdaviolencia.com.br/. Acesso em 03 abr. 2019.

Patrícia Helena Maia Alves de Andrade. 2015. Aplicação de Técnicas de Mineração de Textos para Classificação de Documentos: um Estudo da Automatização da Triagem de Denúncias na CGU. Master’s thesis. Universidade Federal de Brasília. https://repositorio.unb.br/handle/10482/21004 Mestrado Profissional em Computação Aplicada.

Secretaria Especial de Políticas para as Mulheres. 2006. Lei Maria da Penha. [link].

V. D. Gil, J. D. Betancur, I. C. Puerta, L.M. Montoya, and J.M. Sepulveda. 2018. The Feminicide in Colombia and Mexico: a Text Mining Analysis. The Turkish Online Journal of Design, Art and Communication - TOJDAC (Mar 2018), 170–177. http://www.tojdac.org/tojdac/VOLUME8-MRCHSPCL_files/tojdac_v080MSE121.pdf

João Ribeiro Carrilho Junior. 2007. Desenvolvimento de uma Metodologia para Mineração de Textos. Master’s thesis. Pontifícia Universidade Católica do Rio de Janeiro. https://doi.org/10.17771/PUCRio.acad.11675 Programa de Pós graduação em Engenharia Elétrica.

George Karystianis, Armita Adily, Peter Schofield, Lee Knight, Clara Galdon, David Greenberg, Louisa Jorm, Goran Nenadic, and Tony Butler. 2018. Automatic Extraction of Mental Health Disorders From Domestic Violence Police Narratives: Text Mining Study. J Med Internet Res 20, 9 (13 Sep 2018), e11548. https://doi.org/10.2196/11548

George Karystianis, Armita Adily, Peter W Schofield, David Greenberg, Louisa Jorm, Goran Nenadic, and Tony Butler. 2019. Automated Analysis of Domestic Violence Police Reports to Explore Abuse Types and Victim Injuries: Text Mining Study. J Med Internet Res 21, 3 (12 Mar 2019), e13067. https://doi.org/10.2196/13067

Joel Neto, Neto Alexandre, D Santos, Celso Kaestner, and Alex Freitas. 2000. Document Clustering and Text Summarization. (06 2000).

Lucas Rodrigues, Jorge da Silva Júnior, and Fábio Lobato. 2019. A culpa é dela! É isso o que dizem nos comentários das notícias sobre a tentativa de feminicídio de Elaine Caparroz. In Anais de VIII Brazilian Workshop on Social Network Analysis and Mining (Belém). SBC, Porto Alegre, RS, Brasil, 47–58. https://doi.org/10.5753/brasnam.2019.6547

Dilma Rousseff. 2015. LEI Nº 13.104. http://www.planalto.gov.br/cCivil_03/_Ato2015-2018/2015/Lei/L13104.htm

Gaspar Silva, Nadia Felix, and Márcio Dias. 2018. Deteção de bullying: Como identificar automaticamente essa prática em redes sociais?Revista de Sistemas de Informação da FSMA (06 2018), 11–19.

Ian Witten and Eibe Frank. 2010. Data Mining: Practical Machine Learning Tools and Techniques.

Jia Xue, Junxiang Chen, and Richard Gelles. 2019. Using Data Mining Techniques to Examine Domestic Violence Topics on Twitter. Violence and Gender 6 (02 2019). https://doi.org/10.1089/vio.2017.0066