Entidades nomeadas e extração de informação no auxílio às investigações de crimes de lavagem de dinheiro

• Ano de defesa: 2020
• Autor(a) principal: Silva, Fábio Moreira Freitas da
• Orientador(a): Vieira, Renata
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Pontifícia Universidade Católica do Rio Grande do Sul
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação
• Departamento: Escola Politécnica
• País: Brasil
• Palavras-chave em Português: Processamento de linguagem natural; Crime; Lavagem de dinheiro; Investigação; Polícia Federal
• Palavras-chave em Inglês: Natural Language Processing; Crime; Money laundering; Investigation; Federal Police
• Área do conhecimento CNPq: CIENCIA DA COMPUTACAO::TEORIA DA COMPUTACAO
• Link de acesso: http://tede2.pucrs.br/tede2/handle/tede/9259

With the increasing technological advancement, as criminal organizations are increasingly using technology to commit crimes, especially the crime of money laundering, provided for in Law 9613/98 and amended by Law 12683/12, due to its complexity and sophistication. The volume of information, coming from open sources and materials seized by public security bodies, especially by the Federal Police, presents a challenge for an investigative analysis. In order to offer greater technological support to investigations of agents, this work presents a study applied to the Named Entity Recognition (NER) models, which consists of locating and categorizing important names and unique names in free texts and other Natural Language Processing techniques (NLP) and information visualization. Experiments were carried out using parts produced and supplied by the Federal Police. The results demonstrate the possibilities of applying computational techniques in the identification of elements that include a criminal authority and material, assist as investigation teams in the elucidation of complex crimes, such as money laundering.