A conceptual framework combining visual text mining and causal mapping to support decision process: a case study in the education system in Brazil
Data
2023-12-18
Autores
Abuabara, Leila 
Tipo
Tese de doutorado
Título da Revista
ISSN da Revista
Título de Volume
Resumo
A Pesquisa Operacional (PO) é uma área de conhecimento multidisciplinar que se desenvolve ao apoiar situações complexas do mundo real. Por esta razão, ela evolui com o movimento do mundo e suas questões. Nas últimas quatro décadas, a PO tradicional (PO Hard) alcançou diversas áreas que a complementaram. Uma dessas áreas é a ciência de dados, já que cada vez mais instituições e organizações estão disponibilizando grande volume de dados e informações ao mesmo tempo que contamos com ferramentas analíticas cada vez mais poderosas. No outro extremo, temos a PO Soft, também denominada Métodos de Estruturação de Problemas (PSM), um conjunto de metodologias sistêmicas e predominantemente qualitativas que lidam com as complexidades dos chamados problemas ‘confusos’ que são tão abundantes no mundo contemporâneo com muitas partes interessadas, objetivos conflitantes e interesses diversos. Conectar estas duas áreas (ciência de dados e PSM) metodologicamente e em uma aplicação real é uma contribuição relevante para a PO. Esta pesquisa tem dois objetivos. O primeiro é propor e discutir uma estrutura metodológica conceitual que combina essas duas abordagens, neste caso, a mineração de texto e o mapeamento causal para dados de mídia. O segundo é aplicar esta abordagem em um estudo de caso real e de interesse social. Neste sentido,
utilizamos entrevistas públicas com especialistas da área de educação no contexto do mau desempenho dos estudantes brasileiros no Programa Internacional de Avaliação de Estudantes (PISA) em 2018. Além de provocar o debate sobre o valor desta combinação multimetodológica que pode ser útil em diversos contextos, a aplicação destacou os principais temas de atuação dentro da educação brasileira que merecem atenção em políticas educacionais. Dentre elas, os benefícios e o prazer da leitura; o engajamento das famílias desde muito cedo; as práticas pedagógicas baseadas em uma aprendizagem com significado em um ambiente contemporâneo e tecnológico; os mecanismos educacionais que
focam na formação de um cidadão pleno, incluindo uma formação profissional e assim contribuindo para o desenvolvimento social e econômico do país. Finalmente, esperamos que os resultados desta pesquisa possam ser considerados em planos de governos futuros.
Operations Research (OR) is a multidisciplinary area of knowledge that develops supporting complex real-world situations. For this reason, it evolves with the movement of the world and its issues. Over the last four decades, traditional (Hard) OR has reached diverse areas that complemented it. One of these areas is data science as more and more institutions and organizations make large amounts of data and information available while we also have increasingly powerful analytical tools. At the other extreme, we have Soft OR, also labeled as Problem Structuring Methods (PSM), a set of systemic and predominantly qualitative methodologies that deal with the complexities of the so called ‘wicked’ problems which are so abundant in our contemporary world, with many stakeholders, conflicting goals, and diverse interests. Connecting these two areas (data science and PSM) methodologically and into a real application seems to be a relevant contribution to OR. This research has two objectives. The first is to propose and discuss a conceptual methodological framework that combines these two approaches, in this case text mining and causal mapping for media data. The second is to apply this approach to a real case study of social interest. In this sense, we used public interviews with experts in the field of education in the context of the poor performance of Brazilian students in the Program for International Student Assessment (PISA) in 2018. In addition to sparking debate about the value of this multi-methodological combination that can be useful in different contexts, the application highlighted the main themes within Brazilian education that deserve attention in educational policies. Among them, the benefits and pleasure of reading; the engagement of families from a very early age; pedagogical practices based on meaningful learning within a contemporary and technological environment; educational mechanisms that focus on the formation of a full citizen, including professional training and thus contributing to the social and economic development of the country. Finally, we expect that the outcomes of this research can be considered in future government plans.
Operations Research (OR) is a multidisciplinary area of knowledge that develops supporting complex real-world situations. For this reason, it evolves with the movement of the world and its issues. Over the last four decades, traditional (Hard) OR has reached diverse areas that complemented it. One of these areas is data science as more and more institutions and organizations make large amounts of data and information available while we also have increasingly powerful analytical tools. At the other extreme, we have Soft OR, also labeled as Problem Structuring Methods (PSM), a set of systemic and predominantly qualitative methodologies that deal with the complexities of the so called ‘wicked’ problems which are so abundant in our contemporary world, with many stakeholders, conflicting goals, and diverse interests. Connecting these two areas (data science and PSM) methodologically and into a real application seems to be a relevant contribution to OR. This research has two objectives. The first is to propose and discuss a conceptual methodological framework that combines these two approaches, in this case text mining and causal mapping for media data. The second is to apply this approach to a real case study of social interest. In this sense, we used public interviews with experts in the field of education in the context of the poor performance of Brazilian students in the Program for International Student Assessment (PISA) in 2018. In addition to sparking debate about the value of this multi-methodological combination that can be useful in different contexts, the application highlighted the main themes within Brazilian education that deserve attention in educational policies. Among them, the benefits and pleasure of reading; the engagement of families from a very early age; pedagogical practices based on meaningful learning within a contemporary and technological environment; educational mechanisms that focus on the formation of a full citizen, including professional training and thus contributing to the social and economic development of the country. Finally, we expect that the outcomes of this research can be considered in future government plans.
Descrição
Citação
ABUABARA, Leila. A conceptual framework combining visual text mining and causal mapping to support decision process: a case study in the education system in Brazil. 2023. 144f. Thesis of Doctor of Science – Instituto Tecnológico de Aeronáutica and Universidade Federal de São Paulo, São José dos Campos.