Advanced search
Start date
Betweenand

Social diagnosis with data mining

Grant number: 19/23039-6
Support type:Research Grants - Innovative Research in Small Business - PIPE
Duration: December 01, 2020 - May 31, 2022
Field of knowledge:Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
Principal Investigator:Edimilson Ricardo Azevedo Novais
Grantee:Edimilson Ricardo Azevedo Novais
Company:Orion Soluções em Gestão Ltda. - EPP
CNAE: Suporte técnico, manutenção e outros serviços em tecnologia da informação
Pesquisa e desenvolvimento experimental em ciências sociais e humanas
Pesquisas de mercado e de opinião pública
City: Santana de Parnaíba
Associated scholarship(s):20/15998-0 - Social Diagnosis with Data Mining, BP.TT
20/16744-2 - Social Diagnosis with Data Mining, BP.TT
21/00054-0 - Social Diagnosis with Data Mining, BP.TT

Abstract

One of the main characteristics of the Brazilian reality is the diversity that exists between regions, states and municipalities. In recent years, Brazil has made significant progress in the social area, there have been improvements, but the country still suffers from many problems that are difficult to control. Violence, criminality, education, housing, health, racism, hunger are some of the major social problems facing Brazilians. The issue of child and adolescent rights has occupied significant space in the social area since the creation of the Child and Adolescent Statute (ECA). Even so, children and adolescents are still unprotected due to countless factors in society. We can still point out that poverty in childhood and adolescence is complex and has multiple dimensions that go beyond money and legislation (UNICEF, 2019). Given this difficulty, diagnosis is one of the main demands of public organizations or civil society involved in the definition of public policies. Consistent and up-to-date diagnostics make it possible to define these policies, justify project demands, generate proactivity in social care, and also allow systematic planning. This research project aims to propose a software platform that assists municipal councils of rights, municipalities and social organizations in diagnosing the care of the population of Brazilian cities, particularly children and adolescents efficiently. The research originated from the demand of civil society and organizations involved in the definition of public policies, confirmed in situational diagnosis projects of the reality of children and adolescents recently implemented in some municipalities by the proposing company. The main function of the proposed system is to integrate and view population service data through data mining to improve the outcomes of the social diagnostic process. The methodology of this proposal is developed in three stages: (1) selection and integration of social data; (2) data mining; and (3) data interpretation and diagnosis. In this context, the system works with data from data warehouses composed of public and local indicators. This research work began in 2015 and evolved with the results of the execution of diagnostics performed by the company. Recently the company applied data classification using decision tree (algorithm C4.5) for automatic detection of violation of child and adolescent rights. The application of data mining allowed the validation of the indicators and showed new information on the number of violations of the rights of children and adolescents. Overall, the project contributed to strengthening the guarantee of the rights of children and adolescents in the municipalities participating in the research due to the positive results of the diagnosis made. The diagnoses made resulted in an effective management of public resources in the municipalities served, and increased the revenues generated by the proposing company. The next step is to continue research with machine learning algorithms and build an intelligent software platform consisting of preprogrammed data mining techniques to discover new relationships between social diagnostic data that cannot be perceived by humans. depending on the diversity and complexity of the information.Keywords: Social Diagnosis; Data mining; Machine Learning. (AU)