Topics on Brazilian Education:
What do Open Access Databases allow Data Mining to contribute to Public Education Policies?
Date:
Tuesday, January 24, 2017 - 12:00am
Location:
Lemann Center - Conference Room
In this talk Paulo Adeodato presents an overview of Data Mining (DM) as a process of preparation of actionable knowledge from experts and extracted from data to be embedded in a Decision Support System (DSS). Rather than on techniques, this talk focuses on demands from the application domain as an instantiation of a Domain-Driven Data Mining (D3M) approach to Education, which has been recently called Educational Data Mining. Instead of using state-of-the-art algorithms, in this talk, Paulo addresses the issue of the system supplying functionalities (requirements) and expressing the extracted knowledge in a form UNDERSTANDABLE by EXPERTS on EDUCATION. He presents several applications of binary decision solutions he has been developing for the last 10 years, mainly on open access databases from INEP, and the the difficulties faced in Brazil, ranging from Ill-defined problems to disconnected data files. He concludes the presentation discussing how DDSs can help education and (DSS) challenges to be tackled, some immediate and some to appear soon.