Project abstract
Education is essential for country’s development. Education provides children, youth and adults with the knowledge and skills to be active citizens and to fulfil themselves as individuals. Furthermore, education has the ability to change and to induce change and progress in society.
One of the Europe 2020 targets stipulates that at least 40% of the population aged 30-34 should have tertiary education attainment by 2020. In this context, in the last decade it has been conducted a deep analysis, particularly on higher Education, which forced the evaluation, review and reformulation of the processes used to guarantee the quality of the education services provided. In Portugal, this reflection has been encouraged by the publication of a legal framework on quality assessment in higher education and the creation of the Agency for Assessment and Accreditation of Higher Education – A3ES. The Agency has promoted the establishment of internal quality assurance systems, fostering the creation of a systematic collection of data that may enable to identify the main constraints and problems, enhancing the decision-making process.
Having a better understanding of which students are more likely to face difficulties in their educational process and identifying the factors that influence these difficulties, higher education institutions will be able to timely develop strategies to increase the graduation rate and mitigate their attrition rates. This will contribute to the achievement of satisfactory levels of attainment.
Currently, high education institutions have made a big effort and investment on creating systems to collect education related data. However, institutions have not been able to analyze this data and turn it into valuable information. Therefore, data analysis in this context is promising, as it enables institutions to discover and extract hidden knowledge of students’ patterns from educational environment. Data mining is a computer-assisted process of digging through and analyzing enormous sets of data and then extracting the meaning of the data. Data mining tools predict behaviors and future trends, allowing decision makers to make proactive, knowledge-driven decisions.
This project aims at exploring the use of educational data (e.g. social-economic, demographic, higher education access average and academic results) to identify ‘bottlenecks’ (at the curricular unit level) that constraint academic sucess and to predict students’ academic performance. Furthermore, this project aims discussing the main factors that underlie academic performance. The models developed will be supported by data mining techniques and markov chains.
keywords: Education and Big Data; Business Intelligence applied to education; Educational Data Mining; Predictive moddeling; Learning Analytics; Academic performance
Team
Supervisor: Vera Miguéis (DEGI-FEUP)
Students: André Filipe Roque Silva ; Pedro Afonso Paulino Ferreira de Castro
Contributors: Ana Freitas (LEA, FEUP), Paulo Garcia (DEF-FEUP; LEA-FEUP); UPortoDigital
Dates: September 2015 to …
Outcomes
This project has led to 2 Master Dissertations at FEUP and a paper