OqKay: A semi-automatic system approach to question extraction

Abstract

Basic education teachers carry out routine activities to develop evaluative exercises. In High School, this task must be aligned with entrance exams. Such practice generally demands recurrent research and intensive question selection. In this sense, this study presents the design of a semi-automatic system approach for extracting Enem questions available in internet repositories. The methodology employed web scraping methods to investigate educational materials, gather evidence of contest editions, employ regular expressions to examine the content and organize the data found within the documents. It also stored a repository of questions for utilization in the proposed system. When evaluating the approach, we obtained an acceptance and performance rate of 68.9% of questions extracted assertively. Thus, automating part of the specialized activity required to extract evaluative questions from online repositories is possible.