Abstract:
The manual correction of discursive questions is not an easy task. In educational area is seen as a process that contributes to student learning. The time spent to correc...Show MoreMetadata
Abstract:
The manual correction of discursive questions is not an easy task. In educational area is seen as a process that contributes to student learning. The time spent to correct an assessment is high and the teacher cannot maintain the same level of stiffness in corrections during all his work. This article presents a new approach for automatic corrections of discursive questions proposing use of a virtual learning environment, named in this study as “LabSQL”. The approach is based on the use of n-grams and functions in PHP (soundex, similar_text and srt_compare) to measure the similarity between texts, with a system accuracy of 92.15%, compared to the result of a manual work. This result is higher than similar studies that reached up 82.10% of accuracy.
Date of Conference: 19-22 June 2013
Date Added to IEEE Xplore: 17 October 2013
Electronic ISBN:978-989-98434-0-0
Print ISSN: 2166-0727
Conference Location: Lisboa, Portugal