Off-line restricted-set handwritten word recognition for student identification in a short answer question automated assessment system

Off-Line Handwritten Bilingual Name Recognition for Student Identification in an Automated Assessment System

“…6). In addition, both Thai and English names are available for each student name identification, even if some of the name components are misrecognised, the other name A related work found in the literature was performed in the context of automated assessment in the English language [1] with a different dataset size of 2,040 words. To be able to compare recognition rates attained from the proposed WRLGGF, more experiments were carried out by using the proposed WRLGGF together with WRGGF, LGGF and GGF techniques using the SVM on the English dataset of 2,040 words [1].…”

“…In addition, both Thai and English names are available for each student name identification, even if some of the name components are misrecognised, the other name A related work found in the literature was performed in the context of automated assessment in the English language [1] with a different dataset size of 2,040 words. To be able to compare recognition rates attained from the proposed WRLGGF, more experiments were carried out by using the proposed WRLGGF together with WRGGF, LGGF and GGF techniques using the SVM on the English dataset of 2,040 words [1]. Recognition rates of the proposed WRLGGF together with a comparison of the existing system [1] and the proposed system are displayed in Table II. As can be seen, when using WRLGGF on the English SIS dataset [1], and applying it to SVMs, a higher recognition rate of 99.61% was obtained compared to 98.59% for MDF, 93.62% for GGF using ANNs, and 99.55% of the original GGF using SVMs.…”

“…The English name component dataset was obtained from the authors' previous work [1] plus 6 additional English name components from 3 writers. To sum up, the bilingual dataset was increased to 4,120 (412 name components  10 samples of each name component) samples.…”

“…Off-line Automatic Assessment Systems (OFAAS) and Student Identification Systems (SIS) benefit from off-line handwriting recognition techniques. There is not much work that has been reported on SIS [1]. In addition, there is only a small amount of research focusing on OFAAS [1], [2], [3], [4].…”

“…There is not much work that has been reported on SIS [1]. In addition, there is only a small amount of research focusing on OFAAS [1], [2], [3], [4]. Commonly for manual assessment, when the marking of examination papers is concluded, the marker has to rewrite each student's mark onto a marking summary sheet.…”