Open AccessAssessing the Line-By-Line Marking Performance of n-Gram String Similarity Method
Author(s) -
Arsmah Ibrahim,
Zainab Abu Bakar,
Nuru’l–‘Izzah Othman,
Nor Fuzaina Ismail
Publication year - 2009
Publication title -
scientific research journal/scientific research journal
Language(s) - English
Resource type - Journals
eISSN - 2289-649X
pISSN - 1675-7009
DOI - 10.24191/srj.v6i1.5636
Subject(s) - dice , computer science , software , string (physics) , closeness , similarity (geometry) , line (geometry) , gauge (firearms) , process (computing) , n gram , artificial intelligence , engineering drawing , engineering , mathematics , programming language , statistics , mathematical analysis , image (mathematics) , geometry , archaeology , mathematical physics , history , language model
Manual marking of free-response solutions in mathematics assessments is very demanding in terms of time and effort. Available software equipped with automated marking features to mark open-ended questions has very limited capabilities. In most cases the marking process focuses on the final answer only. Few available software are capable of marking the intermediate steps as is norm in manual marking. This paper discusses the line-by-line marking performance of the n_gram string similarity method using the Dice coefficient as means to measure similarity. The marks awarded by the automated marking process are compared with marks awarded by manual marking. Marks awarded by manual marking are used as the benchmark to gauge the performance of the automated marking technique in terms of its closeness to manual marking.