TY - JOUR
T1 - Automated essay scoring using incremental latent semantic analysis
AU - Zhang, Mingqing
AU - Hao, Shudong
AU - Xu, Yanyan
AU - Ke, Dengfeng
AU - Peng, Hengli
PY - 2014
Y1 - 2014
N2 - Writing has been increasingly regarded by the testers of language tests as an important indicator to assess the language skill of testees. As such tests become more and more popular and the number of testees becomes larger, it is a huge task to score so many essays by raters. So far, many methods have been used to solve this problem and the traditional method is Latent Semantic Analysis (LSA). In this paper, we introduce a new incremental method of LSA to score essays effectively when the dataset is massive. By comparison of the traditional method and our new incremental method, concerning the running time and memory usage, experimental results make it obvious that the incremental method has a huge advantage over the traditional method. Furthermore, we use real corpora of test essays submitted to the MHK test (Chinese Proficiency Test for Minorities), to demonstrate that the incremental method is not only efficient but also effective in performing LSA. The experimental results also show that when using incremental LSA, the scoring accuracy can reach 88.8%.
AB - Writing has been increasingly regarded by the testers of language tests as an important indicator to assess the language skill of testees. As such tests become more and more popular and the number of testees becomes larger, it is a huge task to score so many essays by raters. So far, many methods have been used to solve this problem and the traditional method is Latent Semantic Analysis (LSA). In this paper, we introduce a new incremental method of LSA to score essays effectively when the dataset is massive. By comparison of the traditional method and our new incremental method, concerning the running time and memory usage, experimental results make it obvious that the incremental method has a huge advantage over the traditional method. Furthermore, we use real corpora of test essays submitted to the MHK test (Chinese Proficiency Test for Minorities), to demonstrate that the incremental method is not only efficient but also effective in performing LSA. The experimental results also show that when using incremental LSA, the scoring accuracy can reach 88.8%.
KW - Automated essay scoring
KW - Incremental latent semantic analysis
KW - Singular value decomposition
UR - http://www.scopus.com/inward/record.url?scp=84894431865&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84894431865&partnerID=8YFLogxK
U2 - 10.4304/jsw.9.2.429-436
DO - 10.4304/jsw.9.2.429-436
M3 - Article
AN - SCOPUS:84894431865
SN - 1796-217X
VL - 9
SP - 429
EP - 436
JO - Journal of Software
JF - Journal of Software
IS - 2
ER -