期刊名称：Anais do Workshop de Desafios da Computação Aplicada à Educação
出版年度：2012
卷号：0
期号：0
页码：10-19
语种：Portuguese
出版社：Anais do Workshop de Desafios da Computação Aplicada à Educação
摘要：In order to attend the virtual learning environment needs, this paper presents the LSA (Latent Semantic Analysis) application to estimate scores au- tomatically in open ended questions, because still there is not a method with a acceptable accuracy for practical uses. It was developed a model based in LSA composed of six steps: pre-processing, normalization, weighting, singular value decomposition, classification and model fitting. As result obtained a study about the LSA parameters adjustment for practical uses in automatic assessment of discursive responses, and as best result was obtained a accuracy of 87,35%
其他摘要：In order to attend the virtual learning environment needs, this paper presents the LSA (Latent Semantic Analysis) application to estimate scores au- tomatically in open ended questions, because still there is not a method with a acceptable accuracy for practical uses. It was developed a model based in LSA composed of six steps: pre-processing, normalization, weighting, singular value decomposition, classification and model fitting. As result obtained a study about the LSA parameters adjustment for practical uses in automatic assessment of discursive responses, and as best result was obtained a accuracy of 87,35%
