29 mars 2007
info:eu-repo/semantics/OpenAccess
Sonia Mandin et al., « Modeling summarization assessment strategies with LSA », HAL-SHS : sciences de l'éducation, ID : 10670/1.yetun2
This paper presents a model based on LSA which attempts to simulate the way humans assess student summaries. It is based on the automatic detection of 5 cognitive operations that student may use in writing a summary. Comparisons with dataand the construction of a new proposition denoting a global from 33 human raters show the strengths and limits of this approach.