A semantic grading model integrated local and global semantic relatedness for English text

This paper proposed a semantic grading model (SGM) for English text based on its content in different views. The SGM not only assesses the global semantic relatedness between ideal text and tested essay, local semantic relatedness between adjoining text segments, as well as the semantic relationship between vary text segments and the whole texts are also taken into account. Local and global semantic relatedness are combined to measure the overall quality of text contents on the basis of semantic. Our experiment shows that the SGM is competent in assessing the content of English text in terms of semantic grading.