Title :
A special parser for learning English composition - Error analysis & learners´ model for ILTS
Author :
Chen, Liang ; Tokuda, Naoyuki
Author_Institution :
Comput. Sci. Dept., Univ. of Northern British Columbia, Prince George, BC, Canada
Abstract :
By embedding the function of automatically correcting nearly free format English translations of given Japanese sentences, we have developed a new ILTS(intelligent language tutoring system) for improving English Writing Skills in a L2 tutoring environment. In addition to a diagnostic engine capable of identifying grammatical, lexical and word usage errors of students´ input translations and returning error contingent feedback, we have developed a simple table look-up parser for displying the grammar trees. The table look-up parser parses a user input, which is always erronous, by simply matching the extended part-of-speech tag sequence of a closing sentence in a template of ILTS to the entries of a look-up-table, in which each extended part-of-speech tag array corresponds to one grammar tree. The complexity of the table look-up parser is O(n), where n denotes the length of a sentence. This shows a marked improvement over the general purpose parser of which the complexity is O(n3).
Keywords :
computational complexity; grammars; intelligent tutoring systems; language translation; learning (artificial intelligence); linguistics; natural language processing; trees (mathematics); English Writing Skills; English translations; Japanese sentences; L2 tutoring environment; error analysis; error contingent feedback; grammar trees; intelligent language tutoring system; learners model; learning English composition; machine learning; machine translation; natural language processing; part-of-speech tag sequence; table look-up parser; Cybernetics; Error analysis; Machine learning; intelligent language learning system; machine learning; machine translation; natural language processing;
Conference_Titel :
Machine Learning and Cybernetics, 2009 International Conference on
Conference_Location :
Baoding
Print_ISBN :
978-1-4244-3702-3
Electronic_ISBN :
978-1-4244-3703-0
DOI :
10.1109/ICMLC.2009.5212797