I am initiating a funded research project on helping Chinese ESL learners improve their writings , e.g. locating grammatical errors and correcting semantically misusing words. We have already gotten a ESL corpus but it contains a lot of tagging mistakes. Would you please point out any related work or tool or corpus? If you did such kind of research before, welcome any of your research data and publications.
I am afraid that this will be very hard to do with an automated system.
Of course, Microsoft Word has basic grammar checking ability, but as a native speaker who writes at an advanced level, I find that Word often makes incorrect suggestions.
The only sure way would be to have native speakers review the writing...and even they might disagree on some points of grammar or word use due to the complexities of the English language.
Thanks. how native speakers make reviews is important in deciding the upper bound of a system.
Sure, I can understand the universal ambiguity on each level of language use. Within a specific range of ESL/EFL learners, should we try an imperfect but appropriate solution to the issue?
Добрый день! Я, к сожалению, не знаю английского языка. Но я занимаюсь обучением китайских студентов русскому языку как иностранному. Имею большое количество материалов для исследования.
Да. Россия. Томск. Мои студенты (3 курс) готовы помогать в исследовании, могут переводить с русского на китайский. Студенты из Урумчи.
Thanks for your input.
As for the goal of artificial ESL/EFL adviser, simulation never goes beyond replacement.
SLA theory and practice does not subscribe to 'error' or 'mistake' correction. Learner 'approximative'/developmental systems are considered as fully-fledged grammars in themselves, though they have been based upon erratic learner hypotheses. What a tutor needs to do is to design a program of 'un-marking' the marked areas of any aspect of Language, since these are known to correlate with learning difficulty. Provision of more and more positive evidence of appropriate forms is probably the only effective intervention a tutor can give; this is because learners have their own inbuilt cognitive system of processing language data.
I am with dr Griggs. One must correct and discuss the corrections with the student if possible. It sounds and is old fashioned, but works best.
When it comes to speaking and writing, there is no evidence to prove that a correct-all or any inconsistent haphazard method will lead to any type of improvement. However, there is research to show that targeted and consistent types of correction are useful in both speaking and writing.
I recommend to consult Dana Ferris' work on error correction in L2 writing.
I do agree that Grammarly is helpful fro advanced writers, though we do not know if it leads to improvement.
Check the work of Granger, S. https://www.researchgate.net/profile/Sylviane_Granger/publications
Error correction in Second Language acquisition and learning is not advisable. This is because research in SLA has proven that there is little or no value in negative evidence, or correction, in the natural process of acquisition. What the teacher of a second language should do is to present increased positive evidence through purposefully establishing authentic communicative situations in which learners engage. Simulated, cybernetic exercises may be of little value; this is because they may be 'thrice-removed' from the communicative reality.
The correction in a second language adquisition should be pragmatically oriented.
- Badges
- Science topic
- Similar topics
- Correction