• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site

Book chapter

Correcting collocation errors in learners’ writing based on probability of syntactic links

P. 58-68.
Bolshakova E. I., Azimov A.
The paper describes a novel method for automatic collocation error correction in NL texts that are written by language learners or are translated from another NL with the aid of machine translators. We assume that the main reason of collocation errors is the strategy of word-by-word translation used by authors of the texts or the machine translators, so the errors essentially depends on the source language. While processing a sentence from the text, the method considers as potential correcting variants all its paraphrases that have the same syntactic structure and are built by replacing of all sentence words by their substitute words. Substitute words are automatically generated using word translation equivalents taken from a particular translation dictionary. For detecting an error in the sentence, we propose a relevance degree function computed from the probability of word syntactic links and applied to the sentence and its paraphrases. If the value of the function for the sentence is less than for some of its paraphrases, our method signals an error, then it is corrected by appropriate sentence paraphrase. The method was approved for correction collocation errors in English texts written by Russian speakers. Stanford Parser and English text collection were used to gather statistics and compute probability of English word syntactic links. Within certain limitation, the experiments gave promising results: there were detected about 80% of collocation errors (with words of various POS) and 87% of proposed correcting paraphrases included “gold” correction.