Automatic Correction of English Composition Grammatical Errors Based on Multiple Features

Abstract

An automatic correction method for English composition grammatical errors based on multiple features, including a grammatical error correction preprocessing module, a grammatical error correction model training module, and a grammatical error checking and correction module; wherein the grammatical error correction preprocessing module performs word part-of-speech tagging on the input training text set , sentence syntax analysis and word frequency statistics; the grammar error correction model training module extracts words and their part-of-speech context grammatical features, words and Its part-of-speech structure depends on grammatical features, words and grammatical features of their parts of speech, calculates the weight of word grammatical features, and outputs a grammatical error correction statistical model; the grammatical error checking and correction module uses the grammatical error correction statistical model and the grammatical error correction rule model to treat error-correcting compositions Grammatical error correction, output the correction results of grammatical errors in English compositions. This method realizes the automatic correction of 11 common English grammatical errors in English composition.

Description

technical field [0001] The invention relates to natural language processing technology and statistics, in particular to a method for automatically correcting grammatical errors in English compositions based on multiple features. Background technique [0002] At present, the automatic correction methods of English composition grammatical errors are mainly divided into two categories, namely: the automatic correction method of English composition grammatical errors based on rules and the automatic correction method of English composition grammatical errors based on statistics. Among them, the rule-based automatic correction method for English composition grammatical errors is: by analyzing the grammatical features of English sentences, summarizing and summarizing fixed English grammatical rules to construct an English sentence grammatical rule error correction model, when correcting grammatical errors in English compositions , build an error correction model of English sentenc...

AI Technical Summary

Problems solved by technology

[0005] Disadvantages: Due to the flexibility and variability of English usage, it takes a long time to build a model of English grammar rules, and the summarized English grammar rules can only cover limited grammar rules, and the types of English grammar error correction are limited

[0008] Disadvantages: The correct rate of English grammar error correction is affected by the size of the training text set, the breadth and accuracy of grammatical feature extraction, and the quality of the grammatical statistical error correction model built by training, so the correct rate of error correction of this method needs to be improved

Images