Disturbance Suppressor based Pelican Optimization Algorithm and Self-moving Window Long Short-Term Memory for English Grammar Error Detection

The English language is a common platform in the world which connects every person in this world. Advancement in technologies has introduced various machine translators where the peoples are able to obtained the translated version of any languages. However, these translators are unable to find the English Grammar errors during the translation. The existing models tried to detect the English Grammar error in the machine translators but failed to get the expected performance due to inappropriate extraction and selection of features. To overcome the problem of existing methods for the English Grammar error detection, the Disturbance Suppressor based Pelican Optimization Algorithm and Self-moving Window Long Short-Term Memory (DS-POA and SWLSTM) was used improve the error detection performance in machine translators. The subword skip n-gram was used in the feature extraction process to transform the text data to vector form. The DS-POA was used in feature selection process to obtain optimal features from the data. The SWLSTM interpreted the patterns of features by capturing the long-range information of the features and detected the English Grammar error. The proposed DS-POA and SWLSTM has obtained 98.95% accuracy, 99.37% precision, 99.49% recall, and 99.42% f1-score on English Grammar dataset compared to existing traditional methods.