Learning Sentential Patterns of Various Rhetoric Moves for Assisted Academic Writing
Jim Chang, Hsiang-Ling Hsu, Joanne Boisson, Hao-Chun Peng, Yu-Hsuan Wu, Jason S. Chang · 2015
We introduce a new method for extracting rep-resentative sentential patterns from a corpus for the purpose of assisting ESL learners in academic writing. In our approach, sentences are transformed into patterns for statistical analysis and filtering, and then are annotated with relevant rhetoric moves. The method in-volves annotating every sentence in a given corpus with part of speech and base phrase information, converting the sentence into for-mulaic patterns, and filtering salient patterns for key content words (verbs and nouns). We display the patterns in the interactive writing environment, WriteAhead, to prompt the user as they type away. 1