Logo image
Mining Collocation-Syntax Patterns for Assisted Academic Writing
Thesis

Mining Collocation-Syntax Patterns for Assisted Academic Writing

Huang, Guan Cheng
Masters, 國立清華大學, 資訊工程學系
2014

Abstract

自然語言處理 學術英文寫作 計算輔助語言學習 Natural Language Processing Academic Writing Computer Assisted Language Learning
We introduce a new method for learning to find collocation-syntax patterns from a corpus for the purpose of assisting English as a Foreign Language (EFL) learners in writing fluently and avoiding errors. In our approach, each word in sentences is transformed into elements aimed at generating candidate patterns for statistical analysis and filtering. The method involves generating candidate patterns from elements of each word, extracting patterns matching predefined templates, and filtering salient collocations and complementational patterns for each word that we focus on. We extract grammatical patterns, collocation patterns, and collocation-syntax patterns to provide writing suggestions. We present an interactive writing environment, WriteAhead, that automatically displays examples grouped by patterns to prompt the users as they write away, or mouse around a draft during self-editing. Preliminary experiments and evaluation show that WriteAhead with the acquired patterns and examples has the potential to improve writing and improve writing skills in the long run.

Metrics

1 Record Views

Details

Logo image