Developing an open-source, rule-based proofreading tool

摘要

In this paper, we show how an open-source, language-independent proofreading tool has been built. Many languages lack contextual proofreading tools; for many others, only partial solutions are available. Using existing, largely language-independent tools and collaborative processes it is possible to develop a practical style and grammar checker and to fight the digital divide in countries where commercial linguistic application software is unavailable or too expensive for average users. The described solution depends on relatively easily available language resources and does not require a fully formalized grammar nor a deep parser, yet it can detect many frequent context-dependent spelling mistakes, as well as grammatical, punctuation, usage, and stylistic errors.