Colloquial expressions, abbreviations and slang are integrated into daily language for teenagers. Such behavior may lower the Chinese language writing proficiency of many students. In view of the problem, CUHK developed an error detection system called ACT, which is based on Big Data Mining and Deep Learning, and in conjunction with a unique intelligent algorithm and a novel scoring mechanism. ACT is the first of its kind in Hong Kong and targets local students. It took just a few seconds to analyze an article with thousand of words. For every detected error, ACT will also try to offer replacement suggestions.