A New Form of Web Corpus: Display of Search Results Based on English Language Quality

A new form of Web corpus for English writing support is presented. When a Web search engine is used to refer to English sentence samples, identifying suitable samples from among the numerous search results is a troublesome task. In particular, nonnative English speakers have difficulty in determining the quality of English of each search result. The corpus system proposed in this paper evaluates the English writing quality of documents by using fuzzy inference; it displays to the corpus users both English sentence samples and information on the quality of the sentences. This enables users to identify good sentence samples with less time and effort