The screenshot above shows a Detection Region with three extraction fields: "chinese name", "english name" and "name" that contains the previous two. To filter out parts that are not needed(i.e. no Chinese in English name, and vice versa), they are both set to "Script" where some JavaScript is written to do the job. Since the "Name" field takes both the English and Chinese names, it can simply be set to "Block of English and Chinese".