Only keywords, terms and n-grams whose %[frequency]% in the focus corpus is equal or lower than this setting will be included in the result.