Use this option to include a different %[attribute]% or more attributes into the result. For example, lemma, word form and part-of-speech tag. When using this option with a very large corpus, the frequency will be computed from a maximum of 10 million concordnce lines.