For collections that have valid text (corrected OCR or proofed input), a "wordwheel" can be created. The XPat indexed text is used by a script to build an index, with counts of occurrences, of individual words that appear in various regions (main text or author or title fields). This data is packaged into SGML and indexed with XPat. The Wordwheel program is then used to access this data. All of this allows a user to "browse" through and find words that may be similar to a word s/he is looking for.