Problem: Search engine is not made for dictionary needs.
Who would benefit: Contributors and readers
Proposed solution: Having an internal anagram and advanced search such as the one we have in French Wiktionary, to find anagrams but also words with a specific sound or with a sequence of letter and a grammatical class for example. It could be based on a parsed dump to have part of speech distinction as well (only in nouns, only in verbs, etc.).
I feel this proposal would mainly target readers, as there is plenty website that reuse Wiktionary database to offer such service of adapted search tool. It is a real need. A UX study on that should provide more evidence of this. Noé (talk) 11:47, 15 November 2019 (UTC)[reply]
The pronunciation-based researches would need to develop an ETL to parse each Wiktionary differently and store each word pronunciation into a new table with an index. FYI, I have already created a draft project which does this extraction in PHP from the fr.wikt and en.wikt XML dumps, along with languages codes and grammatical classes. But I didn't publish the code for now, I'm still choosing a front-end interface to browse these tables. JackPotte (talk) 13:59, 21 November 2019 (UTC)[reply]