Since this was not accepted as a GSoC project, it has been moved to a volunteer project: Wikispeech (Tool for suggesting missing lexicon entries).
Wikispeech will have a crowdsourcing part so that volunteers can help improve the text-to-speech experience. Volunteers will be able to improve the lexicon in a few different ways (see M197). For the volunteer that want to make the most useful contributions (with limited time on their hands) we see a need to help with prioritizing what to focus on.
One great tool would be an option to highlight what words would be most valuable to add in the lexicon - order to improve the text-to-speech experience for a specific article (e.g. the name of the place the article is about). Then the volunteer could quickly add a few words to the lexicon through the Wikispeech editor and a valuable contribution has been made.
The code for the project can be found here: https://github.com/wikimedia/mediawiki-extensions-Wikispeech
Possible mentor: @Sebastian_Berlin-WMSE
Co-mentors: @Lokal_Profil
Skills required: analysis, basic software architecture, development
Microtasks: