Wikipedia, the free online encyclopedia, has partnered with the swedish KTH (Kungliga Tekniska Högskolan) Royal Institute of Technology to develop a crowdsourced speech engine for visually impaired users.
By September 2017 people with reading difficulty or visual impairment will be able to access Wikipedia and get articles read in a Swedish, English, and an Arabic voice.
The new Wikispeech platform will be optimized for Wikipedia but published as open source, therefore every site that uses the MediaWiki software should be able to integrate Wikispeech.
Naturally, Swedish and English speech engines will be easier to develop while Arabic proves more difficult due to the sparse open source resources currently available. So expect the Arabic version to be a rudimentary one. While Arabic is the fastest growing language on the Internet, Arab speaking users tend to produce far less content in their native language than other nations.
According to Wikimedia Sweden, 125 million users per month (or 25% of overall users) would prefer a spoken version of the website(s). Once synthesized speech works in Swedish, English, and Arabic it could be extended to all of Wikipedia’s 280 languages. Of course, Wikipedia would rely heavily on the help of its users to get such a huge task off the ground.
All material produced will be freely licensed and can be used for free by anyone, in line with the rules of Wikimedia Commons.
Although Wikimedia insists that it is in no way interested in competing with Google, the parallels and rivalries of Wikipedia and Google’s projects like the Knowledge Graph and Google Translate are evident.
If you’re interested in learning more about crowdsourcing in education and how to make it work, visit our partner site EDUKWEST. They published an article with various interesting examples of crowdsourced content in the education space.
Picture by Susanne Nilsson via Flickr