Grants:Impact/Diversity and Inclusion/Wikispeech

From Meta, a Wikimedia project coordination wiki
Impact of Grants
Becoming inclusive of the visually impaired
Location:  Sweden
Grantee:  Wikimedia Sverige
Grant information
  • Program: Annual Plan Grants
  • Fiscal year: 2015 and ongoing

Wikimedia projects are built on the assumption that a person can see.

According to the World Health Organization (WHO), an estimated 253 million people live with a vision impairment.[1] Factors which contribute to the proliferation of visual impairment include lack of access to health care, lack of educational resources, and lack of government involvement, including funding.

Historically, being visually impaired meant being confined, with no access to education or employment.[2] While some countries have made improvements, people with physical disabilities are still marginalized and isolated in small and large ways by their communities, their employers, their institutions, and the public.

Wikimedia projects are not exempt from this marginalizing behavior. “Anyone can edit” is our motto, but all contribution online depends on a person’s ability to see. Reading. Writing. Translating. Uploading. Discussing. These are the main activities of editors on our projects, and yet none of them are friendly to the visually impaired.

In 2015, John Andersson of Wikimedia Sverige (Sweden) proposed a solution: an open-source text-to-speech software that would be integrated with Wikimedia projects. The idea came from a personal experience of John’s while in university, working with individuals with different disabilities. One popular and appreciated tool among them was a service in Sweden that read news out loud and was sent in audio format. This seemingly simple service helped people feel included and engaged with their country and current events.

From this idea, Wikispeech was born.

Wikispeech is a text-to-speech solution developed through grants from Wikimedia Foundation and the Swedish Post and Telecom Authority (PTS), and in partnership with STTS (Speech Technology Services) and KTH Royal Institute of Technology – Speech, Music and Hearing. It allows Wikimedia projects to be more accessible for people that have difficulties reading, by converting text into spoken voice output. Wikispeech is a server based solution, meaning that no software is required to install it, allowing local access.

These kind of text-to-speech solutions benefit those with visually impairment, but they also benefit those who struggle to read or cannot read their own language. As such, Wikispeech has received interest from Wikimedia communities with low literacy, such as Odia, Bengali, Telugu communities, and from WikiMed, the Wikimedia organization focused on translating medical knowledge into languages and communities with minimal access to it. To date, Wikispeech has been piloted as an extension for MediaWiki (the underlying platform of all Wikimedia projects), and has received positive feedback from Wikimedia communities, organizations promoting disability access, and groups working with marginalized or oral cultures around the world.


While Wikispeech isn’t fully operationally, the potential long-term impact of the project can be partially gleaned from looking at the effect of other text to speech software.

The Web Accessibility Initiative, has documented several testimonies of the impact of text to speech software, including employment opportunities and retention, advancement within organizations, and being connected to the world as a whole. The benefits ot text to speech are well documented, and provide accessibility for several different populations, including but not limited to: people with learning disabilities, literacy difficulties, individuals who speak a language but do not read it, and people with different learning styles.

Wikispeech, when fully operationally will have benefits for all user groups, creating tangible pathways specifically for individuals with reading disabilities.


References[edit]