Sage Concept Grants: Swara

Swara is one of the 5 winners of the £2,000 Sage Concept Grant in 2023.

Mrs Husnul Fitri, School of Strategic and Global Studies, University of Indonesia

Swara (the Sankrit’s word for “sound, voice or tone”) is a web-based/mobile self-completion survey software/application that utilizes voice user interface technology. Swara advances disability-friendly innovations to social science research tools that benefits both persons of disability and researchers. Unlike conventional data collection, Swara employs voice overs to read survey questions, records and analyzes (i.e. sorting, categorizing) speech-based responses, and displays data on spreadsheets. This technology will facilitate people with disabilities, particularly those who are visually impaired and have limited hand mobility, to participate more conveniently in survey research. It will also facilitate researchers to collect and analyze data from persons of disability more quickly.

I was extremely grateful and honored to be announced as a recipient of Sage Concept this year's winners. Swara is inspired by the notion that technological advancements in the research field should benefit a more diverse population, allowing us to better comprehend our society. On the other hand, technology must play a significant role in facilitating research without sacrificing its quality.

Swara is not a completely novel technology, as it has already been utilized in the revolution of speech-to-text technology. However, the essence of this voice recognition technology can be expanded in the research field by adapting it to our needs and the population that will be greatly aided by it.

Since submitting Swara to Sage Concept Grants, I've been talking with colleagues who are experts in AI about the possibility of bringing this technology to real life. Obviously, more time is required to discuss the optimal form of Swara in order for it to achieve its intended purpose as an application that will aid researchers and benefit a specific population. There are several aspects that require exploration and careful consideration: What is the most optimal program required to establish a connection between voice recognition and text as command code, with the purpose of categorizing, summarizing, and processing the voice response as data sources according to research need? To what extent do variations in the human voice, such as timbre, pitch, modulation, impact Swara's capacity to accurately interpret and provide suitable responses? What kind of alternative response or option is required for Swara to anticipate the problem of participant fatigue?   

Undoubtedly, those factors undertake considerable challenges. More collaboration and discussion is needed to create Swara so that it can really speak in the actual world . Therefore, I am presently inviting for those interested in Swara development to provide feedback and suggestions with the aim of testing this concept and eventually building a prototype for Swara to practical application. Feedback can be provided by using the provided form, and any contributions made will be much appreciated.

Let's give Swara a Voice!