Privacy-preserving voice-based search over mHealth data

Abstract Voice-enabled devices have a potential to significantly improve the healthcare systems as smart personal assistants. They usually come with a hands-free feature to add an extra level of usability and convenience to elderly, disabled people and patients. In this paper, we propose a privacy-preserving voice-based search scheme to enhance the privacy of in-home healthcare applications. We consider an application scenario where patients use the devices to communicate with their caregivers by recording and uploading their voices to the servers, where the caregivers can search the interested voices of their patients based on the voice content, mood, tone and background sounds. Our scheme preserves the richness and privacy of voice data and enables accurate and efficient voice-based search, while in current systems that use speech recognition, the richness and privacy of voice data are compromised. Specifically, our scheme achieves the privacy by employing a privacy-preserving voice feature matching technique and a novel category-based encryption; only encrypted voice data is uploaded to the server who is unable to access the original voice data. In addition, our scheme enables the server to selectively and accurately respond to caregivers' queries on the voice data based on voice similarities. We evaluate our scheme through real experiments and show that our scheme even with privacy preservation can successfully match similar voice data at an average accuracy of 80.8%.