The Swedish based audio streaming application Spotify has been granted a patent for a new technology that can suggest users audio tracks by monitoring their speech. Spotify will monitor the user's intonation, stress, rhythm, and other speech units to know if they are happy, angry, sad or neutral. The technology also monitors environmental metadata and identifies the user's surroundings through ambient sounds. Other basic information like the user's gender is also added in the data pool. The technology uses this data in addition to the user's previous audio track requests, listening and rating history, links associated to profiles of the user's friends or colleagues, their existing music collection etc. to customise the recommendations.
Spotify could play music according to the user's social settings, such as 'alone', 'in a small group' or 'at a party'.
"It is common for a media streaming application to include features that provide personalised media recommendations to a user," the audio streaming company says in the patent description. Instead of inputting answers for multiple queries, a technology that evaluates the user and finds her taste could perform efficiently, says the patent filing document.
Regarding privacy concerns, Spotify clarified in one of its research papers, "We disavow any future research or applications that violate ethical standards of data usage and are not transparent about privacy to its users." Using voice recognition and complex options like creating an 'emotion-tree' and linking the user's information such as environment, age, gender and even accent to collect taste attributes of a user has generated concerns, despite similar algorithms like YouTube's recommendations.
Spotify applied for this patent in February 2018, and it was granted by the United States Patent and Trademark Office this January 2021. It is currently unclear how or when Spotify will implement the technology.