Spotify has obtained a patent for a new technology that can recommend songs to users based on analysis of user voice data. The global audio and music streaming media giant first applied for this patent in February 2018 and was authorized by the United States Patent and Trademark Office (USPTO) this month. The patent outlines technologies that can recognize various personal characteristics (such as the user’s “emotional state”) and suggest content based on this model. Previously, Spotify was granted another quite disturbing patent, which involved planning the user’s Spotify experience based on the user’s personality characteristics.

The patent, titled “Identifying Taste Attributes from Audio Signals,” describes a method for processing audio signals input by users (including voice content and background noise) to identify playable content based on input.

In the patent specification, the Swedish audio and music streaming giant explained: “Media streaming applications usually include the function of providing users with personalized media recommendations.” The common method Spotify uses to identify listeners’ music tastes is to use basic information such as age or gender.

This new patent takes it one step further and uses voice recognition technology to “collect user taste attributes.” This patented technology is designed to detect the “intonation, pressure, rhythm, etc. of the voice unit” and combine it with the “sound information in the hidden Markov model architecture” so that Spotify can classify the user’s mood as ” happy”. , Angry, sad or neutral”.

Spotify analyzed in detail the emotional state, gender, age and accent of the speaker who provided the voice. The metadata collected by the technology will also indicate the number of people in the environment where the audio signal is input.

See also  Legion of the Dead release date set on May 21 on Netflix

Collect and analyze all this metadata, as well as Spotify listener’s “previous requests”, “listening and rating history”, “link to associated personal data (such as the user’s friend or colleague’s personal data)” and “user’s Existing music collection and/or library”.

According to the Spotify patent application, the output will be the audio output of the music corresponding to the preference, or the display of the recommended next music track corresponding to the preference. The output can also be a simple suggestion on a visual display.

Spotify has applied for another patent earlier, the technology can recommend music to users based on their personal characteristics. The patent aims to classify users based on different personality models such as openness, friendliness, extroversion, neuroticism and conscientiousness. Because of the seemingly unpleasant personality classification method used by Spotify, this type of classification has attracted people’s attention.

The audio and music streaming giant is known for its algorithms that suggest new audio content to users. Although the new Spotify patent is not surprising, the effectiveness or problematic factors brought about by the new technology are only known to users when it enters the market.

Readers should note that the company has conceptualized its patented technology, but may not use the concept in the market. Therefore, we urge readers to obtain Spotify’s latest patent and point out that this patent sees the company as the only possibility in the future.


Is LG Wing’s unique design enough to help it succeed in India? We discussed it on the weekly technical podcast Orbital, you can subscribe via Apple Podcast, Google Podcast or RSS, download the episode, or click the play button below.