Methods and systems for personalizing content in accordance with divergences in a user's listening history

• US 10,412,183 B2
• Filed: 08/23/2017
• Issued: 09/10/2019
• Est. Priority Date: 02/24/2017
• Status: Active Grant

• Current Assignee: Spotify AB (Spotify Technology SA)
• Original Assignee: Spotify AB (Spotify Technology SA)
• Inventors: Gibson, Clay, Shapiro, Will, Gil, Santiago, Anderson, Ian, Mpossi, Margreth, Semerci, Oguz, Wolf, Scott
• Primary Examiner(s): Bengzon, Greg C
• Application Number:
  US15/684,841
• Publication Number:
  US 20180248965A1
• Time in Patent Office: 748 Days
• CPC Class Codes: G06F 16/435 Filtering based on additional data, e.g. user or group profiles
• CPC Class Codes: G06F 16/635 Filtering based on additional data, e.g. user or group profiles
• CPC Class Codes: H04L 65/61 for supporting one-way streaming services, e.g. Internet radio
• CPC Class Codes: H04L 67/01 Protocols
• CPC Class Codes: H04L 67/52 specially adapted for the location of the user terminal
• CPC Class Codes: H04L 67/535 Tracking the activity of the user network monitoring arrangements H04L43/00; recording of computer activity G06F11/34

Abstract

An electronic device accesses a listening history of a user of a media-providing service, and for each characteristic of a set of characteristics describing tracks provided by the media-providing service, determines a measure of divergence with respect to listening contexts for the listening history of the user. The electronic device identifies a characteristic that has a respective measure of divergence that satisfies a threshold indicative of a lack of uniformity in the listening history with respect to the listening contexts. The electronic device determines a representative value of the identified characteristic for a listening context corresponding to a current listening context of the user. The electronic device provides personalized content to a client device of the user based on the current listening context and representative value of the identified characteristic.

Claims

1. A method, comprising

at an electronic device associated with a media-providing service and having one or more processors and memory storing instructions for execution by the one or more processors:

accessing a listening history of a user of the media-providing service;

for each characteristic of a set of characteristics describing tracks provided by the media-providing service, determining, for the user, a measure of divergence indicating a degree of lack of uniformity in the listening history of the user with respect to listening contexts, the determining including normalizing, based on the listening history, a calculated divergence for the characteristic relative to measures of divergence of other users in the listening contexts, wherein each characteristic has a plurality of possible values;

identifying a characteristic, of the set of characteristics, having a respective measure of divergence that satisfies a threshold indicative of the lack of uniformity in the listening history with respect to the listening contexts;

determining a representative value of the identified characteristic for a listening context corresponding to a current listening context of the user;

in the current listening context, providing personalized content to a client device of the user based on the representative value of the identified characteristic, including adjusting an order of tracks within a playlist to promote a track within the playlist, the track having the representative value of the identified characteristic for the listening context corresponding to the current listening context and

streaming one or more tracks within the playlist to the client device of the user in accordance with the adjusted order.

2. The method of claim 1, wherein the set of characteristics includes genre and mood.

3. The method of claim 1, wherein the set of characteristics includes one or more characteristics selected from the group consisting of a degree to which a track is acoustic, a strength of a beat, a degree to which a track sounds choppy, suitability for dancing, positions of a track where downbeats occur, duration of a track, average dynamic range, a measure of energy of a track, a degree to which the twelve pitches are balanced, a degree to which a track contains vocals, key, a degree to which a live audience is present, loudness, a degree of consistency of timing, fade-in point, fade-out point, modal scale, suitability for running, a degree to which a track includes spoken words, tempo, time signature, and a degree to which a track is cheerful.

4. The method of claim 1, wherein:

the listening contexts include respective time periods during a day; and

the current listening context includes a current time of day.

5. The method of claim 1, wherein:

the listening contexts include respective days of a week; and

the current listening context includes a current day of the week.

6. The method of claim 1, wherein:

the listening contexts include locations of the user; and

the current listening context includes a current location of the user.

7. The method of claim 1, wherein:

the listening contexts include weather; and

the current listening context includes current weather at a current location of the user.

8. The method of claim 1, wherein:

the listening contexts include playback devices; and

the current listening context includes a current playback device.

9. The method of claim 1, wherein providing the personalized content comprises forgoing to provide a recommended playlist.

10. The method of claim 9, wherein forgoing to provide a recommended playlist comprises forgoing to provide a playlist in which a number of tracks having the representative value of the identified characteristic does not satisfy a threshold.

11. The method of claim 1, wherein providing the personalized content comprises:

selecting a message for the user based on the representative value of the identified characteristic; and

providing the selected message to the client device of the user.

12. The method of claim 11, wherein the message is associated in the media-providing service with the representative value of the identified characteristic.

13. The method of claim 1, wherein the calculated divergence comprises a Jensen-Shannon Divergence for the characteristic with respect to the listening contexts.

14. The method of claim 1, wherein identifying the characteristic comprises:

ranking characteristics of the set of characteristics that have measures of divergence that satisfy the threshold by the measures of divergence; and

selecting the characteristic with the greatest measure of divergence as the identified characteristic.

15. The method of claim 1, wherein determining the representative value of the identified characteristic for the listening context corresponding to the current listening context of the user comprises:

determining, for the listening context corresponding to the current listening context of the user, the mode of the plurality of possible values of the identified characteristic in the listening history of the user; and

assigning the mode as the representative value.

18. The method of claim 1, wherein normalizing, based on the listening history, the calculated divergence for the characteristic relative to measures of divergence of other users in the listening contexts is performed in accordance with a determination that the user has listened to tracks provided by the media-providing service for a threshold amount during a preceding time window.

16. An electronic device, comprising:

one or more processors; and

memory storing one or more programs for execution by the one or more processors, the one or more programs including instructions for;

accessing a listening history of a user of the media-providing service;

for each characteristic of a set of characteristics describing tracks provided by the media-providing service, determining, for the user, a measure of divergence indicating a degree of lack of uniformity in the listening history of the user with respect to listening contexts, the determining including normalizing, based on the listening history, a calculated divergence for the characteristic relative to measures of divergence of other users in the listening contexts, wherein each characteristic has a plurality of possible values;

identifying a characteristic, of the set of characteristics, having a respective measure of divergence that satisfies a threshold indicative of the lack of uniformity in the listening history with respect to the listening contexts;

determining a representative value of the identified characteristic for a listening context corresponding to a current listening context of the user;

in the current listening context, providing personalized content to a client device of the user based on the representative value of the identified characteristic, including adjusting an order of tracks within a playlist to promote a track within the playlist, the track having the representative value of the identified characteristic for the listening context corresponding to the current listening context and

streaming one or more tracks within the playlist to the client device of the user in accordance with the adjusted order.

17. A non-transitory computer-readable storage medium, storing one or more programs for execution by one or more processors of an electronic device, the one or more programs including instructions for:

accessing a listening history of a user of the media-providing service;

for each characteristic of a set of characteristics describing tracks provided by the media-providing service, determining, for the user, a measure of divergence indicating a degree of lack of uniformity in the listening history of the user with respect to listening contexts, the determining including normalizing, based on the listening history, a calculated divergence for the characteristic relative to measures of divergence of other users in the listening contexts, wherein each characteristic has a plurality of possible values;

identifying a characteristic, of the set of characteristics, having a respective measure of divergence that satisfies a threshold indicative of the lack of uniformity in the listening history with respect to the listening contexts;

determining a representative value of the identified characteristic for a listening context corresponding to a current listening context of the user;

in the current listening context, providing personalized content to a client device of the user based on the representative value of the identified characteristic, including adjusting an order of tracks within a playlist to promote a track within the playlist, the track having the representative value of the identified characteristic for the listening context corresponding to the current listening context and

streaming one or more tracks within the playlist to the client device of the user in accordance with the adjusted order.