Seeding and generating suggested content collections

Patent metadata

Field	Value
Publication number	US-11947601-B2
Application number	US-202217815478-A
Country	US
Kind code	B2
Filing date	Jul 27, 2022
Priority date	Jul 27, 2022
Publication date	Apr 2, 2024
Grant date	Apr 2, 2024

Abstract

The present disclosure is directed toward systems, methods, and non-transitory computer readable media for generating and suggesting content collections for user accounts of a content management system using combinations of content-based features such as textual signals and visual signals. In some embodiments, the disclosed systems select a seed content item from among a plurality of content items associated with a user account within a content management system. From the seed content item, the disclosed systems can determine one or more germane topics and can cluster additional content items in relation to the germane topic(s). In addition, the disclosed systems can select one or more content items from a content cluster to provide as a suggested content collection.

First claim

What is claimed is:

1. A computer-implemented method comprising:
selecting one or more candidate seed content items for seeding generation of a content collection from a plurality of content items associated with a user account of a content management system;
generating, based on the one or more candidate seed content items, one or more content clusters from the plurality of content items associated with the user account by:
determining a first relevance score based on comparing a topic for a content item from the plurality of content items with a germane topic of one or more candidate seed content items;
determining a second relevance score based on comparing an object classification for the content item from the plurality of content items with the germane topic of the one or more candidate seed content items; and
generating a combined relevance score by utilizing a hybrid relevance model to combine the first relevance score and the second relevance score;

identifying a suggested content collection for the user account from a content cluster determined by the combined relevance score from among the one or more content clusters seeded by the one or more candidate seed content items; and
providing a notification corresponding to the suggested content collection for display on a client device associated with the user account.

2. The computer-implemented method of claim 1 , wherein selecting the one or more candidate seed content items for seeding generation of the content collection comprises:
determining relevance scores for one or more content items within the plurality of content items associated with the user account; and
selecting the one or more candidate seed content items based on the relevance scores for the one or more content items.

3. The computer-implemented method of claim 1 , wherein selecting the one or more candidate seed content items for seeding generation of the content collection comprises utilizing a seed prediction machine learning model to predict a content item from among the plurality of content items that the user account will access.

4. The computer-implemented method of claim 1 , wherein generating the one or more content clusters comprises:
determining a germane topic from a candidate seed content item of the one or more candidate seed content items based on relevance to the user account; and
clustering the plurality of content items in relation to the germane topic of the candidate seed content item.

5. The computer-implemented method of claim 1 , wherein generating the one or more content clusters comprises:
determining an object classification for an object depicted within a candidate seed content item of the one or more candidate seed content items; and
clustering the plurality of content items in relation to the object classification of the candidate seed content item.

6. The computer-implemented method of claim 1 , wherein identifying the suggested content collection comprises:
ranking the one or more content clusters seeded by the one or more candidate seed content items based on relevance to the user account;
selecting, based on ranking the one or more content clusters, a content cluster from the one or more content clusters for generating the suggested content collection; and
generating, for the suggested content collection, a virtual folder comprising references to one or more content items within the content cluster selected based on ranking the one or more content clusters.

7. The computer-implemented method of claim 1 , wherein providing the notification corresponding to the suggested content collection comprises providing, for display on the client device, an interface element selectable to, with a single client device interaction:
generate a virtual folder for a new content collection associated with the user account within the content management system; and
add references to one or more content items within the suggested content collection to the virtual folder of the new content collection.

8. A system comprising:
at least one processor; and
a non-transitory computer readable medium comprising instructions that, when executed by the at least one processor, cause the system to:
select a seed content item for seeding generation of a content collection from a plurality of content items associated with a user account of a content management system;
generate a content cluster seeded by the seed content item comprising a set of content items from the plurality of content items associated with the user account by:
determining a first relevance score based on comparing a topic for a content item from the plurality of content items with a germane topic of the seed content item;
determining a second relevance score based on comparing an object classification for the content item from the plurality of content items with the germane topic of the seed content item; and
generating a combined relevance score by utilizing a hybrid relevance model to combine the first relevance score and the second relevance score;

generate, based on the combined relevance score, a suggested content collection for the user account from the content cluster seeded by the seed content item; and
provide a notification corresponding to the suggested content collection for display on a client device associated with the user account.

9. The system of claim 8 , further comprising instructions that, when executed by the at least one processor, cause the system to select the seed content item for seeding generation of the content collection by utilizing a seed prediction machine learning model trained to predict content items that the user account will access to select the seed content item from a plurality of candidate seed content items.

10. The system of claim 8 , further comprising instructions that, when executed by the at least one processor, cause the system to generate the content cluster seeded by the seed content item by clustering the set of content items based on:
extracting topics from the plurality of content items;
determining classifications for objects depicted within the plurality of content items; and
comparing the topics and the classifications with a germane topic associated with the seed content item to identify the set of content items as corresponding to the seed content item.

11. The system of claim 10 , further comprising instructions that, when executed by the at least one processor, cause the system to generate the content cluster seeded by the seed content item by further identifying co-access patterns between the seed content item and the set of content items within the content cluster.

12. The system of claim 8 , further comprising instructions that, when executed by the at least one processor, cause the system to:
receive a client device interaction from the client device declining the suggested content collection; and
based on the client device interaction declining the suggested content collection, generate a new suggested content collection that thematically differs from the suggested content collection.

13. The system of claim 8 , further comprising instructions that, when executed by the at least one processor, cause the system to:
identify at least one archival content item within the plurality of content items; and
exclude the at least one archival content item from the suggested content collection.

14. The system of claim 8 , further comprising instructions that, when executed by the at least one processor, cause the system to generate the suggested content collection by identifying a subset of content items from the content cluster to suggest based

Assignees

• Dropbox Inc

Inventors

• Baek Jongmin
• Ding Jiarui

Classifications

• G06F16/906 Primary

  Clustering; Classification · CPC title

• G06F3/0482

  Interaction with lists of selectable items, e.g. menus · CPC title

• G06F16/9035

  Filtering based on additional data, e.g. user or group profiles · CPC title

• G06N20/00 Primary

  Machine learning · CPC title