Traitement en cours

Veuillez attendre...

Paramétrages

Paramétrages

Aller à Demande

1. WO2019118032 - PRÉDICTIONS INTELLIGENTES CENTRÉES SUR DES PERSONNES DANS UN ENVIRONNEMENT COLLABORATIF

Note: Texte fondé sur des processus automatiques de reconnaissance optique de caractères. Seule la version PDF a une valeur juridique

[ EN ]

CLAIMS

What is claimed is:

1. A method comprising:

identifying, for a user of a cloud-based content management platform, a plurality of other users of the cloud-based content management platform that have a relationship with the user and are associated with a plurality of documents hosted by the cloud-based content management platform;

predicting, by a processor, one or more collaborators for the user based on collaboration attributes of the plurality of other users; and

providing, by the processor, for presentation to the user, information identifying the predicted one or more collaborators to direct the user to a subset of documents from the plurality of documents hosted by the cloud-based content management platform, the subset of documents each being associated with one of the predicted one or more collaborators.

2. The method of claim 1, wherein the identifying of the plurality of other users of the cloud-based content management platform comprises:

determining the plurality of other users from at least one or more of:

actions of the user associated with the plurality of documents, a calendar, and a communication tool hosted by the cloud-based content management platform, wherein the actions of the user are from a recent predetermined time period;

affiliations of the user in association with the cloud-based content management platform; or

a predetermined number of recently shared documents with the user by the plurality of other users.

3. The method of claim 2, wherein:

the affiliations of the user include one or more of:

a contact list of the user having a list of profiles of users of the cloud-based content management platform, the contact list including a level of affinity with the user for each profile; or

a membership to a group space in the cloud-based content management platform that allows members of the group space to store, search, and access documents in the group space; and

the actions of the user include one or more of:

a collaboration of the user in association with the plurality of documents hosted by the cloud-based content management platform,

a document search action on the plurality of documents hosted by the cloud-based content management platform based on a document owner,

a future collaborative calendar event attendance action, or

a communication action via a communication tool hosted by the cloud- based content management platform.

4. The method of claim 2, further comprising:

determining, for each of the plurality of other users of the cloud-based content management platform, a set of collaboration attributes.

5. The method of claim 4, wherein the set of collaboration attributes of a respective other user comprises at least two or more of:

a frequency of collaboration with the user;

a recency of collaboration with the user;

a responsiveness of the user to acti ons of the respective other user associated with the plurality of documents hosted by the cloud-based content management platform;

a concurrent i nteraction of the respecti ve other user and the user with one or more documents from the plurality of documents hosted by the cloud-based content management platform;

an overlap between i) the actions and affiliations of the user in association with the cloud-based content management platform and ii) actions and affiliations of the respective other user in association with the cloud-based content management platform; or

a level of affinity of the user for the respective other user based on the affiliations of the user in association with the cloud-based content management platform .

6. The method of claim 4, wherein the predicting of the one or more collaborators for the user based on collaboration attributes of the plurality of other users comprises:

providing the set of collaboration attributes of each of the plurality of other users as input to a trained machine learning model; and

obtaining one or more outputs from the trained machine learning model, the outputs indicating, for each of the plurality of other users, a probability that the user is to collaborate with a respective other user.

7. The method of claim 6, further comprising:

providing training data to train the machine learning model on a set of training inputs and a set of target outputs, wherein:

the set of training inputs comprises collaboration attributes of a second plurality of other users; and

the set of target outputs comprising selections by the user of previous users of the second plurality of other users as collaborators.

8. The method of claim 1, wherein the predicting of the one or more collaborators for the user based on collaboration attributes of the plurality of other users further comprising:

determining, based on the collaboration attributes of the plurality of other users, a number of documents shared between the user and each of the plurality of other users;

determining, based on the collaboration attributes of the plurality of other users, a recency of each collaboration associated with the documents shared between the user and each of the plurality of other users;

ranking the plurality of other users based on the number of documents shared between the user and each of the plurality of other users, the recency of each collaboration associated with the documents shared between the user and each of the plurality of other users, and one or more rules defining a correlation between the number of documents and the recency; and

selecting one or more other users based on rankings of the plural ity of other users, the selected one or more other users representing the predicted one or more collaborators.

9. The method of claim 8, wherein the one or more rules comprise weights associated with the number of documents and the recency.

10. The method of claim 8, further comprising:

determining whether a selection of a collaborator by the user matches any of the predicted one or more collaborators; and

modifying the one or more rules responsive to the selection of the collaborator by the user being different from the one or more predicted collaborators.

11. A method comprising:

i dentifying, for a first user of a cloud-based content management platform, a plurality of pending actions directed to the first user by a second user of the cloud-based content management platform, wherein the plurality of pending acti ons by the second user are associated with a subset of a plurality of documents hosted by the cloud-based content management platform;

predicting, by a processor, one or more responses of the first user to the plurality of pending actions of the second user based on action attributes of the plurality of pending actions of the second user; and

providing, by the processor, a user interface (UI) for presentation to the first user, the UI comprising one or more UI components to be activated by the first user to provide the predicted one or more responses to at least a subset of the plurality of pending actions of the second user.

12. The method of claim 11 , wherein the plurality of pending actions directed to the first user by the second user comprises one or more of:

one or more invitations, by the second user, to the first user to share respective one or more documents from the subset of the plurality of documents,

one or more comments, by the second user, to the first user in relation to the subset of the plurality of documents, or

one or more edits, by the second user, for the first user with regards to the subset of the plurality of documents.

13. The method of claim 11, wherein the one or more action attributes include at least two or more of:

an action type of a respective action;

a recency of the respective action; or

an identity of a user of the cloud-based content management platform who initiated the respective action.

14. The method of claim 1 1 , wherein the predicting of the one or more responses of the first user to the plurality of pending actions of the second user is further based on:

response history of the first user to a plurality of previous actions of one or more other users including the second user, wherein the plurality of previous actions of the one or more other users are associated with the plurality of documents hosted by the cloud-based content management platform.

15. The method of claim 14, further comprising:

determining the response history of the first user to the plurality of previous actions, wherein the response history of the first user to the plurality of previous actions comprises one or more of:

a frequency the first user has responded to a comment by each of the one or more other users:

a ratio of a number of times the first user has edited the plurality of documents for each of the one or more other users to a number of times the first user has commented on the plurality of documents for a respective other user;

a ratio between the first user providing a comment in response to previous actions of the plural ity of previous acti ons of the one or more other users and the first user ignoring previous actions of the plurality of previous actions of the one or more other users;

a ratio between the first user editing a document in response to previous actions of the plurality of previous actions of the one or more other users and the first user ignoring previous actions of the plurality of previous actions of the one or more other users;

a ratio between the first user opening a document in response to previous actions of the plurality of previous actions of the one or more other users and the first user ignoring previous acti ons of the pl urality of previous actions of the one or more other users;

a frequency the first user has commented on the plurality of documents; a frequency the first user has responded to a comment based on a length of the comment; or

a frequency the first user has edited the plurality of documents.

16. The method of claim 14, wherein the one or more other users either only include the second user, or include the second user and one or more additional users.

17. The method of claim 1 1 , wherein the predicting of the one or more responses of the first user to the plurality of pending actions of the second user comprises:

providing one or more action attributes of each of the plurality of pending actions directed to the first user by the second user with respect to a corresponding document of the subset of documents as an input to a trained machine learning model; and

obtaining one or more outputs from the trained machine learning model, the outputs indicating, for each of the plurali ty of pending acti ons of the second user, a probability that the first user is to provide a response to a respective pending action.

18. The method of claim 17, further comprising:

providing training data to train the machine learning model on a set of training inputs and a set of target outputs, wherein:

the set of training inputs comprises one or more action attributes of a plurality of previous actions directed to the first user by other users; and

the set of target outputs comprising responses by the first user to the plurality of previous actions directed to the first user by the other users.

19. The method of claim 14, wherein the predi cting of the one or more responses of the first user to the plurality of pending actions of the second user comprises:

identi fying rules for predicting the one or more respon ses to pending acti ons of the second user, the rules comprising a ranking rule and an expected response rule, wherein the ranking rule defines how to rank a pending action based on a correlation between an action type and a recency of the pending action, and the expected response rule define expected responses to pending actions of different types, the ranking rule and the expected response rule being derived based on the response hi story of the first user to the plurality of previous actions;

determining an action type and a recency of each of the plurality of pending actions directed to the first user by the second user;

ranking each of the plurality of pending actions directed to the first user by the second user based on the ranking rule and the action type and recency of each of the plurality of pending actions directed to the first user by the second user;

selecting a subset of the plurality of pending actions based on rankings of the plurality of pending actions directed to the first user by the second user; and

identifying expected responses to the selected subset of pending actions based on the expected response rule and the type of each pending action, the identified responses representing the predicted one or more responses of the first user to the plurality of pending actions of the second user.

20. The method of claim 19, further comprising:

determining whether the first user has responded as intended to any of the selected subset of pending actions;

modifying the ranking rule responsive to determining that the first user has not responded as intended to any of the selected subset of pending actions

determining whether a selection of a response by the first user matches any of the identified responses; and

modifying the expected response rule responsive to determining that the selection of the response by the first user does not match any of the identified responses.