Automated social message stream population

Patent metadata

Field	Value
Publication number	US-10565180-B2
Application number	US-201816004897-A
Country	US
Kind code	B2
Filing date	Jun 11, 2018
Priority date	Oct 30, 2014
Publication date	Feb 18, 2020
Grant date	Feb 18, 2020

Abstract

A messaging system automatically populates a stream of messages using only a seed selected by the requesting account holder. In one embodiment, the seed includes the streams of one or more of the “top” accounts of the messaging system. Here, “top” is according to any one of a number of different metrics stored in the messaging system. With knowledge of the seed, the messaging system automatically populates a stream for the requesting account holder, without requiring any other input. As a result, an account holder is provided with a fully functioning stream with very little effort or knowledge required on their part.

First claim

What is claimed is:

1. A method performed by a messaging system for automatically populating a stream of messages for a requesting user of the messaging system, using a seed selected by the requesting user, the method comprising:
providing, by a computer server, to a client device, a list of top accounts of the messaging system for selection by a user of the client device, the top accounts of the messaging system being a subset of accounts that are top when measured by numbers of followers or when measured by numbers of engagements on messages of the accounts;
receiving, by the computer server, a request for a stream of messages from the user of the client device, the request specifying one or more of the top accounts on the list to use as a seed;
determining a feature vector for each top account, each feature vector being a normalized feature vector for a particular one of the top accounts;
determining a seed vector, the seed vector being a normalized vector that is a single feature vector or a normalized sum of multiple feature vectors for the top account or accounts selected by the requesting user as the seed;
determining a respective relevance to the seed of each of a plurality of top accounts by computing, for each of the plurality of top accounts, a function that combines the seed vector and the feature vector of the top account in accordance with
R S ( S,FV i )=sign( S,FV i )×| S·FV i | α

where S is the seed vector, FV i is the feature vector for the i-th top account, α is a breadth factor, and R S is a relevance of the seed to the feature vector; and
ranking messages authored by top accounts for inclusion in the stream of messages for the user using each relevance of the seed to a respective feature vector of each of the top accounts.

2. The method of claim 1 , wherein the messaging system populates the stream of messages for the requesting user without requiring any input from the requesting user other than the selection of the seed.

3. The method of claim 1 , wherein determining the feature vectors of the top accounts comprises:
constructing a follow matrix F having N general accounts on a first axis and T top accounts on a second axis, wherein the value of each cell in the matrix is a binary value that depends on whether or not the account of the first axis follows the account of the second axis or vice versa;
factoring F into an approximate matrix product UΣV to identify A top dimensions of F, where U is an N×A matrix, Σ is a diagonal A×A matrix, and V is an A×T matrix, wherein A is a predetermined number of dimensions; and
determining the feature vectors as the normalized row vectors of a matrix M=(Σ·V) T .

4. The method of claim 3 , wherein determining the feature vectors of the top accounts is completed before the receiving request from user is received, the method comprising:
storing the matrix M in a database; and
determining a feature vector for each top account comprises reading the matrix M from the database.

5. The method of claim 3 , wherein factoring F comprises factoring F by singular value decomposition.

6. The method of claim 3 , wherein the factoring of F is done by a distributed factorization method.

7. The method of claim 3 , comprising trimming M before using the feature vectors in M by removing a column that indicates how many followers the top accounts have.

8. The method of claim 3 , comprising not including in M a column that indicates how many followers the top accounts have.

9. The method of claim 3 , wherein A is in a range of 100 to 500.

10. A messaging system for automatically populating a stream of messages for a requesting user of the messaging system, using a seed selected by the requesting user, the messaging system comprising:
multiple computing devices, each computing device including computer hardware, the multiple computing devices being configured to perform messaging system operations comprising:
providing to a client device, a list of top accounts of the messaging system for selection by a user of the client device, the top accounts of the messaging system being a subset of accounts that are top when measured by numbers of followers or when measured by numbers of engagements on messages of the accounts;
receiving a request for a stream of messages from the user of the client device, the request specifying one or more of the top accounts on the list to use as a seed;
determining a feature vector for each top account, each feature vector being a normalized feature vector for a particular one of the top accounts;
determining a seed vector, the seed vector being a normalized vector that is a single feature vector or a normalized sum of multiple feature vectors for the top account or accounts selected by the requesting user as the seed;
determining a respective relevance to the seed of each of a plurality of top accounts by computing, for each of the plurality of top accounts, a function that combines the seed vector and the feature vector of the top account in accordance with
R S ( S,FV i )=sign( S,FV i )×| S·FV i | α

where S is the seed vector, FV i is the feature vector for the i-th top account, α is a breadth factor, and R S is a relevance of the seed to the feature vector; and
ranking messages authored by top accounts for inclusion in the stream of messages for the user using each relevance of the seed to a respective feature vector of each of the top accounts.

11. The messaging system of claim 10 , wherein the operations comprise:
populating the stream of messages for the requesting user without requiring any input from the requesting user other than the selection of the seed.

12. The messaging system of claim 10 , wherein determining the feature vectors of the top accounts comprises:
constructing a follow matrix F having N general accounts on a first axis and T top accounts on a second axis, wherein the value of each cell in the matrix is a binary value that depends on whether or not the account of the first axis follows the account of the second axis or vice versa;
factoring F into an approximate matrix product UΣV to identify A top dimensions of F, where U is an N×A matrix, Σ is a diagonal A×A matrix, and V is an A×T matrix, wherein A is a predetermined number of dimensions; and
determining the feature vectors as the normalized row vectors of a matrix M=(Σ·V) T .

13. The messaging system of claim 12 , wherein determining the feature vectors of the top accounts is completed before the receiving request from user is received, the method comprising:
storing the matrix M in a database; and
determining a feature vector for each top account comprises reading the matrix M from the database.

14. The messaging system of claim 12 , wherein factoring F comprises factoring F by singular value decomposition.

15. The messaging system of claim 12 , wherein the factoring of F is done by a distributed factorization method.

16. The messaging system of claim 12 , wherein trimming M before using the feature vectors in M by removing a column that indicates how many followers the top accounts have.

17. The messaging system of claim 12 , wherein not including in M a column that indicates how many followers the top accounts have.

18. The messaging system of claim 12 , wherein A is in a range of 100 to 500.

19. A non-transitory computer-readable storage medium storing computer software for automatically populating a stream of messages for a requesting user of the messaging system, using a seed selected by the requesting user, th

Assignees

• Twitter Inc

Inventors

• Fleischman Michael Ben
• Miller Matthew
• Whitcomb Jr Richard Douglas
• Watabe Mark
• Sciola Anthony

Classifications

• G06Q10/10

  Office automation; Time management · CPC title

• H04L51/02

  using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages · CPC title

• G06F3/0482

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

• G06Q10/40

  Business processes related to social networking or social networking services · CPC title

• H04L51/32

  Electricity · mapped topic