# US 10,455,033 B2 · Systems and methods for evaluating user activity > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 10,455,033 B2 - Document ID: US10455033B2 - Status: Granted - Kind: B2 - Application number: 15/156,173 - Filed: 2016-05-16 - Published: 2017-11-16 - Granted: 2019-10-22 - Earliest claimed: 2016-05-16 - Assignees: Facebook, Inc. - Applicants: Facebook, Inc. - Inventors: Akos Lada; Alexander Peysakhovich - Web page: https://patentgrep.com/patent/US10455033B2 ## CPC - G06N 20/00: Machine learning - G06Q 10/04: Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" (market predictions or forecasting for commercial activities G06Q30/0202) - G06Q 10/0639: Performance analysis of employees; Performance analysis of enterprise or organisation operations - G06Q 10/101: Collaborative creation, e.g. joint development of products or services - G06Q 10/107: Computer-aided management of electronic mailing [e-mailing] - G06Q 10/40: Business processes related to social networking or social networking services - H04L 41/16: using machine learning or artificial intelligence - H04L 67/535: {Tracking the activity of the user (network monitoring arrangements H04L43/00; recording of computer activity G06F11/34)} - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR - H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M) ## Abstract Systems, methods, and non-transitory computer-readable media can determine an event that may affect at least one activity being performed by a first group of users through the computing system. A set of first measurements of the at least one activity being performed by the first group of users over a period of time are determined. A set of second measurements of the at least one activity for the first group of users over the period of time are generated, wherein the set of second measurements are predicted based at least in part on a machine learning model. Data describing an impact of the event on the first group of users is generated based at least in part on the set of first measurements and the set of second measurements. ## Claims ### Claim 1 A computer-implemented method comprising: training, by a computing system, a machine learning model to output forecasted measurements of at least one activity for at least a first group of users, wherein the at least one activity includes a number of likes performed by the first group of users, and wherein the machine learning model is trained using a set of training examples that each include respective counts of the at least one activity having been performed by each group of users in a set of groups over some period of time that precedes an event, the set of groups including the first group of users and a plurality of different groups of users; determining, by the computing system, the event that may affect at least one activity being performed by a first group of users through the computing system; determining, by the computing system, a set of first measurements of the at least one activity being performed by the first group of users over a first period of time; generating, by the computing system, a set of second measurements of the at least one activity for the first group of users over a second period of time, wherein the set of second measurements are predicted based at least in part on the trained machine learning model, and wherein the set of second measurements for the first group of users are predicted based at least in part on the at least one activity having been performed by one or more groups of users in a set of groups, the set of groups not including the first group of users; and generating, by the computing system, data describing an impact of the event on the first group of users based at least in part on the set of first measurements and the set of second measurements. ### Claim 2 (depends on 1) The computer-implemented method of claim 1, wherein training the machine learning model further comprises: generating, by the computing system, a set of training examples that each include respective counts of the at least one activity having been performed by each group of users in a set of groups over a period of time that precedes the event, the set of groups including the first group of users and a plurality of different groups of users. ### Claim 3 (depends on 1) The computer-implemented method of claim 1, wherein generating the set of second measurements of the at least one activity for the first group of users further comprises: providing, by the computing system, respective counts of the at least one activity having been performed by each group of users in a set of groups during a unit of time to the machine learning model, the set of groups not including the first group of users, wherein the machine learning model outputs a predicted measurement of the at least one activity for the first group of users corresponding to the unit of time. ### Claim 4 (depends on 1) The computer-implemented method of claim 1, wherein the predicted measurement is a prediction interval within which a count of the at least one activity for the first group of users was expected had the event not occurred. ### Claim 5 (depends on 1) The computer-implemented method of claim 1, wherein generating data describing the impact of the event on the first group of users further comprises: generating, by the computing system, a visual graph that plots the set of first measurements and the set of second measurements over the period of time. ### Claim 6 (depends on 1) The computer-implemented method of claim 1, wherein the first group of users includes users that are associated with a geographic region, users that share at least one specified demographic attribute, or users that share at least on specified behavioral attribute. ### Claim 7 (depends on 1) The computer-implemented method of claim 1, wherein determining the event that may affect at least one activity being performed by the first group of users through the computing system further comprises: determining, by the computing system, a date on which a feature is available to the group of users. ### Claim 8 (depends on 1) The computer-implemented method of claim 1, wherein determining the event that may affect at least one activity being performed by the first group of users through the computing system further comprises: determining, by the computing system, a date on which a real-world event was experienced by at least some of users in the first group of users. ### Claim 9 (depends on 1) The computer-implemented method of claim 1, wherein the at least one activity corresponds to: a number of likes performed, a number of comments posted, a number of content items shared, a number of content items accessed, a number of content items of a specified type accessed, a number of content items in a specified category accessed, an amount of time spent while accessing content items, or any combination of thereof. ### Claim 10 A system comprising: at least one processor; and a memory storing instructions that, when executed by the at least one processor, cause the system to perform: training a machine learning model to output forecasted measurements of at least one activity for at least a first group of users, wherein the at least one activity includes a number of likes performed by the first group of users, and wherein the machine learning model is trained using a set of training examples that each include respective counts of the at least one activity having been performed by each group of users in a set of groups over some period of time that precedes an event, the set of groups including the first group of users and a plurality of different groups of users; determining the event that may affect at least one activity being performed by a first group of users through the computing system; determining a set of first measurements of the at least one activity being performed by the first group of users over a first period of time; generating a set of second measurements of the at least one activity for the first group of users over a second period of time, wherein the set of second measurements are predicted based at least in part on the trained machine learning model, and wherein the set of second measurements for the first group of users are predicted based at least in part on the at least one activity having been performed by one or more groups of users in a set of groups, the set of groups not including the first group of users; and generating data describing an impact of the event on the first group of users based at least in part on the set of first measurements and the set of second measurements. ### Claim 11 (depends on 10) The system of claim 10, wherein training the machine learning model further causes the system to perform: generating a set of training examples that each include respective counts of the at least one activity having been performed by each group of users in a set of groups over a period of time that precedes the event, the set of groups including the first group of users and a plurality of different groups of users. ### Claim 12 (depends on 10) The system of claim 10, wherein generating the set of second measurements of the at least one activity for the first group of users further causes the system to perform: providing respective counts of the at least one activity having been performed by each group of users in a set of groups during a unit of time to the machine learning model, the set of groups not including the first group of users, wherein the machine learning model outputs a predicted measurement of the at least one activity for the first group of users corresponding to the unit of time. ### Claim 13 (depends on 10) The system of claim 10, wherein the predicted measurement is a prediction interval within which a count of the at least one activity for the first group of users was expected had the event not occurred. ### Claim 14 A non-transitory computer-readable storage medium including instructions that, when executed by at least one processor of a computing system, cause the computing system to perform a method comprising: training a machine learning model to output forecasted measurements of at least one activity for at least a first group of users, wherein the at least one activity includes a number of likes performed by the first group of users, and wherein the machine learning model is trained using a set of training examples that each include respective counts of the at least one activity having been performed by each group of users in a set of groups over some period of time that precedes an event, the set of groups including the first group of users and a plurality of different groups of users; determining the event that may affect at least one activity being performed by a first group of users through the computing system; determining a set of first measurements of the at least one activity being performed by the first group of users over a first period of time; generating a set of second measurements of the at least one activity for the first group of users over a second period of time, wherein the set of second measurements are predicted based at least in part on the trained machine learning model, and wherein the set of second measurements for the first group of users are predicted based at least in part on the at least one activity having been performed by one or more groups of users in a set of groups, the set of groups not including the first group of users; and generating data describing an impact of the event on the first group of users based at least in part on the set of first measurements and the set of second measurements. ### Claim 15 (depends on 14) The non-transitory computer-readable storage medium of claim 14, wherein training the machine learning model further causes the computing system to perform: generating a set of training examples that each include respective counts of the at least one activity having been performed by each group of users in a set of groups over a period of time that precedes the event, the set of groups including the first group of users and a plurality of different groups of users. ### Claim 16 (depends on 14) The non-transitory computer-readable storage medium of claim 14, wherein generating the set of second measurements of the at least one activity for the first group of users further causes the computing system to perform: providing respective counts of the at least one activity having been performed by each group of users in a set of groups during a unit of time to the machine learning model, the set of groups not including the first group of users, wherein the machine learning model outputs a predicted measurement of the at least one activity for the first group of users corresponding to the unit of time. ### Claim 17 (depends on 14) The non-transitory computer-readable storage medium of claim 14, wherein the predicted measurement is a prediction interval within which a count of the at least one activity for the first group of users was expected had the event not occurred.