Machine learning system for configuring social media campaigns

A system comprising:

• at least one device including a hardware processor;
• the system being configured to perform operations comprising:
• performing, by a social relationship management (SRM) service, supervised machine learning to generate a learned model for configuring social media campaigns, at least by:
• generating a plurality of feature vectors, based on training data comprising campaign configuration data and one or more campaign success metrics;
• and
• performing pattern recognition on the plurality of feature vectors to determine one or more preferred campaign configurations;
• publishing a plurality of messages to at least one social media platform, wherein each message in the plurality of messages is associated, respectively, with a social media campaign in a plurality of social media campaigns;
• receiving user interaction data associated with a plurality of interactions, by a plurality of users of the at least one social media platform, with the plurality of messages;
• vectorizing quantified measurements being stored in a cloud storage environment, wherein the cloud storage environment is access-controlled to secure the vectorized quantifications for access to perform unsupervised machine learning;
• performing unsupervised machine learning to update the learned model based at least in part on quantified measurements of the user interaction data,
• wherein the plurality of social media campaigns are clustered using an unsupervised clustering algorithm to detect a particular cluster of campaigns having similar characteristics,
• wherein quantified measurements for the particular cluster of campaigns are expected to be higher than for other clusters for particular request parameters;
• receiving, by the SRM service, a request to configure a particular social media campaign, wherein the request specifies certain parameters;
• applying data associated with the request to the learned model to determine a preferred campaign configuration, from the one or more preferred campaign configurations, for the particular social media campaign, at least by:
• selecting campaign characteristics based on the particular cluster of campaigns having quantified measurements that are expected to be higher than for other clusters for the particular request parameters;
• and
• configuring the particular social media campaign based on the preferred campaign configuration.

A method comprising:

• performing, by a social relationship management (SRM) service, supervised machine learning to generate a learned model for configuring social media campaigns, at least by:
• generating a plurality of feature vectors, based on training data comprising campaign configuration data and one or more campaign success metrics;
• and
• performing pattern recognition on the plurality of feature vectors to determine one or more preferred campaign configurations;
• publishing a plurality of messages to at least one social media platform, wherein each message in the plurality of messages is associated, respectively, with a social media campaign in a plurality of social media campaigns;
• receiving user interaction data associated with a plurality of interactions, by a plurality of users of the at least one social media platform, with the plurality of messages;
• vectorizing quantified measurements being stored in a cloud storage environment, wherein the cloud storage environment is access-controlled to secure the vectorized quantifications for access to perform unsupervised machine learning;
• performing unsupervised machine learning to update the learned model based at least in part on quantified measurements of the user interaction data,
• wherein the plurality of social media campaigns are clustered using an unsupervised clustering algorithm to detect a particular cluster of campaigns having similar characteristics,
• wherein quantified measurements for the particular cluster of campaigns are expected to be higher than for other clusters for particular request parameters;
• receiving, by the SRM service, a request to configure a particular social media campaign, wherein the request specifies certain parameters;
• applying data associated with the request to the learned model to determine a preferred campaign configuration, from the one or more preferred campaign configurations, for the particular social media campaign, at least by:
• selecting campaign characteristics based on the particular cluster of campaigns having quantified measurements that are expected to be higher than for other clusters for the particular request parameters;
• and
• configuring the particular social media campaign based on the preferred campaign configuration,
• wherein the method is performed by at least one device comprising a hardware processor.