Apparatus, method and article to effect electronic message reply rate matching in a network environment

A system to enhance messaging between users, the system comprising:

• a communication network that includes servers, processors, display devices, input devices, and system memory devices to send messages between users and autonomously collect actual historical messaging data;
• at least one processor;
• and
• at least one non-transitory processor-readable medium communicatively coupled to the least one processor, wherein the at least one processor:
• forms an initial data set, the initial data set comprising user profile information for a plurality of users, user behavioral information for at least some of the plurality of users, and message data indicative of messaging activity between at least some of the plurality of users;
• organizes the initial data set into a training data set and a test data set, wherein in order to organize the initial data set into a training data set and a test data set, the at least one processor randomly separates the selected users for representation in one or the other of the training data set or the test data set;
• generates a response predictive model using a machine learning system from the training data set which includes the user profile information, the user behavioral information and message data;
• evaluates the response predictive model generated from the training data set against the corresponding test data set, wherein at least some of the test data set is based on actual historical messaging data;
• and
• determines an accuracy of the response prediction model using the actual historical messaging data.

A system to enhance messaging between users, the system comprising:

• a communication network that includes servers, processors, display devices, input devices, and system memory devices to send messages between users and autonomously collect actual historical messaging data;
• at least one processor;
• and
• at least one non-transitory processor-readable medium communicatively coupled to the least one processor, wherein the at least one processor:
• forms an initial data set, the initial data set comprising user profile information for a plurality of users, user behavioral information for at least some of the plurality of users, and message data indicative of messaging activity between at least some of the plurality of users;
• organizes the initial data set into a training data set and a test data set, wherein in order to organize the initial data set into a training data set and a test data set, the at least one processor randomly selects for inclusion in either the training data set or the test data set the user profile information, the user behavioral information, and the messaging data from all of the selected users on a message by message basis;
• generates a response predictive model using a machine learning system from the training data set which includes the user profile information, the user behavioral information and message data;
• evaluates the response predictive model generated from the training data set against the corresponding test data set, wherein at least some of the test data set is based on actual historical messaging data;
• and
• determines the accuracy of the response prediction model using the actual historical messaging data.

A system to enhance messaging between users, the system comprising:

• a communication network that includes servers, processors, display devices, input devices, and system memory devices to send messages between users and autonomously collect actual historical messaging data;
• at least one processor;
• and
• at least one non-transitory processor-readable medium communicatively coupled to the least one processor, wherein the at least one processor:
• forms an initial data set, the initial data set comprising user profile information for a plurality of users, user behavioral information for at least some of the plurality of users, and message data indicative of messaging activity between at least some of the plurality of users;
• organizes the initial data set into a training data set and a test data set, wherein to organize the initial data set into a training data set and a test data set, the at least one processor further, for each of a first set of randomly selected users, includes all messages for each of the users in the first set of users in a training data set, and for a second set of users, different than the first set of users, includes all messages for each of the users in the second set in a test data set;
• generates a response predictive model using a machine learning system from the training data set which includes the user profile information, the user behavioral information and message data;
• evaluates the response predictive model generated from the training data set against the corresponding test data set, wherein at least some of the test data set is based on actual historical messaging data;
• and
• determines the accuracy of the response prediction model using the actual historical messaging data.