Method and a system for selecting a targeted message to be included within a web resource

A system for training a machine-learning algorithm (MLA) for determining a targeted message server storing one or more targeted messages to be included within a web resource hosted by a web server;

• the web server being communicatively coupled to a first targeted message server and a second targeted message server,
• the first targeted message server storing a first plurality of targeted messages and the second targeted message server storing a second plurality of targeted messages;
• each targeted message of the first plurality of targeted messages and second plurality of targeted messages being associated with a context parameter and a bid price;
• the system comprising at least one processor and at least one non-transitory computer-readable medium comprising executable instructions, which, when executed by the at least one processor, cause the system to:
• generate a training dataset by:
• retrieving, from a database, a processing log, the processing log comprising:
• one or more previously received requests for targeted messages, each of the one or more previously received requests including a respective target context parameter and a respective target floor price,
• the one or more previously received requests having been previously completed by at least one of the first targeted message server and the second targeted message server;
• and
• the respective target context parameter comprising a topic of a given targeted message responsive to a respective previously received request;
• inputting the training dataset to the MLA, thereby determining a set of features associated with the training dataset, the set of features including at least:
• (i) the respective target context parameter and the respective target floor price of a given previously received request, (ii) a respective previously identified targeted message responsive to the given previously received request, and (iii) a success feature indicating which of the first targeted message server or the second targeted message server stored the respective previously identified targeted message;
• generating an inferred function based on the set of features, the inferred function being configured, during an in-use phase, to enable the MLA to determine, from the first and second targeted message servers, the targeted message server storing the one or more targeted message to be included within the web resources, by:
• receiving, from the web server, a first request for the one or more targeted messages, the first request including a target context parameter and a target floor price;
• applying, to the target context parameter and target floor price, the inferred function to determine:
• a first confidence parameter indicative of a first likelihood value of the first targeted message server storing, being a targeted message meeting at least a first condition and a second condition;
• the first condition corresponding to the context parameter of the targeted message matching the target context parameter, and the second condition corresponding to the bidding price being above the target floor price;
• a second confidence parameter indicative of a second likelihood value of the second targeted message server storing the targeted message meeting at least the first condition and the second condition;
• transmitting, a second request only to a selected one of the first targeted message server and the second targeted message server based on the first confidence parameter and the second confidence parameter, the second request including the target context parameter and the target floor price;
• receiving, the targeted message from the selected one of the first targeted message server and the second targeted message server;
• and
• transmitting, the targeted message to the web server for inclusion within the web resource.