Computerized systems and methods for using artificial intelligence to generate product recommendations

A method for AI-based product recommendation generation, comprising:

• generating a model by training the model with training data, wherein the training data comprises at least one of historical queries, historical filter applications, or historical product specifications of a plurality of customers;
• receiving, via a server, a query;
• extracting at least one keyword from the query;
• based on the extracted at least one keyword, determining a user identifier and a product identifier;
• based on the determined user identifier and the determined product identifier, searching a database associated with the user identifier and the product identifier;
• based on the user identifier and the product identifier, predicting, using the generated model, a plurality of recommended products, by:
• determining a time span associated with the user identifier;
• retrieving at least one query submitted by a user associated with the user identifier during the time span associated with the user identifier;
• retrieving at least one filter application submitted by the user during the time span associated with the user identifier;
• retrieving at least one product specification associated with at least one product associated with the retrieved at least one query or the retrieved at least one filter application;
• extracting attributes from the retrieved at least one query, the retrieved at least one filter application, and the retrieved at least one product specification;
• determining a first product category and a second product category based on the extracted attributes;
• determining a first frequency in engagement, via a user device, in the time span associated with the user identifier between the user and the first product category;
• determining a second frequency in engagement, via the user device, in the time span associated with the user identifier between the user and the second product category;
• determining that the first frequency is greater than the second frequency in the time span associated with the user identifier;
• generating a list of products based on the determined first product category, wherein products belonging to the second category are excluded from the generated list of products;
• generating a list of recommended products based on the generated list of products;
• and
• modifying a user interface element associated with a webpage, via the server, to include the generated list of recommended products.

A computer-implemented system for AI-based product recommendation generation, the system comprising:

• a memory storing instructions;
• and
• at least one processor configured to execute the instructions to:
• generate a model by training the model with training data, wherein the training data comprises at least one of historical queries, historical filter applications, or historical product specifications of a plurality of customers;
• receive, via a server, a query;
• extract at least one keyword from the query;
• based on the extracted at least one keyword, determine a user identifier and a product identifier;
• based on the determined user identifier and the determined product identifier, search a database associated with the user identifier and the product identifier;
• based on the user identifier and the product identifier, predict, using the generated model, a plurality of recommended products, by:
• determining a time span associated with the user identifier;
• retrieving at least one query submitted by a user associated with the user identifier during the time span associated with the user identifier;
• retrieving at least one filter application submitted by the user during the time span associated with the user identifier;
• retrieving at least one product specification associated with at least one product associated with the retrieved at least one query or the retrieved at least one filter application;
• extracting attributes from the retrieved at least one query, the retrieved at least one filter application, and the retrieved at least one product specification;
• determining a first product category and a second product category based on the extracted attributes;
• determining a first frequency in engagement, via a user device, in the time span associated with the user identifier between the user and the first product category;
• determining a second frequency in engagement, via the user device, in the time span associated with the user identifier between the user and the second product category;
• determining that the first frequency is greater than the second frequency in the time span associated with the user identifier;
• generating a list of products based on the determined first product category, wherein products belonging to the second category are excluded from the generated list of products;
• generating a list of recommended products based on the generated list of products;
• calculating a similarity score for each product on the generated list;
• removing products with similarity scores below a threshold from the generated list;
• and
• modifying a user interface element associated with a webpage, via the server, to include the generated list of recommended products.