Arranging content on a user interface of a computing device

A non-transitory computer readable storage medium tangibly embodying a computer readable program code having computer readable instructions that, when executed, causes a computer device to carry out a method of presenting content on a user interface of a computing device, the method comprising:

• receiving historical data comprising (i) data of a plurality of customers and (ii) data of a plurality of products;
• creating a hybrid graph based on the received historical data, the hybrid graph comprising:
• one or more customer nodes and one or more product nodes;
• for each product purchased, a directed edge between a corresponding customer node and a product node;
• and
• for each product returned, a non-directed edge between a corresponding customer node and a product node;
• and
• between each two customers of the plurality of customers, applying a customer weight factor on the hybrid graph, wherein the customer weight factor indicates a degree of similarity between the two customers;
• between each two products of the plurality of products, applying a product weight factor on the hybrid graph, wherein the product weight factor indicates a degree of similarity between the two products;
• receiving a request for a seed product from a computing device of a requestor customer;
• identifying one or more products related to the seed product;
• for each identified related product, determining a return affinity score towards the requestor customer by performing a random walk analysis initiated from the seed product to find a cluster around the seed product, wherein the return affinity score is a similarity function used to quantify the similarity between the two products and the two customers by evaluating a customer's return attribute probability for the products based on the hybrid graph;
• ranking each related product based on their return affinity score;
• and
• sending a representation of the related products to be displayed on the computing device of the requestor customer based on the ranking, where the related products are emphasized and/or suppressed based on the return affinity score.

A computer implemented method, comprising:

• receiving historical data comprising (i) data of a plurality of customers and (ii) data of a plurality of products;
• creating a hybrid graph based on the received historical data, the hybrid graph comprising:
• one or more customer nodes and one or more product nodes;
• for each product purchased, a directed edge between a corresponding customer node and a product node;
• and
• for each product returned, a non-directed edge between a corresponding customer node and a product node;
• and
• between each two customers of the plurality of customers, applying a customer weight factor on the hybrid graph, wherein the customer weight factor indicates a degree of similarity between the two customers;
• between each two products of the plurality of products, applying a product weight factor on the hybrid graph, wherein the product weight factor indicates a degree of similarity between the two products;
• receiving a request for a seed product from a computing device of a requestor customer;
• identifying one or more products related to the seed product;
• for each identified related product, determining a return affinity score towards the requestor customer by performing a random walk analysis initiated from the seed product to find a cluster around the seed product, wherein the return affinity score is a similarity function used to quantify the similarity between the two products and the two customers by evaluating a customer's return attribute probability for the products based on the hybrid graph;
• ranking each related product based on their return affinity score;
• and
• sending a representation of the related products to be displayed on the computing device of the requestor customer based on the ranking, where the related products are emphasized and/or suppressed based on the return affinity score.