Product click rate prediction method and computing device

The invention provides a click rate prediction method of a product and computing equipment. The method comprises the following steps: acquiring scene features and non-scene features of a target scene, wherein the non-scene features comprise features of a product; performing weight adjustment on non-scene features of the target scene according to the representation of the target scene to obtain non-scene features of an implicit scene corresponding to the target scene, the representation of the target scene being determined according to the scene features; fusing the representation of the non-scene feature of the implicit scene in the representation of the target scene to obtain the representation of the implicit scene; and determining a click rate prediction value of the product in the target scene according to the representation of the implicit scene. According to the method, the implicit scene division is performed on the explicit target scene, and the click rate of the product in the target scene is predict