| CPC G06F 16/1744 (2019.01) [G06N 5/022 (2013.01)] | 15 Claims |

|
1. A data compression method performed by a computing device, comprising:
receiving, by a first processor of the computing device, to-be-compressed data;
obtaining, by the first processor of the computing device, M data type identification models;
determining, by the first processor of the computing device, M probability sets corresponding to the to-be-compressed data based on the M data type identification models;
identifying, by the first processor of the computing device, a data type of the to-be-compressed data by obtaining P intermediate values based on the M probability sets and weight ratios of the M data type identification models, the data type of the to-be-compressed data corresponding to an intermediate value in the P intermediate values and that is greater than an accuracy threshold, the data type including a text type, a picture type, or a video type;
selecting, by the first processor of the computing device, a data compression model based on the data type;
sending, by the first processor of the computing device, the to-be-compressed data and information related to the selected data compression model to a second processor of the computing device;
compressing, by the second processor of the computing device, the to-be-compressed data to obtain compressed data based on the selected data compression model; and
sending, by the second processor of the computing device, the compressed data to the first processor of the computing device.
|