Recommendation system and multi-algorithm fusion recommendation processing flow

Abstract

The invention discloses a recommendation system and a multi-algorithm fusion recommendation processing flow. A plurality of feature conversion algorithms and a plurality of model training algorithms are integrated together by a feature conversion and model training module, separate loading is not needed during calling, the practicability is good, a new model can be obtained only by modifying a configuration file and performing training storage during model optimization, in the actual application process, a user only needs to pay attention to input data and an output result, intermediate feature processing and model training are packaged, and the cost of independent maintenance is saved. A model file storage module stores a model file in a distributed file system and records basic information of the model file into a relational database, a model reading module can conveniently read the model basic information and load a corresponding model file from the distributed file system accordingto the model basic information, and reading is convenient and fast, so that an API service module obtains the loaded model, receives a network request and returns a recommendation result to the frontend.

Description

technical field [0001] The invention relates to a recommendation system and a multi-algorithm fusion recommendation processing flow. Background technique [0002] In recent years, with the development and popularization of mobile Internet technology, more and more user behavior data have been generated, and users are surrounded by a lot of information. At this time, recommendation systems come into play. In essence, the recommendation system is to find the information of interest for users from a large amount of information with the explosive growth of user scale and the increasing types of items provided by suppliers when the needs of users are not clear. Provide accurate personalized recommendations. [0003] Users have dynamic and static attributes, static attributes such as age, gender, region, etc., dynamic attributes such as historical behavior, context information (login time, login device, etc.), corresponding items also have dynamic and static attributes, static at...

AI Technical Summary

Problems solved by technology

Feature processing, model training, and recommendation services are completely separated, and each item requires independent maintenance, which is costly and inconvenient to adjust feature processing and hyperparameters

Images