Product testing method and device based on MOA model, and equipment

Abstract

The invention discloses a product test method and device based on an MOA model, and equipment, which can cover the use scene of a user as much as possible, accurately determine the function effect of an algorithm product and further improve the accuracy of a test result. The method comprises the following steps: firstly, determining a to-be-tested target algorithm product, and then dividing the motivation of at least one target user using the target algorithm product to obtain a first division result; dividing the opportunity of using the target algorithm product by at least one target user to obtain a second division result; dividing the ability of the at least one target user to use the target algorithm product to obtain a third division result, then generating a user group portrait of the at least one target user according to the first division result, the second division result and the third division result, and further generating a test case according to the user group portrait of the at least one target user; and testing the target algorithm product by using the trial case to obtain a test result.

Description

technical field [0001] The present application relates to the field of computer technology, in particular to a product testing method, device and equipment based on the MOA model. Background technique [0002] With the rapid development of social economy, more and more functional products composed of various algorithms have brought great convenience to people's lives. The testing of new product algorithms has long been a key step before new products go online. . [0003] At present, in the testing phase, for the algorithm models newly adopted by products (such as various mobile phone function software), such as collaborative filtering, machine learning algorithms, etc., many projects only return to the functions and processes, and do not evaluate the specific algorithm models; Generally, the algorithm model is evaluated through the recommendation effect published online or in gray scale. Since the testers are involved in the testing and analysis of related products in the ...

AI Technical Summary

Problems solved by technology

Since the testers are involved in the testing and analysis of related products in the algorithm market, there are many kinds of algorithms and the complexity is relatively large, and the testers cannot fully cover the test scenarios of the algorithms on the basis of traditional test analysis. The user experience of the product is covered on the basis of the test algorithm and the function of the algorithm market itself. If you encounter a project with a short and frequent iteration cycle, it is difficult to obtain such test analysis requirements only by relying on traditional test analysis and algorithm test frameworks. Satisfy

Images