An application ranking method, device and computing device

Abstract

The invention discloses an application ranking method and device and a computing device. The computing device comprises a data storage information which stores the application information of a plurality of applications, wherein the application information of each application comprises the application identification of the application, an application active amount and an accessed SDK (Software Development Kit) sign. The method comprises the following steps that: according to the application active amount, grouping each application to obtain a first amount of application groups; for each application group, according to the accessed SDK (Software Development Kit) sign, dividing the applications in the application group into the corresponding accessed SDK application group or the non-accessed SDK application group; for the accessed SDK application group or the non-accessed SDK application group in each application group, carrying out differentiated analysis to obtain inter-group difference metric values; if the inter-group difference metric value smaller than a first threshold value is in the presence, carrying out rectification processing on the application active amount of the application, which corresponds to the inter-group difference metric value, in the accessed SDK application group; and sorting the application active amount of each application according to a descending order to generate a corresponding rank.

Description

technical field [0001] The invention relates to the technical field of the Internet, in particular to an application ranking method, device and computing equipment. Background technique [0002] With the continuous development of Internet technology and hardware technology, more and more people begin to use mobile terminal devices such as smart phones and tablet computers, and various mobile applications that meet people's daily needs are rapidly increasing in number. Correspondingly, the application market for mobile devices has also been rapidly and fully developed, with a wide variety of applications, and there are many types of applications in the same field and the competition is fierce. Generally speaking, if the abnormal situation in the data collection process is not taken into account, the actual operating conditions of an application, such as the active volume of the application (the number of installations or the number of times the application is opened and run w...

AI Technical Summary

Problems solved by technology

However, it is still difficult to obtain the real operating conditions of applications from the outside. The collected data is not accurate and deviates from the actual situation. It is difficult to give a ranking with high accuracy and credibility for all applications in the application market.

[0003] The existing application ranking methods are mainly divided into two types. The first one is that the relevant demand personnel empirically estimate the application activity, and rank according to the estimated results. This method is highly subjective, the estimation efficiency is low, and it cannot be proved and discussed. Accuracy, the second is to observe the application behavior by connecting the application to the SDK (Software Development Kit, software development kit), realize automatic data collection, and in principle, be able to measure the operation indicators of a large number of applications, such as Analysis and calculation of application activity

[0004] However, the second application ranking method cannot accurately understand the application activity of applications without SDK access. In order to solve this problem, based on an application that accesses the SDK, other applications can be used through data processing technology. Active volume estimation for app ranking

Although this solution solves the problem of ranking apps that are not connected to the SDK, it introduces system errors, resulting in high application activity for apps connected to the SDK, and the noise impact that may be caused by the data collection mechanism itself. , so that the final ranking results have large errors, and even reversed rankings

Images