Method and device of determining stability of search function

Abstract

The embodiment of the invention discloses a method and a device of determining stability of a search function. The method of the embodiment comprises steps of obtaining a first document list corresponding to a benchmark version of an application and a second document list corresponding to a target version of the application; calculating a sub-differential value between a first sub-list and a second sub-list both corresponding to a same search word in a search word set, wherein the first document list comprises the first sub-list corresponding to each search word in the search set, and the second document list comprises the second sub-list corresponding to each search word in the search set; determining difference information of the first document list and the second document list both corresponding to the search word set based on the sub-differential value which corresponds to each search word; and determining the stability of the search function of the target version of the application based on the difference information, wherein the difference information is in negative correlation with the stability of the search function of the new version of the application. The embodiment ofthe invention also provides a method of determining the stability of the search function; the assessment standard of the stability is quantified by indexes, and the accuracy is improved.

Description

technical field [0001] The invention relates to the field of computers, in particular to a method and device for determining the stability of a search function. Background technique [0002] In the service iteration process of the application including the search service, updating the version of the service is a relatively common process. The stability evaluation of the new version of the search service is also an important link. If the stability of the new version of the search service is uncertain, problems may occur after the new version is launched. For example, taking a video application as an example, the current video application needs to be upgraded, and the old version is upgraded to a new version. Suppose the old version of the video application receives a search term, the search term is "morning at dawn", and the old version of the video application The video application feeds back search results, and the search results are related videos of the TV series "Mornin...

AI Technical Summary

Problems solved by technology

[0004] Man-made subjective judgments on the stability of the new version of the application search service waste manpower and time costs, and are highly subjective. Different testers may have different judgments on the same result, and the probability of judgment errors is high.

Images