Data query method and device, and readable medium

Abstract

The invention discloses a data query method and device, and a readable medium, and belongs to the technical field of data processing. In the method and the device provided by the invention, after a query statement in a query request is acquired, the confidence probability of at least one candidate segment can be determined according to a correlation model and a posterior recall quality model whichare obtained by pre-training, and an optimal query result of the query statement is determined according to a query result obtained by querying the candidate segment with the confidence probability meeting the screening condition and a query result obtained by querying by using the query statement. In the data query method and the device provided by the invention, the correlation between the candidate segments and the query statement can be evaluated; the posterior recall quality model can evaluate the ability of the candidate segments to recall documents; and based on this, the candidate segment with the confidence probability meeting the screening condition is used for querying, the number of the obtained recall documents is high, and the relevance of the recalled documents is also large, so that the relevancy of the query results is guaranteed.

Description

technical field [0001] The present invention relates to the technical field of data processing, in particular to a data query method, device and readable medium. Background technique [0002] In the search, the document recall is obtained by intersecting the inverted list of each keyword obtained by segmenting the query statement (query). However, in a medium-length query, if the inverted lists of all keywords participate in the intersection, it will result in insufficient recall documents or related documents cannot be recalled. For example, if the query is "Can swisse vitamins be eaten by pregnant women during breastfeeding?", if all the keywords of the query are included in the recall, the returned query result will be empty. Refer to Figure 1a As shown, but if only the three keywords "swisse vitamin lactation period" are allowed to participate in the cross-request recall, some documents that do not exactly match the query but are also very relevant to the query requirem...

AI Technical Summary

Problems solved by technology

[0004] Embodiments of the present invention provide a data query method, device, and readable medium to solve the problem in the prior art that the query results obtained by using keywords in the query statement are not highly relevant

Images