Text retrieval method and device

Abstract

The invention discloses a text retrieval method and device. The method comprises the following steps: retrieving at least one text according to a keyword and retrieval dimension information after acquiring the keyword and the retrieval dimension information required for retrieving, thereby obtaining at least one hit text including the keyword, arranging at least one hit text according to the retrieval dimension information to obtain the sorting of each hit text in all hit texts, wherein the retrieval dimension information is used for indicating the performing of the retrieval through at leastone retrieval dimension of the text and indicating the display important degree of each retrieval dimension in at least one retrieval dimension; which of retrieval dimensions of the text can be selected to perform retrieval and the display important degree of each retrieval dimension can be set according to the user intention, and then the keyword retrieval can be performed on the content corresponding to the retrieval dimension selected by the user, and the hit texts can be arranged according to the set display important degree, thereby improving the user experience.

Description

technical field [0001] The present invention relates to the technical field of data retrieval, and more particularly, to a text retrieval method and device. Background technique [0002] Text retrieval refers to retrieving each text in the text collection according to the text content to obtain the text that meets the user's intention. For example, each text in the text collection can be retrieved according to keywords or semantics to obtain text that meets the user's intention. After obtaining the text that meets the user's intention, the text that meets the user's intention can be quantified according to the degree of relevance to a certain extent. [0003] The so-called quantification according to the degree of relevance refers to scoring the relevance of the text according to the degree of relevance between the text and the user's intention. It is considered that the higher the relevance score, the better the text is in line with the user's intention. Therefore, sorting...

AI Technical Summary

Problems solved by technology

[0004] However, in actual retrieval, multiple parts of any text will be retrieved, such as a keyword query for the title and text of any text at the same time, and the content of the text is far more than the content of the title, so the text part The probability of hitting keywords is far greater than the probability of hitting keywords in the title part, so that the relevance score of the body part is much higher than that of the title part. In this way, there will be a situation: all texts that are displayed preferentially will have The title of some text does not include keywords, so that users cannot determine whether the text meets their own intentions through the title. Therefore, users need to click to view the text to determine whether the text that is displayed first best meets the user's intention, which reduces user experience.

Images