Keyword determination method and keyword determination device

Abstract

The embodiment of the invention discloses a keyword determination method and a keyword determination device. The keyword determination method comprises the steps of acquiring an entry to be searched; segmenting the to-be-searched entry of a user according to the preset sequential character segmentation rule to obtain a to-be-searched sub-entry set, wherein the to-be-searched sub-entry set comprises at least one to-be-searched sub-entry, the to-be-searched sub-entry is the partial content or the whole content of the to-be-searched entry; searching a target sub-entry which is the same as the to-be-searched sub-entry in the acquired to-be-searched sub-entry set in a target word bag stored in advance and at least comprising one target sub-entry; and determining the searched same target sub-entry as a keyword corresponding to the to-be-searched entry after searching the same target sub-entry. The method in the embodiment of the invention can improve the data processing speed, and the accuracy of the determined keyword is high.

Description

technical field [0001] The embodiments of the present invention relate to the field of keywords, and in particular to a method and device for determining keywords. Background technique [0002] With the increase of big data, users have higher and higher requirements for processing big data methods. In practical applications, there is often a need to determine the same terms that exist in the user's search term and the given word package. Hereinafter, these same terms are called keywords, and the determined keywords It can be used to analyze user behavior characteristics, recommend information to users, etc. [0003] There are two existing methods for determining keywords: [0004] First, find the same sub-entry in the user's to-be-searched term by looping through the sub-entries in the given word bag, and then determine the same sub-entry found as a keyword, for example: a The user's entry to be searched is "the name is Li Mingming", and there are 1,000 entries in the wor...

AI Technical Summary

Problems solved by technology

[0004] First, find the same sub-entry in the user's to-be-searched term by looping through the sub-entries in the given word bag, and then determine the same sub-entry found as a keyword, for example: a The user's entry to be searched is "the name is Li Mingming", and there are 1,000 entries in the word bag, then each sub-entry in the word bag needs to be searched in the word to be searched, so that 1,000 entries are searched. Times, this is only for one entry to be searched, for multiple entries to be searched, the number of searches will be more, the increase in the number of searches will increase the complexity of the search algorithm, and the increase in search time will slow down the data processing speed

[0005] Second, find the same sub-entry in a given word bag by looping through the sub-entries of the searched entry after word segmentation, and then determine the same sub-entry found as a keyword, and the word segmentation is based on The entries in the corpus obtained from corpus training are divided into entries to be searched. For example, in the above example, there may be entries such as "name", "yes" and "Li Mingming" in the corpus. After word segmentation, "name is Li Mingming" can be " Name", "Yes", "Li Mingming", this method is to check whether the three sub-entries of "name", "Yes", and "Li Mingming" have the same sub-entries in the word bag, only need to search three times, and Compared with the first method, the obvious reduction in the number of searches reduces the complexity of the algorithm, and the shortening of the search time makes the data processing speed faster. However, the existing word segmentation is limited by the entries in the corpus, and there are often some words after word segmentation. If the sub-entry does not match the meaning of the original entry to be searched, if there is no "Li Mingming" in the corpus, but there are entries such as "Li Ming" and "Ming", the above-mentioned "name is Li Mingming" can be " Name", "Yes", "Li Ming" and "Ming", so if there is "Li Ming" in the given word bag, then "Li Ming" will be determined as the keyword. Obviously, "Li Ming" is the same as the original The meaning of "Li Mingming" in the search term is different, which directly affects the accuracy of the determined keyword

Images