Sentence association method and device

Abstract

The embodiment of the invention discloses a sentence association method and a related device. According to the above content that is on the screen, network search can be carried out by taking the above content as a keyword to obtain corresponding search results. The corpus related to the above content on the network is far more than the model word bank of the local high-frequency sentence model, therefore, according to the search result, the next sentence corresponding to the above content can be determined more possibly, the sentence association candidate item corresponding to the above content is generated according to the determined next sentence, the effect of providing input convenience for the user is achieved, the sentence association function is achieved in a search mode, and the application range of the sentence association function is widened.

Description

technical field [0001] This application relates to the field of input methods, in particular to a sentence association method and device. Background technique [0002] The input method system can associate the following candidates based on the content on the user's screen during the user's input process, so that the user can directly go to the screen by selecting the candidate, which provides convenience for the user's input. Through the sentence association function, the input method system can use the associated sentence as a sentence association candidate. The sentence association candidate includes a sentence, and the sentence can be a phrase or a long sentence composed of multiple words. If the user selects a sentence association candidate, more content can be displayed on the screen at one time, which improves the input efficiency. [0003] At present, the N-Gram model commonly used in large vocabulary continuous speech recognition is mainly used to realize the functi...

AI Technical Summary

Problems solved by technology

[0004] However, since the current model training is based on the user's historical corpus, and the off-line model lexicon occupies strict storage requirements, it is currently only possible to extract high-frequency sentences from the above content that has been uploaded on the screen to train the high-frequency sentence model, and use high-frequency The sentence model implements the sentence association function to determine whether the current above content has sentence association candidates for the following sentences, so that the scope of application of the sentence association function is not large

Images