Method and system for transforming written text into oral text

Abstract

The invention discloses a method for transforming a written text into an oral text. The method comprises the steps that to-be-transformed source text data is received; word segmentation and vectorization processing is performed on the source text data, and a word vector sequence of each source text data sentence is obtained; the word vector sequences of all the source text data sentences are sequentially input into a text transformation model constructed in advance, and target text data corresponding to the source text data is obtained according to output of the text transformation model; andparalanguage information is interposed into the target text data, and oral text data with the paralanguage information is obtained. By use of the method, the oral text obtained after transformation can better conform to the oral expression habit.

Description

technical field [0001] The invention relates to the field of natural language processing, in particular to a method and system for converting written text into spoken text. Background technique [0002] Language is the system people use to communicate and usually has two different forms of expression, spoken and written. Spoken language is spoken language, and written language is written language, both of which have different characteristics. Generally speaking, spoken language is more flexible and shorter than written language, and is more dependent on context. Expressions are often accompanied by paralinguistic phenomena, such as panting, dragging, and pauses, which make spoken language sound more natural and easier to understand than written language. Therefore, in order to facilitate human understanding, the researchers propose that written texts can be converted into colloquial texts. [0003] When existing text conversion methods convert written language into spoken ...

AI Technical Summary

Problems solved by technology

[0004] The existing text conversion methods directly analyze the written text and add the corresponding paralanguage to obtain the converted colloquial text. The converted colloquial text simply adds common paralinguistic information in spoken language to the written text. The colloquial text does not conform to the expression habits of the language. Sometimes the colloquial paralinguistic information in the written text will make the user feel blunt, awkward, and not smooth, which seriously affects the user experience.

Images