Rapid text generation method, computer readable storage medium and electronic equipment

Abstract

The invention discloses a text rapid generation method, a computer readable storage medium and electronic equipment. The text rapid generation method comprises the following steps: obtaining table data; carrying out word segmentation on the content in the table data and then carrying out word vectorization processing to obtain a word vectorization sequence of a table attribute layer and a word vectorization sequence of a table content layer; respectively encoding the word vectorization sequence of the table attribute layer and the word vectorization sequence of the table content layer, and decoding in combination with an attention mechanism to obtain a target word id sequence; and performing text conversion on the target word id sequence, and performing associated word replacement and unknown word copying to generate a final word sequence serving as an output text. Therefore, the text rapid generation method can improve the text generation speed while ensuring that the text generation is correct.

Description

technical field [0001] The invention relates to the technical field of natural language generation, in particular to a method for rapidly generating text, a computer-readable storage medium and an electronic device. Background technique [0002] Natural language generation technology is an important branch of the field of natural language processing, specifically referring to the technology of generating natural language text from machine-readable data. Table-to-text generation is an important research direction in the field of natural language generation. It is often used in some applications, such as search engines and question answering systems, which can help people better understand tables, especially those tables that use complex data, such as Weather data and medical data, etc. [0003] Tabular data is different from text or images, which only have a single-dimensional sequence or have no correlation between dimensions. It is composed of attributes and content. There...

AI Technical Summary

Problems solved by technology

Due to the large amount of vocabulary in the text and the limited size of the target vocabulary, often only the words that appear many times are counted, and words that appear very few times cannot be included. Therefore, in the process of generating text, these words will be used as unknown words. Symbols, but these words may be very important in the text, such as rare names, place names, etc., resulting in inaccurate generated text

Images