Text generation method and device, electronic equipment and storage medium

Abstract

The present disclosure relates to the field of data processing, and specifically discloses a text generation method and device, electronic equipment and storage medium. The method comprises: obtaining source text data, converting the source text data into a source text vector, performing task processing corresponding to a preset task type on the source text vector to obtain a predicted text vector corresponding to the source text vector; querying an associated text vector corresponding to the predicted text vector from a pre-stored vector text mapping file, and obtaining associated text data corresponding to the associated text vector; adjusting the predicted text data corresponding to the predicted text vector according to the associated text data, and obtaining target text data according to the adjustment result. Since the acquisition threshold of unlabeled sample data is lower and the data sources are more extensive, it is convenient to determine text content covering different business fields according to massive corpus data, thereby improving the generality of the text generation method.

Images