FIELD: computer engineering.
SUBSTANCE: group of inventions can be used to generate text. Method comprises steps of: obtaining input data containing a source text in a natural language and a target style of the generated text; source text is encoded; performing tokens vectorization; vector representations of source text tokens are processed using a machine learning model based on a neural network, trained on texts stylized in accordance with a given target style, during which an array of vectorized stylized texts is formed; performing decoding of each vectorized stylized text from array, wherein during decoding performing at least conversion of vectorized stylized text into tokens and detokenization; filtering the array of stylized texts; performing ranking of filtered stylized texts; styled text is generated.
EFFECT: high semantic accuracy of generating a stylized text from a source text.
9 cl, 3 dwg, 1 tbl
Title | Year | Author | Number |
---|---|---|---|
METHOD AND SYSTEM FOR DIGITAL ASSISTANT TEXT GENERATION | 2022 |
|
RU2796208C1 |
METHOD AND SYSTEM FOR PARAPHRASING TEXT | 2023 |
|
RU2814808C1 |
TEXT CLASSIFICATION METHOD AND SYSTEM | 2022 |
|
RU2818693C2 |
SYSTEM AND METHOD FOR CORRECTING SPELLING ERRORS | 2020 |
|
RU2753183C1 |
METHOD AND SYSTEM FOR DEPERSONALIZATION OF CONFIDENTIAL DATA | 2022 |
|
RU2804747C1 |
METHOD AND SYSTEM FOR RETRIEVING NAMED ENTITIES | 2020 |
|
RU2760637C1 |
METHOD AND SYSTEM FOR DEPERSONALIZATION OF CONFIDENTIAL DATA | 2022 |
|
RU2802549C1 |
NAMED ENTITIES FROM THE TEXT AUTOMATIC EXTRACTION | 2014 |
|
RU2665239C2 |
METHOD AND SYSTEM FOR EXTRACTING NAMED ENTITIES | 2021 |
|
RU2823914C2 |
SYSTEM AND METHOD FOR AUGMENTATION OF THE TRAINING SAMPLE FOR MACHINE LEARNING ALGORITHMS | 2020 |
|
RU2758683C2 |
Authors
Dates
2024-04-16—Published
2023-06-02—Filed