Systems and methods of generative machine-learning guided by modal classification
AssigneeClioTech Ltd
InventorGeorgia Helen Kirke
Life of this patent
Estimating the term…
Abstract
Described herein are systems and methods of generative machine-learning guided by modal classification. An apparatus may receive a characterization datum and output an inquiry as a function of the characterization datum. An apparatus may receive an inquiry response and may output a plurality of draft literature modes as a function of the characterization datum and the inquiry response. An apparatus may receive template feedback and may output a literature mode as a function of the template feedback and the plurality of draft literature modes.
Claims18
An apparatus for generative machine-learning guided by modal classification, the apparatus comprising:
- at least a processor;
- and
- a memory communicatively connected to the at least processor, the memory containing instructions configuring the at least processor to:
- receive a characterization datum from a user device operated by a user;
- generate an inquiry as a function of the characterization datum;
- communicate the inquiry to the user device;
- receive an inquiry response from the user device;
- generate a plurality of draft literature modes using a literature mode machine learning model as a function of the characterization datum and the inquiry response, wherein using the literature mode machine learning model comprises:
- training an art element machine learning model on a training dataset including example text characterization data and example inquiry responses, associated with example draft layout templates;
- inputting into the layout template machine learning model the characterization datum and the inquiry response;
- receiving as an output from the art element machine learning model a draft art element;
- receiving user feedback on the draft art element;
- determining an accuracy score of the art element machine learning model based on the user feedback,
- generating an updated training dataset based on the accuracy score;
- and
- iteratively retraining the art element machine learning model using the updated training dataset;
- communicate the plurality of draft literature modes to the user device;
- receive template feedback from the user device;
- and
- output a literature mode as a function of the plurality of draft literature modes and the template feedback.
A method for generative machine-learning guided by modal classification, the apparatus comprising:
- using at least a processor, receiving a characterization datum from a user device operated by a user;
- using at least a processor, generating an inquiry as a function of the characterization datum;
- using at least a processor, communicating the inquiry to the user device;
- using at least a processor, receiving an inquiry response from the user device;
- using at least a processor, generating a plurality of draft literature modes using a literature mode machine learning model as a function of the characterization datum and the inquiry response, wherein using the literature mode machine learning model comprises:
- training an art element machine learning model on a training dataset including example text characterization data and example inquiry responses, associated with example draft layout templates;
- inputting into the layout template machine learning model the characterization datum and the inquiry response;
- receiving as an output from the art element machine learning model a draft art element;
- receiving user feedback on the draft art element;
- determining an accuracy score of the art element machine learning model based on the user feedback,
- generating an updated training dataset based on the accuracy score;
- and
- iteratively retraining the art element machine learning model using the updated training dataset,
- using at least a processor, communicating the plurality of draft literature modes to the user device;
- using at least a processor, receiving template feedback from the user device;
- and
- using at least a processor, outputting a literature mode as a function of the plurality of draft literature modes and the template feedback.
Dependencies are inferred from the claim text.