Systems and methods for synthetic database query generation
A system for returning synthetic database query results. The system may include a memory unit for storing instructions, and a processor configured to execute the instructions to perform operations comprising: receiving a query input by a user at a user interface; determining, based on natural language processing, a type of the query input; determining, based on the received query input and a database language interpreter, an output data format; returning, based on a generation model and the output data format, a result of the query input; providing, to a plurality of training models and based on the determined query type, the query input and the result; and training the training models, based on the query input and the result.
1. A system for generating synthetic results based on a query input, comprising:
one or more processors and memories storing instructions that, when executed by the one or more processors, cause operations comprising:
in response to a user interaction, with a user interface, involving a query input, inputting the query input into a natural language processing model to derive a type of the query input;
routing, via a network, based on the type of the query input, the query input to a trained model of a plurality of models, wherein the trained model is trained to generate synthetic values for a selected subclass within a class and not for other subclasses within the class; and
based on the routing of the query input to the trained model, inputting the query input to the trained model to generate a subclass-specific synthetic dataset that satisfies a statistical similarity criterion associated with both the synthetic dataset and a reference dataset, wherein the statistical similarity criterion is one or more of a statistical correlation score between the synthetic dataset and the reference dataset, a data similarity score between the synthetic dataset and the reference dataset, or a data quality score for the synthetic dataset.
2. The system of claim 1 , the operations further comprising:
returning, via a network, based on the subclass-specific synthetic dataset, one or more synthetic query results.
3. The system of claim 1 , the operations further comprising:
retrieving, from a network database, a data record comprising one or more record portions and one or more sensitive record portions; and
returning, via a network, based on the data record and the subclass-specific synthetic dataset, one or more synthetic query results comprising (i) the one or more record portions without the one or more sensitive record portions and (ii) one or more synthetic data portions of the subclass-specific synthetic dataset in lieu of the one or more sensitive record portions.
4. The system of claim 1 , wherein routing the query input to the trained model comprises inputting one or more query portions of the query input to a classifier model to determine the selected subclass and, based on the selected subclass and the type of the query input, routing the query input to the trained model.
5. The system of claim 1 , wherein the statistical similarity criterion comprises a statistical correlation score between the synthetic dataset and the reference dataset.
6. The system of claim 1 , wherein the trained model comprises a neural network, recurrent neural network, generative adversarial network, kernel density estimator, or random value generator.
7. A method comprising:
executing, via one or more processors, operations comprising:
based on a user interaction, with a user interface, involving a query input, inputting the query input into a natural language processing model to derive a type of the query input;
routing, via a network, based on the type of the query input, the query input to a trained model that is trained to generate synthetic values for a selected subclass within a class and not for other subclasses within the class; and
based on the routing of the query input to the trained model, inputting the query input to the trained model to generate a subclass-specific synthetic dataset that satisfies a statistical similarity criterion associated with both the synthetic dataset and a reference dataset, wherein the statistical similarity criterion comprises one or more of a statistical correlation score between the synthetic dataset and the reference dataset, a data similarity score between the synthetic dataset and the reference dataset, or a data quality score for the synthetic dataset.
8. The method of claim 7 , the operations further comprising:
returning, via a network, based on the subclass-specific synthetic dataset, one or more synthetic query results.
9. The method of claim 7 , the operations further comprising:
retrieving, from a network database, a data record comprising one or more record portions and one or more sensitive record portions; and
returning, via a network, based on the data record and the subclass-specific synthetic dataset, one or more synthetic query results comprising (i) the one or more record portions without the one or more sensitive record portions and (ii) one or more synthetic data portions of the subclass-specific synthetic dataset in lieu of the one or more sensitive record portions.
10. The method of claim 7 , the operations wherein routing the query input to the trained model comprises inputting one or more query portions of the query input to a classifier model to determine the selected subclass and, based on the selected subclass and the type of the query input, routing the query input to the trained model.
11. The method of claim 7 , wherein the statistical similarity criterion comprises a statistical correlation score between the synthetic dataset and the reference dataset.
12. The method of claim 7 , wherein the statistical similarity criterion comprises a data similarity score between the synthetic dataset and the reference dataset.
13. The method of claim 7 , wherein the trained model comprises a neural network, recurrent neural network, generative adversarial network, kernel density estimator, or random value generator.
14. One or more non-transitory computer-readable media storing instructions that, when executed by one or more processors, cause operations comprising:
based on a user interaction, with a user interface, involving a query input, inputting the query input into a natural language processing model to derive a type of the query input;
routing, via a network, based on the type of the query input, the query input to a trained model that is trained to generate synthetic values for a selected subclass within a class and not for other subclasses within the class; and
based on the routing of the query input to the trained model, inputting the query input to the trained model to generate a subclass-specific synthetic dataset that satisfies a statistical similarity criterion associated with both the synthetic dataset and a reference dataset, wherein the statistical similarity criterion comprises one or more of a statistical correlation score between the synthetic dataset and the reference dataset, a data similarity score between the synthetic dataset and the reference dataset, or a data quality score for the synthetic dataset.
15. The one or more non-transitory computer-readable media of claim 14 , the operations further comprising:
returning, via a network, based on the subclass-specific synthetic dataset, one or more synthetic query results.
16. The one or more non-transitory computer-readable media of claim 14 , the operations further comprising:
retrieving, from a network database, a data record comprising one or more record portions and one or more sensitive record portions; and
returning, via a network, based on the data record and the subclass-specific synthetic dataset, one or more synthetic query results comprising (i) the one or more record portions without the one or more sensitive record portions and (ii) one or more synthetic data portions of the subclass-specific synthetic dataset in lieu of the one or more sensitive record portions.
17. The one or more non-transitory computer-readable media of claim 14 , the operations wherein routing the query input to the trained model comprises inputting one or more query portions of the query input to a classifier model to determine the selected subclass and, based on the selected subclass and the type of the query input, routing the query input to the trained model.
18. The one or more non-transitory computer-readable media of claim 14 , wherein the statistical similarity criterion comprises a statistical correlation score between the synthetic dataset and the reference dataset.
19. The one or more non-transitory computer-readable media of claim 14 , wherein the statistical similarity criterion comprises a data similarity score between the synthetic dataset and the reference dataset.
20. The one or more non-transitory computer-readable media of claim 14 , wherein the trained model comprises a neural network, recurrent neural network, generative adversarial network, kernel density estimator, or random value generator.