Computing device for processing data management tasks

Abstract

A computing device for processing data management tasks runs a front-end natural language model. After obtaining an intent semantics, the computing device uses the front-end natural language model to obtain a data management task plan. For each data management sub-task, the computing device obtains a data processing sub-method. Based on the data processing sub-methods, the computing device obtains a data processing method and a data management dialogue script. The computing device obtains and transmits at least one task-oriented dialogue to a user device to obtain at least one user response. For each user response, the computing device determines whether the data processing method passes verification. When the data processing method passes verification, it is designated as an executable data processing method.

Claims

1. A computing device for processing data management tasks, communicatively connected to a user device via a communication network, the computing device comprising: a communication unit communicatively connected to the user device via the communication network; a storage unit storing a front-end natural language model and multiple secondary natural language models; and a processing unit electrically connected to the communication unit and the storage unit; wherein, The processing unit obtains an intent semantic from the user device related to a data management task. Based on the intent semantic, the processing unit uses the front-end natural language model to obtain a data management task plan, which includes multiple data management sub-tasks. For each data management sub-task, the processing unit uses the front-end natural language model and the secondary natural language models to obtain a data processing sub-method corresponding to the data management sub-task. Based on the data processing sub-methods, the processing unit uses the front-end natural language model to generate a data processing method corresponding to the data management task plan and a data management dialogue script corresponding to the data processing method. Based on the data management dialogue script, the processing unit obtains and transmits at least one task-oriented dialogue to the user device to obtain at least one user response. Based on the at least one user response and the data processing sub-methods, the processing unit uses the front-end natural language model and the secondary natural language models to determine whether the data processing method passes verification. When the processing unit determines that the data processing method passes verification, it treats the data processing method as an executable data processing method.

2. The computing device for processing data management tasks as described in claim 1, wherein the processing unit obtains a data management task request using the front-end natural language model based on multiple key semantics contained in the intent semantics, and the processing unit obtains a data management task plan using the front-end natural language model based on the data management task request.

3. The computing device for processing data management tasks as described in claim 1, wherein for each data management subtask, the processing unit obtains a subtask natural language model corresponding to the data management subtask from the secondary natural language models using the front-end natural language model, and the processing unit obtains the data processing sub-method using the subtask natural language model.

4. The computing device for processing data management tasks as described in claim 1, wherein when the processing unit determines that the data processing method has failed verification, it generates at least one supplementary information prompt message related to the requested supplementary information using the front-end natural language model, and transmits the at least one supplementary information prompt message to the user device; after receiving at least one supplementary intent semantic from the user device corresponding to the at least one supplementary information prompt message, the processing unit obtains another data management task plan based on the intent semantic and the at least one supplementary intent semantic using the front-end natural language model, the other data management task plan including multiple other data management sub-tasks, and for each other data management sub-task, the processing unit... Using the front-end natural language model and the secondary natural language models, another data processing sub-method is obtained. The processing unit generates another data processing method and another data management dialogue script based on the other data processing sub-method. The processing unit obtains and transmits at least one other task-oriented dialogue to the user device based on the other data management dialogue script to obtain at least one other user response information. For each other user response information, the processing unit uses the front-end natural language model and the secondary natural language models to determine whether the other data processing method passes verification. When the processing unit determines that the other data processing method passes verification, it uses the other data processing method as another executable data processing method.

5. The computing device for processing data management tasks as described in claim 1, wherein for each user reply information, the processing unit obtains a target sub-method corresponding to the user reply information from the data processing sub-methods using the front-end natural language model based on the user reply information; the processing unit obtains a target natural language model corresponding to the data management sub-task from the secondary natural language models based on the target sub-method; the processing unit determines whether the target sub-method passes verification using the target natural language model based on the user reply information; when the processing unit determines that all target sub-methods pass verification, it determines that the data processing method passes verification and uses the data processing method as the executable data processing method.

Images