• The text content extraction module is used to extract structured data from text, such as user names, mobile phone numbers, etc.

• Example: For example, a dental clinic appointment assistant needs to extract structured information such as name, phone number, time, etc. from user questions, and then call the HTTP interface to make an appointment with the dentist.

history record

Usually some historical records are needed to more completely extract user issues. For example, in the picture above, the name, time, and phone number need to be provided. The user may only provide the time and phone number at the beginning, but not his or her name. After another round of missing prompts, the user entered their name. At this time, it is necessary to combine the last record to fully extract the three contents.

Target field

The target field corresponds to the extracted result. As you can see from the above figure, for each additional field, a corresponding outlet will be added to the output.

• key is the unique identifier of the field and cannot be repeated!

• Field description: Describe what the field is about, such as name, time, phone number, etc.

• Required: Whether to force the model to extract this field, which may be an empty string.

Output introduction

• Complete field extraction: It means that the user's question contains all the content that needs to be extracted.

• Extract field missing: Opposite of Field extracted completely, triggered when there are missing extracted fields.

• Complete extraction result: A JSON string containing the extraction results for all fields.

• Target field extraction results: all types are strings.