Office method and device, electronic equipment, storage medium

Abstract

The application relates to the office automation technical field and discloses an office method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring first configuration information; wherein the first configuration information is obtained in response to a first configuration operation of a user in a preset user configuration interface; acquiring a first user dialogue in response to a first input operation of the user in a preset user dialogue interface; determining a to-be-filled form according to the first user dialogue; determining field content corresponding to each form field according to a form field corresponding to the to-be-filled form, a field attribute corresponding to each form field and the first user dialogue; combining the field content to form a form; and submitting the form to a third-party system according to an access authentication parameter of the third-party system. In this way, the office system can realize a full-automatic process from form formation to submission, and the generation result of the office system meets the user expectation.

Description

Technical Field

[0001] This application relates to the field of office automation technology, and in particular to an office method and apparatus, electronic device, and storage medium. Background Technology

[0002] Currently, with the improvement of large-scale model capabilities, more and more users are starting to use intelligent office systems.

[0003] Building a complete intelligent office system by oneself is very complex, but if you directly use a mature intelligent office system on the market, the generated form results may not fully meet the user's expectations.

[0004] Therefore, it is urgent to solve the problem of how to provide an office system that is easy to use and produces results that meet user expectations.

[0005] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this application, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] To provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not intended as a general commentary, nor is it intended to identify key / important components or describe the scope of protection of these embodiments, but rather as a prelude to the detailed description that follows.

[0007] This application provides an office method and apparatus, electronic device, and storage medium to facilitate the configuration of an office system and ensure that the generated results of the office system meet user expectations.

[0008] This application provides an office method, comprising: obtaining first configuration information; the first configuration information including: access authentication parameters of a third-party system and construction parameters of different forms; the construction parameters including: form fields and field attributes corresponding to the form fields; wherein, in a preset user configuration interface, the first configuration information is obtained in response to a user's first configuration operation; in a preset user dialogue interface, a first user dialogue is obtained in response to a user's first input operation; a form to be filled is determined based on the first user dialogue; the field content corresponding to each form field is determined based on the form fields corresponding to the form to be filled, the field attributes corresponding to each form field, and the first user dialogue; the field content is combined to form a target form; and the form is submitted to the third-party system according to the access authentication parameters of the third-party system.

[0009] In the above embodiments, by configuring different form construction parameters, the office system can automatically extract parameters based on the user's required data, thereby constructing a target form that meets the user's expectations. Furthermore, by configuring access authentication parameters for third-party systems, the office system can automatically submit the target form to the corresponding third-party system, eliminating the need for the user to submit the form manually. In other words, through this application, users can achieve a fully automated process from form creation to submission by simply configuring the office system, and the generated results from the office system meet the user's expectations.

[0010] Further, determining the form to be filled based on the first user dialogue includes: obtaining second configuration information; the second configuration information includes: a purpose description and intent use cases; wherein, in the user configuration interface, the second configuration information is obtained in response to the user's second configuration operation; the purpose description describes what intents are triggered in different scenarios; the intent use cases include dialogue examples of users under different intents; inputting a first prompt word and the first user dialogue into a preset large language model to obtain the user intent; the first prompt word is used to describe the user intent in the user dialogue determined by referring to the purpose description and the intent use cases; and determining the form to be filled based on the user intent.

[0011] In the above implementation, by having the user configure the model's purpose description and intent use cases, the model can fully understand various situations in the user's usage scenario, thereby deriving the form to be filled that best meets the user's needs in the user's configured scenario.

[0012] Furthermore, determining the form to be filled based on the user intent includes: determining the intent type corresponding to the user intent; if the intent type belongs to a submission intent, searching for the form to be filled corresponding to the user intent in a preset form association relationship; the form association relationship is used to reflect the association relationship between the user intent and the form to be filled.

[0013] In the above implementation, by associating the intent type with the configured form to be filled in in advance, the corresponding form to be filled in can be accurately determined by searching, which is convenient and fast.

[0014] Furthermore, determining the field content corresponding to each form field based on the form fields corresponding to the form to be filled, the field attributes corresponding to each form field, and the first user dialogue includes: combining the form fields corresponding to the form to be filled, the field attributes corresponding to the form fields, and a preset prompt word template to obtain a second prompt word; the second prompt word is used to describe the field content corresponding to each form field extracted from the dialogue according to the field attributes of the form fields, and if the field content corresponding to a certain form field does not exist, then the field content corresponding to that form field is empty; inputting the second prompt word and the first user dialogue into a preset large language model to obtain the field content corresponding to each form field.

[0015] In the above implementation, the large language model has excellent learning and understanding capabilities. By combining form fields, the field attributes corresponding to the form fields, and prompt word templates, the large language model can accurately obtain the field content of the form fields to be extracted and output the content according to the field attributes required by the form fields.

[0016] Furthermore, the method also includes: if there is a form field whose content is empty, the form field with empty content is recorded as an incomplete field; the incomplete field is combined with a preset display template to obtain content to be displayed; the content to be displayed is used to prompt the user to fill in the field content of the incomplete field; and the content to be displayed is displayed to the user.

[0017] Considering that users may not provide all the required information for the form to be filled in at once, the above implementation uses the large language model to filter out incomplete fields whose content was not retrieved, and then queries those incomplete fields again. This ensures that the content of the form to be filled in is complete, thus enabling successful submission.

[0018] Furthermore, the method also includes: acquiring a second user dialogue; combining the incomplete field, the field attribute corresponding to the incomplete field, and the prompt word template to obtain a third prompt word; the third prompt word is used to describe the field content corresponding to each incomplete field extracted from the dialogue according to the field attribute of the incomplete field, and if the field content corresponding to a certain incomplete field does not exist, then the field content corresponding to that incomplete field is empty; inputting the third prompt word and the second user dialogue into the large language model to obtain the field content corresponding to each incomplete field respectively.

[0019] In the above implementation, the large language model has excellent learning and understanding capabilities. By combining incomplete fields, the field attributes corresponding to incomplete fields, and prompt word templates, the large language model can accurately obtain the field content of the incomplete fields that need to be extracted, and output the content according to the field attributes required by the form fields.

[0020] Furthermore, the construction parameters also include: form layout; combining the contents of each field to form a target form, including: arranging the contents of each field according to the form layout to form a target form.

[0021] In the above implementation, by allowing users to customize the form layout, the resulting target form can meet their specific needs. Furthermore, by customizing the form layout, subsequent users of the same form can generate target forms in the same format, ensuring consistency and facilitating company management.

[0022] This application embodiment also provides an office device, including: a configuration acquisition module, used to acquire first configuration information; the first configuration information includes: access authentication parameters of a third-party system and construction parameters of different types of forms; the construction parameters include: form fields and field attributes corresponding to the form fields; wherein, in a preset user configuration interface, the first configuration information is acquired in response to a user's first configuration operation; a user input module, used to acquire a first user dialogue in response to a user's first input operation in a preset user dialogue interface; an intent module, used to determine a form to be filled based on the first user dialogue; a text processing module, used to determine the field content corresponding to each form field based on the form fields corresponding to the form to be filled, the field attributes corresponding to each form field, and the first user dialogue; and to combine the field content to form a target form; and a submission module, used to submit the target form to the third-party system according to the access authentication parameters of the third-party system.

[0023] This application provides an electronic device, including a processor and a memory. The memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement the above-described office method.

[0024] In some embodiments, the storage medium stores computer-executable instructions that, when invoked and executed by a processor, cause the processor to implement the above-described office methods.

[0025] The above general description and the description below are exemplary and illustrative only and are not intended to limit this application. Attached Figure Description

[0026] One or more embodiments are illustrated by way of example with reference to the accompanying drawings. These illustrations and drawings do not constitute a limitation on the embodiments. Elements having the same reference numerals in the drawings are considered similar elements. The drawings do not constitute a limitation of scale, and wherein: Figure 1 This is a flowchart illustrating an office method provided in an embodiment of this application; Figure 2 This is a schematic diagram of the structure of an office device provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0027] Figure label: 1: Configuration acquisition module; 2: User input module; 3: Intent module; 4: Text processing module; 5: Submission module; 6: Processor; 7: Memory; 8: Communication interface; 9: Bus. Detailed Implementation

[0028] To provide a more detailed understanding of the features and technical content of the embodiments of this application, the implementation of the embodiments of this application will be described in detail below with reference to the accompanying drawings. The accompanying drawings are for illustrative purposes only and are not intended to limit the embodiments of this application. In the following technical description, for ease of explanation, several details are used to provide a full understanding of the disclosed embodiments. However, one or more embodiments may still be implemented without these details. In other cases, well-known structures and devices may be simplified in their depiction to simplify the drawings.

[0029] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion.

[0030] Unless otherwise stated, the term "multiple" means two or more.

[0031] The term "correspondence" can refer to an association or binding relationship. The correspondence between A and B means that there is an association or binding relationship between A and B.

[0032] Example 1 As large-scale model capabilities continue to improve, more and more organizations are introducing agent platforms to achieve intelligent upgrades of their business systems. Agent platforms can be used to submit processes, fill out forms, and query data via voice dialogue through PCs or mobile devices.

[0033] However, building an intelligent agent platform adapted to an organization's office scenarios typically requires complex configurations. For example, it necessitates integrating organizational structures and personnel permission systems across systems to achieve seamless connectivity between the current intelligent agent platform and the target system's login / business application / data query layers. It also requires constructing an intent recognition system, enabling the model to guide users to different operational intents based on their needs, while maintaining strong configuration flexibility to adapt to the intent recognition requirements of different industry clients in various scenarios. Furthermore, it requires disassembling the data structure of the target system; for instance, to implement conversational form filling, the details of each field in the process form must be broken down; to implement conversational data querying, it is necessary to delve into the database layer to complete data queries or obtain API data, and so on.

[0034] All of these configurations require developers with extensive development and business experience to achieve the goal of building a conversational office assistant based on a general-purpose intelligent agent platform and applying it in business systems. Furthermore, due to the high complexity of intelligent agent platforms, achieving the desired results requires developers to have a thorough understanding of the platform's functions and to invest significant human and time resources.

[0035] This application provides an office method, and the software system that applies this office method is referred to as an office system.

[0036] Before users can automatically generate and submit forms using the office system, certain configurations need to be made in the office system so that the generated results can be adapted to the user's office scenario.

[0037] In one embodiment, first configuration information can be obtained in response to a user's first configuration operation on a preset user configuration interface.

[0038] Given that different software systems often cannot directly communicate with each other, but with increasing business needs and improved software system capabilities, many software systems are beginning to allow external systems to connect through open platforms and other functions to achieve data transmission between different software systems. This application can display third-party system configuration fields in a preset user configuration interface, allowing users to fill in these fields to enable data interaction between the office system and the third-party system.

[0039] Optionally, the user's first configuration operation may include configuration operations on configuration fields of the third-party system. In this case, the first configuration information obtained in response to the user's first configuration operation includes: access authentication parameters of the third-party system.

[0040] Third-party system configuration fields, such as: system type, system name, system address, global variable information, global request header information (global header information), etc.

[0041] The usage of third-party system configuration fields is as follows: System Type: Used to select or enter the system type. Multiple system types are supported. Examples of system types include: various versions of enterprise office automation platforms (OA systems), custom systems, etc.

[0042] System Name: Used to select or enter the name of the system. Custom system names are supported.

[0043] Access Address: Used to fill in the actual URL (Uniform Resource Locator) address for accessing the third-party system.

[0044] Global variables: Used to configure global variables that need to be added when interacting with the system. For example, you can fill in authentication information based on different access protocols.

[0045] The authentication information types may include three types: fixed values, scripts, and requests (authentication information is dynamically obtained from requests).

[0046] Optionally, you can fill in the necessary parameters for accessing the third-party system, such as SecretId (identity identifier) ​​and SecretIdKey (key) to ensure successful access to the third-party system.

[0047] Alternatively, the authentication token can be dynamically constructed using JavaScript (a code format) to carry token information that proves the user's identity when accessing a third-party system, thereby enabling successful access to the third-party system.

[0048] Global Header Information: This is used to input the header name and value as system parameters. These parameters are essential components for data exchange between software systems, enabling the construction of request parameters for each request. Typically, data exchange between software systems is based on HTTP (Hypertext Transfer Protocol) / HTTPS (Hypertext Transfer Secure Protocol) protocols.

[0049] Users only need to configure these third-party system configuration fields in a configurable way to establish a connection with the third-party system, without having to invest a lot of development in the process of synchronizing organizational structure and personnel.

[0050] Considering that different users have different office environments, this application aims to create forms that are best suited to each user's work environment. This is achieved by displaying form configuration fields in a pre-defined user configuration interface, allowing users to easily create the most suitable form.

[0051] Optionally, the user's first configuration operation may include configuration operations on form configuration fields. In this case, the first configuration information obtained in response to the user's first configuration operation includes: form fields for different types of forms, and field attributes corresponding to each form field.

[0052] The form configuration fields include: Form Field Management Item: After clicking on the Form Field Management Item, users can choose to delete or modify existing form fields in the user configuration interface, or add a form field with empty attribute information configuration items.

[0053] For each form field, there are one or more independent attribute configuration items. These attribute configuration items include, for example: field name, field format, field display attributes, field editing attributes, field default value, field association dictionary, and field description.

[0054] Users can edit each attribute configuration item to configure its properties. In subsequent use, the field attributes corresponding to form fields can be at least some of the attribute configuration items. For example, the field attribute corresponding to a form field can be the field format. The field attribute corresponding to a form field can also be the field format and the field association dictionary, etc.

[0055] For example, a preset user configuration interface displays form configuration fields, third-party system configuration fields, etc. A fill-in box can follow each configuration field. This fill-in box can be edited directly, selected from a dropdown menu, browsed using a button, or used for file uploads. When the user selects a dropdown / browse button, a related dictionary appears, allowing them to select or search for information to fill in.

[0056] Therefore, considering that user-entered fields may be of various types, such as text, numbers, and timestamps, using dropdown lists or browse boxes for field entry can lead to excessively long and unsightly dropdowns or browse boxes if a field has many attribute configuration options. Furthermore, finding the desired information can be complex for users. Constructing a dictionary can more conveniently meet users' needs for selecting multiple values ​​within the same field.

[0057] There are three sources of dictionary options depending on the scenario. In scenarios where the number of options is limited and fixed, dictionary options can be added manually. For example, options for a company name field.

[0058] For dictionaries that have already been built, you can also directly synchronize the options of those built dictionaries.

[0059] For scenarios where there are a large number of options that cannot all be listed at once, a real-time search method can be used to retrieve query field options based on keywords in real time.

[0060] Besides using the associated dictionary when filling in content in the input fields, the attribute information configuration item can also configure an associated dictionary. This configured associated dictionary is used for matching when extracting the content of form fields later, to obtain more accurate and consistent information.

[0061] Optionally, after the form fields and their attribute information configuration items are configured, the user's first configuration operation may also include a field combination operation. In this case, the first configuration information obtained in response to the user's first configuration operation includes: form layout.

[0062] During the configuration process, it also supports configuration of form submission (such as operations to be performed before form submission or when the form fails), linkage of form field attributes (field display / hiding triggered by changes in field values), and so on.

[0063] Considering that the same statement may refer to different meanings in different scenarios, this application trains the model to understand the intent of different office scenarios by displaying corpus intent configuration items and usage description configuration items in a preset user configuration interface. This allows the model to correctly understand and enter the corresponding intent when the user expresses themselves in natural language interaction.

[0064] In one embodiment, second configuration information can be obtained in response to a second configuration operation by the user in a preset user configuration interface.

[0065] The second configuration information includes: usage description and intended use cases.

[0066] Correspondingly, the second configuration operation can include uploading a usage description and intent test cases. In this way, by having users install / upload pre-configured usage descriptions and intent test cases, other members of the same organization can easily implement typical office scenarios without complex configuration.

[0067] In this application embodiment, the usage description describes what kind of intent is triggered in different scenarios. The usage description allows the model to understand the core of the user's intent.

[0068] For example, when configuring a leave request intent, the purpose description can be: trigger the leave request intent when a user needs to apply for leave. Guide the user through the leave application process, extract the form fields that need to be filled from the form to be filled out associated with the leave request intent, and extract the field content from the user's dialogue according to the form fields that need to be filled in, record it, and do not return any information.

[0069] Optionally, users can also directly describe the key field information that the model needs to extract in different scenarios in the usage description.

[0070] In one alternative approach, the key field information is at least partially identical to the form fields in a form to be filled out that belongs to the same intent. This allows the model to more clearly define the form fields to be extracted by the user during setup.

[0071] For example, when configuring a leave request intent, the purpose description can be: trigger the leave request intent when a user needs to apply for leave. Guide the user through the leave application process and extract four key pieces of information from the user's conversation: [Leave Type], [Leave Start Time], [Leave End Time], and [Leave Reason], and record them without returning any information.

[0072] With this configuration, when a user's conversation contains elements related to "taking leave," the model can correctly understand and guide the user to the intention to take leave. At the same time, it can correctly extract the field information contained in the conversation based on the key field information in the description.

[0073] In this application's embodiments, intent use cases include dialogue examples of users under different intents.

[0074] Intent use cases, based on intent judgment, provide the large model with more references about common user questioning methods. Through intent use cases, the model can have a stronger understanding of various questioning methods, enhance the model's understanding of the current intent, and even achieve generalization.

[0075] For example, when configuring leave use cases, these intent use cases can be supplemented as: "I need to take a day off because I am sick, please submit my leave application", "I need to take sick leave and go to the hospital tomorrow morning", "I have a child at home, please take paternity leave", "I am sick", etc.

[0076] In this way, when users configure usage descriptions and intent use cases, the model can guide them and provide different feedback for different use cases.

[0077] After configuring the office system, users can then directly interact with the system to have it automatically generate and submit forms. See also... Figure 1 As shown, Figure 1This is a flowchart illustrating the office method provided in this application embodiment, which includes: Step S101: Obtain the first configuration information.

[0078] Optionally, the first configuration information may include: access authentication parameters of the third-party system and construction parameters of different forms.

[0079] The construction parameters can include: form fields and the field attributes corresponding to the form fields.

[0080] The field attributes corresponding to form fields can include: field format.

[0081] Step S102: In the preset user dialogue interface, in response to the user's first input operation, obtain the first user dialogue.

[0082] For example, the user dialogue interface could be a page displayed when an application is opened, or it could be a webpage. Users can click the voice input button in the user dialogue interface to input voice, and the input voice will be used as the first user dialogue.

[0083] Step S103: Determine the form to be filled out based on the first user's dialogue.

[0084] In some embodiments, the second configuration information includes: a usage description and intent use cases. A first prompt word is used to describe the user intent determined in the user dialogue based on the usage description and intent use cases. Determining the form to be filled based on the first user dialogue may include: obtaining the second configuration information; inputting the first prompt word and the first user dialogue into a preset large language model to obtain the user intent; and determining the form to be filled based on the user intent.

[0085] In the above embodiments, inputting the first prompt word and the first user dialogue into a preset large language model to obtain user intent includes: combining the first user dialogue and the preset first prompt word to obtain a first combined text; and inputting the first combined text into the preset large language model to obtain user intent.

[0086] Specifically, a first specific character can be set in the preset first prompt word, and a first user dialogue can be filled in the first specific character, or the first user dialogue can be used to replace the first specific character, so as to achieve a combination of the first user dialogue and the preset first prompt word.

[0087] It is understandable that a specific character can be a single character or a combination of multiple characters, such as double quotes, single quotes, forward slashes, colons, percent signs, letters, etc.

[0088] In the above embodiments, determining the form to be filled based on the user's intent may include: searching for the form to be filled corresponding to the user's intent in a preset form association relationship.

[0089] Among them, the form association is used to reflect the association between the user's intent and the form to be filled out.

[0090] In the above embodiments, the office method further includes: determining the intent type corresponding to the user intent.

[0091] Correspondingly, determining the form to be filled based on the user's intent can include: if the intent type is a submission intent, searching for the form to be filled corresponding to the user's intent in a preset form association relationship.

[0092] Optionally, determining the intent type corresponding to the user intent may include: searching for the intent type corresponding to the user intent in a preset type association relationship.

[0093] The preset type association is used to reflect the relationship between user intent and intent type.

[0094] For example, intent types can be divided into two categories: query intents and submission intents.

[0095] In user conversations, phrases like "I want to check my to-do list" or "I want to find the head of department xx in my organization" are query-type scenarios that do not require submitting a request form or related parameters; these are query intents.

[0096] User conversations mentioning "I want to ask for leave" or "I'm going to Beijing on a business trip" require users to fill in necessary form information and submit it to the business system; these are submission-type intents.

[0097] Users can pre-set the intent type corresponding to each user intent to form a type association relationship, and record the type association relationship in a table format.

[0098] Step S104: Determine the field content corresponding to each form field based on the form fields to be filled, the field attributes of each form field, and the first user dialogue.

[0099] In some embodiments, determining the field content corresponding to each form field based on the form field to be filled, the field attributes corresponding to each form field, and the first user dialogue may include: combining the form field to be filled, the field attributes corresponding to each form field, and a preset prompt word template to obtain a second prompt word; and inputting the second prompt word and the first user dialogue into a preset large language model to obtain the field content corresponding to each form field.

[0100] The second prompt is used to describe extracting the field content corresponding to each form field from the dialog according to the field attributes of the form fields. If the field content corresponding to a certain form field does not exist, then the field content corresponding to that form field is empty.

[0101] In the above embodiment, the attribute information configuration items for form fields include default values. It is possible to extract only the field content of form fields whose default values ​​are empty from all form fields corresponding to the form to be filled.

[0102] In the above embodiments, a second specific character and a third specific character are set in a preset field description template. The field description template is used to prompt the extraction of form fields within the third specific character according to the requirements within the second specific character. A fourth specific character is set in a preset prompt word template. The preset prompt word template is used to describe the content described within the fourth specific character extracted from the dialogue. If the field content corresponding to a certain form field does not exist, then the field content corresponding to that form field is empty.

[0103] The second prompt word is obtained by combining the form fields and their corresponding attributes with a preset prompt word template. This can be done as follows: For each form field: enter the form field in the second specific character, enter the corresponding field attribute in the third specific character, and use the replaced field description template as the second combined text. Then, sequentially enter each of the second combined texts in the fourth specific character to obtain the second prompt word.

[0104] In other embodiments, when setting the purpose description, the user has already simultaneously set the form fields to be extracted for each type of form to be filled out. In this case, determining the field content corresponding to each form field based on the form fields, the field attributes corresponding to each form field, and the first user dialogue may include: inputting the fourth prompt word corresponding to the user's intent and the first user dialogue into a preset large language model to obtain the field content corresponding to each form field.

[0105] The fourth prompt is used to describe the extraction of the field content corresponding to each form field from the dialog according to the preset field attributes of the form fields. If the field content corresponding to a certain form field does not exist, then the field content corresponding to that form field is empty.

[0106] For example, field attributes can include field format. Field format can be text, number, date, time, etc. Field format can also be a set specification. For example, the extracted time needs to be described according to the ISO 8601 standard, that is, the date format: YYYY-MM-DD format, such as 2024-03-19 representing March 19, 2024.

[0107] If a form field has a field format, the large language model extracts the candidate content for that form field from the user's dialogue; it then converts the candidate content according to the field format to obtain the field content of that form field.

[0108] For example, field attributes may include an associated dictionary. If an associated dictionary exists for a field attribute of a form field, the large language model extracts candidate content for that form field from the user's dialogue; it then searches and matches the candidate content in the associated dictionary, and uses the most similar content as the field content of that form field.

[0109] For example, field attributes can include an associated dictionary and a field format. If a form field's field attributes have an associated dictionary and a field format, the large language model extracts candidate content for that form field from the user's dialogue; it searches and matches the candidate content in the associated dictionary; and it transforms the most similar content in the associated dictionary according to the field format to obtain the field content of that form field.

[0110] In some cases, users may omit some information that needs to be filled in during the initial interaction. Therefore, one solution is to: if a form field corresponding to a form to be filled in has empty content, the form field with empty content is considered an incomplete field. This incomplete field is then combined with a preset display template to obtain the content to be displayed. This content is then displayed to the user.

[0111] The content to be displayed is used to prompt users to fill in the missing field information.

[0112] In the above embodiments, a fifth specific character is set in a preset display template. The preset display template is used to prompt the user to supplement the content of the form field within the fifth specific character. The incomplete field is combined with the preset display template to obtain the content to be displayed, including filling the incomplete field into the fifth specific character, so as to realize the combination of the incomplete field and the display template.

[0113] After the content to be displayed is shown to the user, the user will engage in another dialogue. The working method at this time also includes: obtaining the second user dialogue; combining the incomplete fields, the field attributes corresponding to the incomplete fields, and the prompt word template to obtain the third prompt word; inputting the third prompt word and the second user dialogue into the large language model to obtain the field content corresponding to each incomplete field.

[0114] The method for obtaining the third prompt word is similar to that for the second prompt word, and will not be repeated here. If incomplete fields still exist after the user's next conversation, the above steps can be repeated, that is, prompting the user again to complete the incomplete field content, and extracting the incomplete field content from the conversation again, until all the data required for the form to be filled in has been extracted.

[0115] Step S105: Combine the contents of each field to form the target form.

[0116] In some embodiments, the construction parameters further include: form layout. Correspondingly, combining the content of each field to form the target form includes: arranging the content of each field according to the form layout to form the target form.

[0117] Step S106: Submit the target form to the third-party system according to the access authentication parameters of the third-party system.

[0118] By setting access authentication parameters for third-party systems, it's like building plugins that can easily connect to different systems to generate intelligent interaction commands, enabling information exchange between different systems.

[0119] In some embodiments, the office system also supports configuring the content of the interface return results. The obtained interface return results can be written back to the form or the execution results can be displayed to the user to help the user quickly and intuitively obtain the execution results.

[0120] For example, a user asks, "I'm not feeling well and would like to ask for leave tomorrow morning." Referring to the user's configured purpose description and intent use cases, the model understands that the user wants to request leave and enters the leave request intent. The model obtains the structured text form associated with the leave request intent and clarifies that the form requires two form fields: "Leave Time" and "Leave Type."

[0121] The model extracts the content of the "Leave Time" and "Leave Type" form fields from the user's query. After extraction, the model determines whether the user's content is complete. It finds that the user's content is incomplete; only the "Sick Leave" field for "Leave Type" can be extracted, but the "Leave Time" field is missing. The model then fills in the "Sick Leave" field according to the pre-defined form layout and displays the "Leave Time" field to the user for completion, with a message such as "You have not provided leave time, please provide it."

[0122] At this point, the user provides feedback again, such as "tomorrow morning 10:00-12:00".

[0123] The model extracts the "Leave Time" field from the user's query, finding it to be "March 4, 2022, 10:00 AM - 12:00 PM". The model then fills in the corresponding "Leave Time" field, "March 4, 2022, 10:00 AM - 12:00 PM", into the designated location according to the pre-defined form layout. Finally, the complete target form is submitted to a pre-configured third-party system.

[0124] Combination Figure 2 As shown in the figure, this application embodiment provides an office device, including: a configuration acquisition module 1, a user input module 2, an intent module 3, a text processing module 4, and a submission module 5.

[0125] The system comprises: a configuration acquisition module 1 for acquiring first configuration information; a user input module for acquiring first user dialogue in response to the user's first input operation at a preset user dialogue interface; an intent module for determining the form to be filled based on the first user dialogue; a text processing module for determining the field content corresponding to each form field based on the form fields to be filled, the field attributes of each form field, and the first user dialogue; combining the field content to form a target form; and a submission module for submitting the target form to the third-party system according to the access authentication parameters of the third-party system.

[0126] In some embodiments, the intent module 3 is used to determine the form to be filled based on a first user dialogue in the following manner: obtaining second configuration information; the second configuration information includes: a purpose description and intent use cases; wherein, in the user configuration interface, the second configuration information is obtained in response to the user's second configuration operation; the purpose description describes what intents are triggered in different scenarios; the intent use cases include examples of user dialogues under different intents; inputting a first prompt word and the first user dialogue into a preset large language model to obtain the user intent; the first prompt word is used to describe the user intent in the user dialogue determined by referring to the purpose description and intent use cases; and determining the form to be filled based on the user intent.

[0127] In some embodiments, the intent module 3 is used to determine the form to be filled based on the user intent in the following ways: determining the intent type corresponding to the user intent; if the intent type is a submission intent, searching for the form to be filled corresponding to the user intent in a preset form association relationship; the form association relationship is used to reflect the association relationship between the user intent and the form to be filled.

[0128] In some embodiments, the text processing module 4 is used to determine the field content corresponding to each form field based on the form field to be filled, the field attributes corresponding to each form field, and the first user dialogue in the following manner: combining the form field to be filled, the field attributes corresponding to the form field, and a preset prompt word template to obtain a second prompt word; the second prompt word is used to describe the field content corresponding to each form field extracted from the dialogue according to the field attributes of the form field, and if the field content corresponding to a certain form field does not exist, then the field content corresponding to that form field is empty; inputting the second prompt word and the first user dialogue into a preset large language model to obtain the field content corresponding to each form field.

[0129] In some embodiments, the text processing module 4 is further configured to: if the field content of a form field corresponding to a form to be filled is empty, record the form field with empty field content as an incomplete field; combine the incomplete field with a preset display template to obtain the content to be displayed; use the content to be displayed to prompt the user to fill in the field content of the incomplete field; and display the content to be displayed to the user.

[0130] In some embodiments, the text processing module 4 is further configured to: acquire a second user dialogue; combine incomplete fields, the field attributes corresponding to the incomplete fields, and a prompt word template to obtain a third prompt word; the third prompt word is used to describe the field content corresponding to each incomplete field extracted from the dialogue according to the field attributes of the incomplete fields, and if the field content corresponding to a certain incomplete field does not exist, then the field content corresponding to that incomplete field is empty; input the third prompt word and the second user dialogue into a large language model to obtain the field content corresponding to each incomplete field.

[0131] In some embodiments, the construction parameters further include: form layout; the text processing module 4 is used to combine the content of each field to form a target form in the following way: arranging the content of each field according to the form layout to form a target form.

[0132] Combination Figure 3 As shown, this application embodiment provides an electronic device, including a processor 6 and a memory 7. Optionally, the device may further include a communication interface 8 and a bus 9. The processor 6, communication interface 8, and memory 7 can communicate with each other via the bus 9. The communication interface 8 can be used for information transmission. The processor 6 can call logical instructions in the memory 7 to execute the office methods described in the above embodiment.

[0133] Furthermore, the logical instructions in the aforementioned memory 7 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium.

[0134] The memory 7, as a computer-readable storage medium, can be used to store software programs and computer-executable programs, such as program instructions / modules corresponding to the methods in the embodiments of this application. The processor 6 executes functional applications and data processing by running the program instructions / modules stored in the memory 7, thereby implementing the office methods in the above embodiments.

[0135] The memory 7 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the terminal device. Furthermore, the memory 7 may include high-speed random access memory and may also include non-volatile memory.

[0136] Among them, electronic devices can be computers or servers, etc.

[0137] This application provides a storage medium storing computer-executable instructions configured to perform the above-described office methods.

[0138] This application provides a computer program product, which includes a computer program stored on a storage medium. The computer program includes program instructions, which, when executed by a computer, cause the computer to perform the above-described office methods.

[0139] The aforementioned computer-readable storage medium may be a transient computer-readable storage medium or a non-transitory computer-readable storage medium.

[0140] The technical solutions of this application embodiment can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes one or more instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method of this application embodiment. The aforementioned storage medium can be a non-transitory storage medium, including various media capable of storing program code such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks, or it can be a transient storage medium.

[0141] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0142] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application. Furthermore, the above embodiments can be combined with each other to form new embodiments without conflict.

Claims

1. An office method, characterized in that, include: Obtain the first configuration information; The first configuration information includes: access authentication parameters of the third-party system and construction parameters of different forms; the construction parameters include: form fields and field attributes corresponding to the form fields; wherein, in the preset user configuration interface, the first configuration information is obtained in response to the user's first configuration operation; In the preset user dialogue interface, respond to the user's first input operation and obtain the first user dialogue; The form to be filled out is determined based on the first user's dialogue; The field content corresponding to each form field is determined based on the form fields corresponding to the form to be filled, the field attributes corresponding to each form field, and the first user dialogue. Combine the contents of each field to form the target form; The target form is submitted to the third-party system according to the access authentication parameters of the third-party system.

2. The method according to claim 1, characterized in that, The form to be filled out is determined based on the first user dialogue, including: Obtain second configuration information; the second configuration information includes: a usage description and intent use cases; wherein, in the user configuration interface, the second configuration information is obtained in response to the user's second configuration operation; the usage description describes what intents are triggered in different scenarios; the intent use cases include dialogue examples of the user under different intents; The first prompt word and the first user dialogue are input into a preset large language model to obtain the user intent; the first prompt word is used to describe the user intent in the user dialogue determined with reference to the usage description and the intent use case. The form to be filled out is determined based on the user's intent.

3. The method according to claim 2, characterized in that, Determining the form to be filled out based on the user's intent includes: Determine the intent type corresponding to the user intent; If the intent type is a submission intent, the form to be filled corresponding to the user intent is searched in the preset form association relationship; the form association relationship is used to reflect the relationship between the user intent and the form to be filled.

4. The method according to claim 1, characterized in that, Based on the form fields corresponding to the form to be filled, the field attributes corresponding to each form field, and the first user dialogue, the field content corresponding to each form field is determined, including: The form fields corresponding to the form to be filled in, the field attributes corresponding to the form fields, and the preset prompt word template are combined to obtain the second prompt word; the second prompt word is used to describe the field content corresponding to each form field extracted from the dialog according to the field attributes of the form fields. If the field content corresponding to a certain form field does not exist, the field content corresponding to that form field is empty. Input the second prompt word and the first user dialogue into a preset large language model to obtain the field content corresponding to each of the form fields.

5. The method according to claim 4, characterized in that, The method further includes: If any form field corresponding to the form to be filled out has empty content, the form field with empty content is recorded as an incomplete field; The incomplete fields are combined with a preset display template to obtain the content to be displayed; the content to be displayed is used to prompt the user to complete the field content of the incomplete fields. The content to be displayed is shown to the user.

6. The method according to claim 5, characterized in that, The method further includes: Obtain the second user's conversation; The incomplete field, the field attribute corresponding to the incomplete field, and the prompt word template are combined to obtain the third prompt word; the third prompt word is used to describe the field content corresponding to each incomplete field extracted from the dialogue according to the field attribute of the incomplete field. If the field content corresponding to a certain incomplete field does not exist, the field content corresponding to that incomplete field is empty. The third prompt word and the second user dialogue are input into the large language model to obtain the field content corresponding to each incomplete field.

7. The method according to claim 1, characterized in that, The construction parameters also include: form layout; combining the contents of each field to form the target form, including: Arrange the contents of each field according to the form layout to form the target form.

8. An office device, characterized in that, include: The configuration acquisition module is used to acquire the initial configuration information; The first configuration information includes: access authentication parameters of the third-party system and construction parameters of different types of forms; the construction parameters include: form fields and field attributes corresponding to the form fields; wherein, in the preset user configuration interface, the first configuration information is obtained in response to the user's first configuration operation; The user input module is used to respond to the user's first input operation in a preset user dialogue interface and obtain the first user dialogue. The intent module is used to determine the form to be filled out based on the first user dialogue; The text processing module is used to determine the field content corresponding to each form field based on the form fields to be filled, the field attributes corresponding to each form field, and the first user dialogue; and to combine the field content to form a target form. The submission module is used to submit the target form to the third-party system according to the access authentication parameters of the third-party system.

9. An electronic device, characterized in that, It includes a processor and a memory, the memory storing computer-executable instructions that can be executed by the processor, the processor executing the computer-executable instructions to implement the office method according to any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement the office method as described in any one of claims 1 to 7.

Images