A task processing method, device, electronic device, and storage medium

Abstract

The present application discloses a task processing method, apparatus, electronic device, and storage medium. The method applies speech recognition and natural language processing technologies. The method can obtain a task instruction corresponding to a task to be processed, perform intent recognition and context analysis on the task instruction, determine task execution content and target context key information, and based on the target context key information, obtain a target execution object corresponding to the task execution content from preset mapping information, and based on the target execution object, execute the task to be processed. By means of the correspondence between the context key information and the execution object, the method performs standardized conversion to obtain the preset mapping information, which can improve the comprehensiveness and accuracy of data integration for the preset mapping information, and is easy to expand, which can increase the richness of the work scenario. The method determines the target execution object based on the context key information and executes the task to be processed, which can improve the efficiency and accuracy of task processing.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular, to a task processing method, apparatus, electronic device, and storage medium. Background Art

[0002] Enterprise management can manage tasks to be executed and employees in an enterprise. In the scenario of enterprise management, since a lot of information is generated during business interactions in different scenarios, in the prior art, the accumulation and integration of this information are mostly processed manually, resulting in chaotic information management, with many errors and omissions. Therefore, when actually processing tasks and applying this information, the efficiency and accuracy of task processing are reduced. Summary of the Invention

[0003] This application provides a task processing method, apparatus, electronic device, and storage medium, achieving the technical effect of improving the efficiency and accuracy of task processing.

[0004] On the one hand, this application provides a task processing method, which includes:

[0005] Obtain a task instruction corresponding to a task to be processed;

[0006] Perform semantic analysis on the task instruction to determine the task execution content and key information of the target context;

[0007] Based on the key information of the target context, obtain a target execution object corresponding to the task execution content from preset mapping information, where the preset mapping information represents the mapping relationship between multiple context key information and at least one execution object corresponding to each;

[0008] Execute the task to be processed based on the target execution object.

[0009] On the other hand, a task processing apparatus is provided, which includes: a task instruction acquisition module, a task instruction analysis module, a target task search module, and a target task execution module;

[0010] The task instruction acquisition module is used to obtain a task instruction corresponding to a task to be processed;

[0011] The task instruction analysis module is used to perform semantic analysis on the task instruction to determine the task execution content and key information of the target context;

[0012] The target task search module is used to obtain a target execution object corresponding to the task execution content from preset mapping information based on the key information of the target context, where the preset mapping information represents the mapping relationship between multiple context key information and at least one execution object corresponding to each;

[0013] The target task execution module is configured to execute the to-be-processed task based on the target execution object.

[0014] On the other hand, an electronic device is provided, which includes a processor and a memory. At least one instruction or at least one program segment is stored in the memory, and the at least one instruction or the at least one program segment is loaded and executed by the processor to implement the above-mentioned task processing method.

[0015] On the other hand, a computer-readable storage medium is provided, which includes a processor and a memory. At least one instruction or at least one program segment is stored in the memory, and the at least one instruction or the at least one program segment is loaded and executed by the processor to implement a task processing method as described above.

[0016] A task processing method, apparatus, electronic device, and storage medium provided by the present application. The method can obtain a task instruction corresponding to a to-be-processed task, perform intent recognition and context analysis on the task instruction, determine task execution content and target context key information, obtain a target execution object corresponding to the task execution content from preset mapping information based on the target context key information, and execute the to-be-processed task based on the target execution object. This method determines the target execution object based on the context key information and executes the to-be-processed task, which can improve the efficiency and accuracy of task processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 It is a schematic diagram of an application scenario of a task processing method provided by an embodiment of the present application;

[0019] Figure 2 It is a flowchart of a task processing method provided by an embodiment of the present application;

[0020] Figure 3 It is a flowchart of a method for semantic analysis in a task processing method provided by an embodiment of the present application;

[0021] Figure 4 It is a flowchart of a method for determining a target execution object in a task processing method provided by an embodiment of the present application;

[0022] Figure 5Schematic diagram of task processing when the task execution content is query content in a task processing method provided by an embodiment of the present application;

[0023] Figure 6 Schematic diagram of task processing when the task execution content is action content in a task processing method provided by an embodiment of the present application;

[0024] Figure 7 Flowchart of a method for obtaining preset mapping information in a task processing method provided by an embodiment of the present application;

[0025] Figure 8 Schematic diagram of forming preset mapping information based on behavior association information and organizational relationship information in a task processing method provided by an embodiment of the present application;

[0026] Figure 9 Schematic diagram of the application scenario of an enterprise knowledge system in a task processing method provided by an embodiment of the present application;

[0027] Figure 10 Schematic diagram of the structure of a task processing device provided by an embodiment of the present application;

[0028] Figure 11 Schematic diagram of the hardware structure of a device for implementing the method provided by an embodiment of the present application. Detailed implementation manners

[0029] To make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0030] In the description of the present application, it should be understood that the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of such features. Moreover, the terms "first", "second", etc. are applicable to distinguish similar objects and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so that the embodiments of the present application described herein can be implemented in an order different from those illustrated or described herein.

[0031] Please refer to Figure 1, which shows a schematic diagram of an application scenario of a task processing method provided in an embodiment of the present application. The application scenario includes a client 110 and a server 120. The client 110 converts the information input by the user into a task instruction corresponding to the task to be processed in response to the information input by the user. The client 110 sends the task instruction to the server. After receiving the task instruction, the server 120 performs semantic analysis on the task instruction to determine the task execution content and the target context key information. Based on the target context key information, the server 120 obtains the target execution object corresponding to the task execution content from the preset mapping information, and based on the target execution object, executes the task to be processed and feeds back the execution result to the client 110.

[0032] In an embodiment of the present application, the client 110 may include physical devices such as smart phones, desktop computers, tablet computers, laptop computers, digital assistants, and smart wearable devices, or may also include software running on the physical devices, such as application programs. The operating systems running on the physical devices in the embodiments of the present application may include, but are not limited to, Android system, IOS system, linux, Unix, windows, etc. The client 110 includes a UI (User Interface) layer. The client 110 provides information input and display of execution results externally through the UI layer. In addition, the task instruction of the task to be processed is sent to the server 120 based on the API (Application Programming Interface).

[0033] In an embodiment of the present application, the server 120 may include an independently operating server, or a distributed server, or a server cluster composed of multiple servers. The server 120 may include a network communication unit, a processor, a memory, and so on. Specifically, the server 120 may receive the task instruction sent by the client 110, perform semantic analysis on the task instruction, and determine the target execution object according to the result of the semantic analysis, and execute the task to be processed.

[0034] In an embodiment of the present application, when the user input information received by the server is voice information, the voice information is converted into a task instruction based on voice recognition technology. Automatic Speech Recognition (ASR) enables the computer to listen, see, and feel, and is the future development direction of human-computer interaction. Among them, voice has become one of the most promising human-computer interaction methods in the future.

[0035] In the embodiments of the present application, when the server performs semantic analysis on the task instruction, natural language processing technology can be adopted. Natural Language Processing (NLP) is an important direction in the field of computer science and artificial intelligence. It studies various theories and methods that can achieve effective communication between humans and computers in natural language. Natural language processing is a science that integrates linguistics, computer science, and mathematics. Therefore, the research in this field will involve natural language, that is, the language people use in daily life, so it has a close connection with the research of linguistics. Natural language processing technology usually includes technologies such as text processing, semantic understanding, machine translation, robot question answering, and knowledge graph.

[0036] First, the following explanations are made for the relevant terms involved in the embodiments of the present application:

[0037] Context key information: Information obtained after standardizing and converting enterprise knowledge data. Enterprise knowledge data includes conversations, conclusions, context-aware Q&A, audio and video conference content, templates, materials, design documents, code, work records, assigned instructions, organizational information, organizational structure, third-party applications, etc. precipitated in the collaborative work process of enterprise employees.

[0038] Knowledge base: A knowledge base is a system with knowledge and intelligence. The knowledge base can be used to store context key information and the preset mapping relationship between execution objects. At the same time, the knowledge base can also store organizational relationship information and use the organizational relationship information to update the preset mapping relationship. Based on the stored information, the knowledge base can perform different functions such as query, notification, and group pulling.

[0039] Standardized conversion: Based on the preset standardization categories, enterprise knowledge data is respectively corresponded to different standardization categories for differentiation to obtain the context key information and execution objects after standardized conversion. The standardization categories can include timestamp information, intention information, sorting information, constraint information, media information, and object information. When receiving a task instruction sent by a user, the text feature information corresponding to the task instruction can also be standardized and converted to obtain the target context key information and target execution object corresponding to the task instruction.

[0040] Context analysis: Analyze the information related to the occurrence scenario corresponding to the enterprise knowledge data, from which context key information can be extracted. The occurrence scenario can be information such as the context in the conversation, the theme of the meeting, the meeting time, and the participants.

[0041] Please refer to Figure 2 , which shows a task processing method that can be applied to the server side. The method includes:

[0042] S210. Obtain the task instruction corresponding to the task to be processed;

[0043] Further, in response to the information input by the user, the client can convert the information input by the user and generate a task instruction corresponding to the task to be processed. The information input by the user can be text information or voice information. When the information input by the user is text information, the client can directly convert the text information into a task instruction corresponding to the task to be processed. When the information input by the user is voice information, the client performs automatic speech dialogue recognition on the voice information to obtain a speech recognition result, and converts the speech recognition result into a task instruction corresponding to the task to be processed. The client sends the task instruction to the server, and the server can receive the task instruction.

[0044] S220. Perform semantic analysis on the task instruction to determine the task execution content and the key information of the target context;

[0045] In some embodiments, the task execution content represents the execution information related to the task to be processed in the task instruction. The task execution content may include intention information and object information. The intention information may include action content and query content. The key information of the target context represents the context key information in the task instruction. The key information of the target context may include timestamp information, sorting information, constraint information, media information, etc. The timestamp information represents the occurrence time corresponding to the target execution object in the task instruction. For example, the file shown in the meeting held in x year x month. The sorting information represents the information of the time or process related to sorting in the task instruction. The constraint information represents the constraint information related to the context, such as quantity, object, type, etc. The media information represents the type information of the target execution object in the task instruction.

[0046] In some embodiments, please refer to Figure 3 , performing semantic analysis on the task instruction to determine the task execution content and the key information of the target context includes:

[0047] S310. Obtain the text feature information corresponding to the task instruction;

[0048] S320. Perform intention recognition on the text feature information to obtain the task execution content;

[0049] S330. Perform context analysis on the text feature information to determine the key information of the target context.

[0050] In some embodiments, when performing semantic analysis on a task instruction, the task instruction is segmented, and at least one segmentation information is obtained. Based on at least one segmentation information, the text feature information corresponding to the task instruction is determined. When determining the text feature information, the method of Term Frequency-Inverse Document Frequency (TF-IDF) can be used to analyze the segmentation information, obtaining the frequency of each segmentation information in the task instruction and its importance in all task instructions. Based on the frequency of the segmentation information in the task instruction and its importance in all task instructions, the text feature information is calculated. Among them, the importance of the segmentation information in all task instructions can be calculated based on the total number of task instructions and the number of task instructions including the segmentation information. When determining the text feature information, it is also possible to use a feature extraction model of word vectors to map each segmentation information into a word vector, identify the word vectors, and obtain the text feature information based on the similarity between the word vectors.

[0051] Perform intent recognition on the text feature information to identify the intent feature information and object feature information in the text feature information. When performing the recognition, the text feature information can be compared with the preset intent feature information and preset object feature information, so as to obtain the intent feature information and object feature information. Based on the intent feature information and object feature information, the task execution content can be obtained.

[0052] Perform context analysis on the text feature information to identify the time feature information, constraint feature information, or media feature information in the text feature information. When performing the recognition, the text feature information can be compared with the preset time feature information, preset constraint feature information, and preset media feature information, so as to obtain the timestamp information, sorting feature information, constraint feature information, or media feature information in the text feature information. Based on the timestamp information, sorting feature information, constraint feature information, or media feature information, the context key information can be determined. The context key information can include one or more of the timestamp information, sorting information, constraint information, or media information.

[0053] When performing intent recognition and context analysis on the text feature information, the text feature information can be subjected to standardized conversion, and content extraction is performed corresponding to different standardized categories from the text feature information according to the preset standardized categories. The preset standardized categories can include six standardized categories, namely timestamp information, intent information, sorting information, constraint information, media information, and object information. The timestamp information represents the occurrence time of the execution object. For example, the documents shown in the meeting minutes occurring in x year x month, the process information mentioned in the text conversation occurring in x year x month, etc.

[0054] The intent information represents the intent of querying or other actions. Other actions include actions such as sharing, forwarding, group pulling, and business execution. Other actions rely on third-party applications or software and hardware carriers. For example, when the other action is sharing, it can be shared through the sharing function in the software. When the other action is notification, it can be notified through office software or a client. When the other action is group pulling, it can be group pulled through the social function in the software. When the other action is business execution, a third-party interface can be called, such as reporting temperature, booking a restaurant, or taking a leave.

[0055] The sorting information represents information with sorting factors such as time or process, such as "the most recent time", "the next step", etc. The constraint information represents information with constraint factors such as quantity, type, and constraint information on the target execution object. For example, "the nearest to me" in "book the meeting room nearest to me" is the constraint information for "meeting room", "General Manager Wang" and "today" in "check when General Manager Wang is available today" are constraint information, "three" in "three suppliers" is the constraint information for "supplier", and "reimbursement" in "reimbursement process" is the constraint information for the process.

[0056] The media information represents the type of the target execution object, which can be different types such as text, sound, picture, document, compressed package, etc.

[0057] The object information can represent the target execution object, involving sharing object, query object, notification object, etc. For example, "General Manager" in "notify the General Manager" is object information, "electric light" in "turn on the electric light in the meeting room" is object information, and "workstation" in "find the workstation of someone" is object information.

[0058] Perform context analysis on the text feature information, determine the context key information, and can perform standardized conversion on the context key information. The task instruction corresponding target execution object can be obtained in combination with the context, thereby improving the efficiency and accuracy of task processing.

[0059] S230. Based on the target context key information, obtain the target execution object corresponding to the task execution content from the preset mapping information. The preset mapping information represents the mapping relationship between multiple context key information and at least one execution object corresponding to each of them;

[0060] In some embodiments, the preset mapping information can be the mapping information between the execution object and the context key information stored in the knowledge base. Under each context key information, one or more execution objects can be corresponding. For example, when the context key information is the xx meeting minutes of x year x month, File A shown in the meeting can be the execution object, File B shown in the meeting can be the execution object, the knowledge information in the conversation among the participants can be the execution object, and the knowledge information in the annotation information in the meeting can be the execution object.

[0061] In some embodiments, refer to Figure 4 , based on the key information of the target context, obtaining the target execution object corresponding to the task execution content from the preset mapping information includes:

[0062] S410. Extract the object identifier corresponding to the target execution object from the task execution content;

[0063] S420. Based on the preset mapping information, determine at least one execution object that matches the key information of the target context;

[0064] S430. Based on the object identifier, determine the target execution object from at least one matching execution object.

[0065] In some embodiments, according to the object information in the task execution content, the object identifier corresponding to the target execution object can be extracted, and the object identifier can be the name information of the target execution object. According to the mapping information between the context key information and the execution object, the context key information that matches the key information of the target context can be determined, and at least one execution object corresponding to the matching context key information can be obtained, that is, at least one execution object that matches the key information of the target context. Among the at least one matching execution object, the target execution object that matches the object identifier is obtained.

[0066] In some embodiments, if the object information in the task execution content is File A, the target execution object can be determined as the file named "A". If the key information of the target context is the xx meeting minutes of x year x month, according to the timestamp information "x year x month" and the constraint information "xx meeting minutes", the matching context key information can be determined from the preset mapping information, so as to obtain one or more execution objects corresponding to the xx meeting minutes of x year x month. Among these execution objects, the file named "A" is obtained, and the target execution object can be obtained.

[0067] Obtaining the target execution object in combination with the context information can improve the accuracy of identifying task instructions.

[0068] S240. Execute the task to be processed based on the target execution object.

[0069] In some embodiments, after obtaining the target execution object, performing the operation corresponding to the intention information in the task execution content on the target execution object can execute the task to be processed. Among them, there can be multiple operations corresponding to the intention information.

[0070] In some embodiments, the method further includes:

[0071] When the task execution content is a query content and the target execution object includes multiple query results, sort the multiple query results based on a preset priority to obtain a query result sequence;

[0072] Execute the task to be processed based on the query result sequence.

[0073] In some embodiments, when the task execution content is a query content, the target execution object may include one or more query results. When the target execution object includes multiple query results, the multiple query results can be sorted according to a preset priority to obtain a query result sequence, and then the task to be processed is executed, and the query result sequence is sent to the client. The preset priority may be the degree of relevance between the query result and the task to be processed. The query result with a higher degree of relevance to the task to be processed has a higher priority, and the query result with a lower degree of relevance to the task to be processed has a lower priority. The preset priority may be the degree of relevance between the context key information corresponding to the query result and the user of the client. The query result with a higher degree of relevance between the context key information and the user of the client has a higher priority, and the query result with a lower degree of relevance between the context key information and the user of the client has a lower priority. After the task to be processed is completed, the mapping relationship between the target context key information corresponding to the task instruction and the target execution object can be determined, and the preset mapping relationship can be updated according to the mapping relationship.

[0074] For example, please refer to Figure 5 , such as Figure 5The figure shows a schematic diagram of the scenario for obtaining query results according to the query content. The user inputs voice information, and voice recognition is performed at the recognition layer of the server to obtain a task instruction. Semantic analysis is performed on the task instruction to obtain the task execution content and the key information of the target context. At the standardization layer of the server, standardization conversion is performed on the task execution content and the key information of the target context. According to the preset standardization categories, the content corresponding to each standardization category is extracted from the task execution content and the key information of the target context, and the task execution content and the key information of the target context are converted into one or more of intention information, sorting information, constraint information, media information, and object information. When the intention information is the query content, based on one or more context-related information such as sorting information, constraint information, and media information included in the key information of the target context, the query result corresponding to the identification information of the target execution object is determined from the preset mapping relationship stored in the server storage layer. In the case where there are multiple query results, the multiple query results can be sorted according to the preset priority to obtain a query result sequence, and the task to be processed is executed, and the query result sequence is sent to the client. The preset mapping relationship stored in the server storage layer is that the server integrates enterprise knowledge such as conversations, voices, meetings, third-party application operations, enterprise internal employee information, enterprise information, and organizational relationship information, extracts the key information of the context and the execution object from the enterprise knowledge information, generates a preset mapping relationship between the key information of the context and the execution object for storage, and the preset mapping relationship includes the knowledge points obtained by context concatenation, and the enterprise knowledge information can be precipitated.

[0075] For example, as Figure 5 shown, when the task instruction of the task to be processed is "How does the procurement process proceed? What are the requirements and restrictions for the quotation sheets of three suppliers?", the key information of the target context is the constraint information "procurement", the intention information in the task execution content is "query", and the object information in the task execution content is "process". Based on the key information of the target context and the task execution content, the query result is obtained and sent to the client. The query result can be "For this part, you need to first write the procurement requirements, put them into the tender document, obtain quotations from at least three suppliers, and then conduct an open tender."

[0076] When there are multiple procurement processes in the query results, the multiple query results of the procurement processes can be sorted according to the user's historical inquiry records of the procurement process, the conversation records of the users related to procurement, the procurement-related information in the third-party application, and the conversation records of other employees discussing the procurement topic to obtain a sequence of procurement processes.

[0077] When multiple query results are obtained, the query results can be sorted according to the preset priority, thereby improving the accuracy of the query results.

[0078] In some embodiments, the method further includes:

[0079] When the task execution content is action content and the action content corresponds to multiple actions, obtain the execution order of the multiple actions corresponding to the action content.

[0080] Execute the task to be processed based on the execution order and the target execution object.

[0081] In some embodiments, when the task execution content is action content, the action content can correspond to one or more actions. When the action content corresponds to multiple actions, the execution order of the multiple actions corresponding to the action content can be obtained, and the operations corresponding to the action content in the task execution content are performed on the target execution object according to the execution order to execute the task to be processed. The action content can include different actions such as sharing, pulling a group, forwarding, etc. The target execution object is the recipient of the action corresponding to the action content, and the target execution object can also include multiple objects. When the action content corresponds to multiple actions, each action can correspond to different objects.

[0082] For example, please refer to Figure 6 , such as Figure 6 FIG. shows a schematic diagram of a scenario for executing a task to be processed according to action content. The user inputs voice information, performs voice recognition at the recognition layer of the server to obtain a task instruction, performs semantic analysis on the task instruction to obtain task execution content and target context key information. At the standardization layer of the server, the task execution content and target context key information are subjected to standardization conversion, and according to the preset standardization categories, the content corresponding to each standardization category is extracted from the task execution content and target context key information, and the task execution content and target context key information are converted into one or more of intention information, sorting information, constraint information, media information, and object information. Based on one or more context-related information such as sorting information, constraint information, and media information included in the target context key information, the target execution object corresponding to the identification information of the target execution object is determined from the preset mapping relationship stored in the server storage layer. When the intention information is action content, obtain the action order of the multiple actions corresponding to the action content, and execute the task to be processed according to the action order and the target execution object. The preset mapping relationship stored in the server storage layer is that the server integrates enterprise knowledge such as conversations, voices, meetings, third-party application operations, enterprise internal employee information, enterprise information, and organizational relationship information, extracts context key information and execution objects from the enterprise knowledge information, generates a preset mapping relationship between the context key information and the execution objects for storage, and the preset mapping relationship includes knowledge points obtained by context concatenation, which can precipitate the enterprise knowledge information.

[0083] For example, please refer to Figure 6, when the task instruction of the task to be processed is "create a group for the department managers, notify them via text message, and forward the file I sent to Mr. Wang last time to the group", the corresponding actions of the action content include action 1 "create a group", action 2 "notify", and action 3 "forward", the object information includes object 1 "group", object 2 "file", object 3 "text message", the constraint information includes constraint 1 "manager" and constraint 2 "Mr. Wang", and the sorting information includes sorting 1 "last time". Among them, object 1 "group" corresponding to constraint 1 "manager" is related to action 1 "create a group" and action 2 "notify", object 2 "file" corresponding to constraint 2 "Mr. Wang" corresponds to action 3 "forward". According to the execution order of action 1, action 2, and action 3, based on constraint 1 "manager", object 1 "group" is created and object 3 "text message" is sent in object 1 "group" for notification, and then based on constraint 2 "Mr. Wang" and sorting 1 "last time", object 2 "file" is determined, and object 2 "file" is sent to object 1 "group".

[0084] When executing multiple actions, executing based on the action order can improve the orderliness of executing the task to be processed, thereby improving the accuracy of the execution result.

[0085] In some embodiments, please refer to Figure 7 , before obtaining the task instruction corresponding to the task to be processed, the method further includes:

[0086] S710. Obtain the behavior association information corresponding to the target user;

[0087] S720. Perform context analysis on the behavior association information to obtain the corresponding context key information and execution objects;

[0088] S730. Establish a mapping relationship between the corresponding context key information and execution objects to obtain the preset mapping information.

[0089] In some embodiments, please refer to Figure 8 , such as Figure 8 shown is the preset mapping information composed of behavior association information and organizational relationship information. The behavior association information corresponding to the target user represents the knowledge information associated with the interaction behavior of the target user. The behavior association information includes information such as text conversations, links or attachments in the context, voice conversations, and meeting records. After performing speech recognition on the voice conversations and meeting records, they are converted into text information, and semantic analysis and context analysis are performed on the converted text information, text conversations, links or attachments in the context to obtain the context key information and execution objects. As Figure 8 shown, in Figure 8The standardized layer shown can perform standardized conversion on the context key information and the execution object according to the preset standardized categories. The preset standardized categories can include six categories, namely timestamp information, intent information, sorting information, constraint information, media information, and object information. Extract the content corresponding to each standardized category from the context key information and the execution object according to the preset standardized categories. Among them, after performing standardized conversion on the context key information, one or more of the timestamp information, intent information, sorting information, constraint information, and media information can be obtained, and after performing standardized conversion on the execution object, object information is obtained. Based on the mapping relationship between the context key information and the execution object after standardized conversion, the preset mapping information is obtained.

[0090] By performing standardized conversion on the context key information and the execution object, it is easy to expand and reuse the preset mapping information, thereby increasing the richness of the working scenarios of the preset mapping information.

[0091] The text conversation is a text conversation with context information, such as Figure 8 For the shown procurement process "You need to write the procurement requirements first, put them into the tender document, obtain quotations from at least three suppliers, and then conduct an open tender", when performing context analysis on the text conversation, the object of the text conversation, the content of the text conversation, the time of the text conversation, etc. can be obtained. In the standardized layer, the time when the text conversation occurs is stored as timestamp information, the object of the text conversation is stored as object information, the content of the text object "procurement process" is stored as constraint information, and the object information and constraint information are used as the execution object and the context key information. Establish the mapping relationship between the execution object of the text conversation and the context key information in the text conversation to obtain the preset mapping information related to the text conversation.

[0092] The link or attachment in the context can be the link and attachment in the text conversation, such as Figure 8 For the shown procurement template file transmitted as an attachment, in addition to collecting data on the content of the text conversation, one or more links or one or more attachments attached during the text conversation can be obtained, and one or more links or one or more attachments are used as the execution object. In the standardized layer, obtain the text conversation corresponding to the link or attachment, perform context analysis and standardized conversion on the text conversation, classify the content in the text conversation according to the preset standardized categories, and store the attribute of the link or attachment as media information. For example, when the attachment is an image, the attachment type as image is stored as media information. Establish the mapping relationship between the execution object of the link and attachment in the context and the context key information in the corresponding text conversation to obtain the preset mapping information related to the link or attachment in the context.

[0093] The voice conversation is a voice conversation with context information, such asFigure 8 For the voice information shown, after converting the voice conversation into text information through automatic speech recognition technology, performing context analysis on the converted text information can also obtain information such as the object of the voice conversation, the content of the voice conversation, and the time of the voice conversation. At the standardization layer, the time when the voice conversation occurs is stored as timestamp information, the object of the voice conversation is stored as object information, and the content of the voice object is stored as constraint information, sorting information, or intent information, obtaining the execution object and context key information. Establish a mapping relationship between the execution object in the voice conversation and the context key information in the voice conversation to obtain the preset mapping information related to the voice conversation.

[0094] The meeting record can be a voice meeting record or a video meeting record. For example Figure 8 For the meeting record shown, performing speech recognition on the audio information of the voice meeting record or the audio information of the video meeting record to obtain the meeting record text information, and performing context analysis on the meeting record text information can also obtain information such as the object of the meeting, the content of the meeting, and the time of the meeting. At the standardization layer, the time when the meeting occurs is stored as timestamp information, the object of the meeting is stored as object information, and the content of the meeting is stored as constraint information, sorting information, or intent information, obtaining the execution object and context key information. Establish a mapping relationship between the execution object in the meeting record and the context key information in the meeting record to obtain the preset mapping information related to the meeting record.

[0095] The server obtains the preset mapping information related to text conversations, the preset mapping information related to links or attachments in the context, the preset mapping information related to voice conversations, and the preset mapping information related to meeting records from various enterprise knowledge information, and stores the preset mapping information related to text conversations, the preset mapping information related to links or attachments in the context, the preset mapping information related to voice conversations, and the preset mapping information related to meeting records, which can precipitate the enterprise knowledge information.

[0096] When obtaining the preset mapping information, the context key information and the execution object can be obtained from the behavior association information, improving the comprehensiveness and accuracy of data integration for the preset mapping information.

[0097] In some embodiments, the method further includes:

[0098] Obtaining the organizational relationship information corresponding to the target user;

[0099] Updating the preset mapping information based on the organizational relationship information.

[0100] In some embodiments, please refer to Figure 8 , such as Figure 8The figure shows a schematic diagram of constructing preset mapping information based on behavior association information and organizational relationship information. The organizational relationship information can be stored as resource data, and based on the resource data, the preset mapping information can be updated. The resource data can also include system data that can be used by employees in an enterprise, such as internal search engines, internal document libraries, internal design material libraries, internal encyclopedias, training videos, meeting materials, internal emails, etc. Among them, the resource data can be knowledge data in a knowledge base. As Figure 8 shown, in the standardization layer, according to the preset standardization categories, the organizational relationship information is converted into object information for storage.

[0101] The organizational relationship information is used to represent the role information and relationship network information of the target user. The organizational relationship information can include address books, organizational structures, organizational information, personal information, reporting relationships, third-party application permissions, etc. The personal information can include basic information, positions, titles, mobile phone numbers, etc. The organizational relationship information can be used as the target execution object or the context key information. For example, when "sending a text message to Zhang San", the mobile phone number of Zhang San can be obtained from the organizational relationship information. At this time, the mobile phone number of Zhang San is the target execution object. When "querying the process information uploaded by the departing employees", the departing employees can be obtained from the organizational relationship information, and then the people who have uploaded the process information among the departing employees can be queried to obtain the process information. At this time, the departing employees obtained from the organizational relationship information are the context key information.

[0102] The third-party application permissions represent the user permissions of internal systems or external systems, and can include meeting room reservation systems, email, parking reservation systems, training systems, encyclopedias, etc. If, when the client sends a task instruction to the server and the task instruction involves operations on internal systems or external systems based on third-party application permissions, and the server determines that the user corresponding to the client has no permissions, the task instruction can be not executed and the user can be reminded.

[0103] When updating the preset mapping information, the context key information and the execution object can be obtained from the organizational relationship information, which improves the comprehensiveness and accuracy of data integration for the preset mapping information.

[0104] In some embodiments, the task processing method may be applied to an enterprise knowledge system, which includes a user layer, an identification layer, a standardization layer, and a storage layer. The user layer is used to obtain a task instruction, behavior association information, and organizational relationship information corresponding to a task to be processed. The back-end server includes an identification layer, a standardization layer, and a storage layer. The back-end server can identify, perform context analysis, and standardize the task instruction, behavior association information, and organizational relationship information to obtain one or more pieces of information such as timestamp information, intention information, sorting information, constraint information, media information, and object information. After the back-end server performs context analysis and standardization on the behavior association information and organizational relationship information, an execution object and context key information are obtained, and a mapping relationship between the execution object and the context key information is established to obtain preset mapping information, which is stored in the storage layer. After the back-end server performs context analysis and standardization conversion on the task instruction, target context key information and task execution content are obtained. The back-end server obtains a target execution object corresponding to the task execution content from the preset mapping information based on the target context key information, and the back-end server can execute the task to be processed based on the target execution object.

[0105] Please refer to Figure 9 , such as Figure 9 shown in the application scenario diagram of the enterprise knowledge system. When the enterprise knowledge system performs an action, it can interface with a third-party application to execute functions such as restaurant push, meeting room reservation, or leave application through the third-party application. For example, when the task instruction is "push the connected Sichuan cuisine restaurants within 500 meters nearby to my secretary", the enterprise knowledge system can interface with a map software to obtain Sichuan cuisine restaurants within 500 meters from the map software. When the task instruction is "find a currently idle meeting room near me and help me reserve and occupy it first", the enterprise knowledge system can interface with a meeting room reservation system to determine an idle meeting room nearby and change the usage status of the meeting room. When the task instruction is "I'm not feeling well this afternoon, help me apply for a leave with my boss and submit a leave form", the enterprise knowledge system can interface with a leave application system to obtain a leave form template file and fill in the leave form template file to generate a leave form.

[0106] When the enterprise knowledge system conducts information query, it can query conversations or meetings. For example, "help me list the conversations about 'product manager recruitment requirements'", "help me find a meeting with someone participating and mention the record about 'deployment cost'", "find the word document with the name containing 'new infrastructure'". When the enterprise knowledge system conducts information query, it can also query organizational relationship information. For example, "query whether someone's boss is someone", "send a message to the secretary of the R & D center and leave my mobile phone number".

[0107] The enterprise knowledge system may also include a knowledge base. The organizational relationship information can be stored as data in the knowledge base, and the preset mapping relationship can be updated based on the data in the knowledge base. Based on the Q&A instructions corresponding to the information input by the user, the interface person, process, tutorial, statistical results, etc. can be directly queried from the knowledge base of the enterprise knowledge system, and the query results are fed back to the user as the Q&A results. For example, when the task instruction is "Who is the person in charge of the procurement contract approval? Who should I ask for the procurement supplier requirement document?", the interface person for the procurement contract approval can be queried and the relevant information of this interface person is fed back to the user. When the task instruction is "How does the procurement process proceed? What are the requirements and restrictions for the quotation forms of three suppliers?", the procurement process can be queried and the relevant information of the procurement process is fed back to the user. When the task instruction is "How many people are there in the design center in total and what is the proportion of the total number of people in the department?", the number of people in the design center can be queried, and the proportion of the number of people in the design center to the total number of people in the department is calculated, and the statistical results are fed back to the user. When the task instruction is "Where can I find the design materials for advertising? What software is generally used?", the tutorial information on how to obtain the design materials can be obtained, and the tutorial on how to obtain the design materials is fed back to the user.

[0108] An embodiment of the present application provides a task processing method, which includes: obtaining a task instruction corresponding to a task to be processed, performing intention recognition and context analysis on the task instruction to determine the task execution content and the key information of the target context, based on the key information of the target context, obtaining a target execution object corresponding to the task execution content from the preset mapping information, and based on the target execution object, executing the task to be processed. This method can obtain the preset mapping information through the corresponding relationship between the key information of the context and the execution object, which can improve the comprehensiveness and accuracy of data integration of the preset mapping information. Moreover, this method performs standardized conversion on the key information of the context and the execution object, making the preset mapping information easy to expand and reuse, thereby improving the richness of the work scenario. This method also determines the target execution object based on the key information of the context and executes the task to be processed, which can improve the efficiency and accuracy of task processing.

[0109] An embodiment of the present application also provides a task processing device, please refer to Figure 10 , the device includes: a task instruction acquisition module 1010, a task instruction analysis module 1020, a target task search module 1030, and a target task execution module 1040;

[0110] The task instruction acquisition module 1010 is used to obtain a task instruction corresponding to a task to be processed;

[0111] The task instruction analysis module 1020 is used to perform semantic analysis on the task instruction to determine the task execution content and the key information of the target context;

[0112] The target task search module 1030 is used to obtain a target execution object corresponding to the task execution content from the preset mapping information based on the target context key information, where the preset mapping information represents the mapping relationship between multiple context key information and at least one execution object corresponding to each of them;

[0113] The target task execution module 1040 is used to execute the task to be processed based on the target execution object.

[0114] Furthermore, the task instruction analysis module includes:

[0115] The text feature extraction unit is used to obtain the text feature information corresponding to the task instruction;

[0116] The intention recognition unit is used to perform intention recognition on the text feature information to obtain the task execution content;

[0117] The context analysis unit is used to perform context analysis on the text feature information to determine the context key information.

[0118] Furthermore, the target task search module includes:

[0119] The object identifier extraction unit is used to extract the object identifier corresponding to the target execution object from the task execution content;

[0120] The context key information matching unit is used to determine at least one execution object that matches the target context key information based on the preset mapping information;

[0121] The target execution object determination unit is used to determine the target execution object from at least one execution object that matches based on the object identifier.

[0122] Furthermore, the device includes:

[0123] The query result sorting module is used to sort multiple query results based on a preset priority to obtain a query result sequence when the task execution content is query content and the target execution object includes multiple query results;

[0124] The target task execution module includes:

[0125] The query result processing unit is used to execute the task to be processed based on the query result sequence.

[0126] Furthermore, the device further includes:

[0127] The execution order acquisition module is used to obtain the execution order of multiple actions corresponding to the action content when the task execution content is action content and the action content corresponds to multiple actions;

[0128] The target task execution module includes:

[0129] An execution order processing unit for executing a task to be processed based on an execution order and a target execution object.

[0130] Furthermore, the device further includes:

[0131] A behavior association acquisition module for acquiring behavior association information corresponding to a target user;

[0132] A behavior association analysis module for performing context analysis on the behavior association information to obtain corresponding context key information and execution objects;

[0133] A mapping information acquisition module for establishing a mapping relationship between the corresponding context key information and execution objects to obtain preset mapping information.

[0134] Furthermore, the device further includes:

[0135] An organizational relationship acquisition module for acquiring organizational relationship information corresponding to a target user;

[0136] A mapping information update module for updating the preset mapping information based on the organizational relationship information.

[0137] The device provided in the above embodiment can execute the method provided in any embodiment of the present application, and has corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in the above embodiment can be found in a task processing method provided in any embodiment of the present application.

[0138] This embodiment also provides a computer-readable storage medium storing computer-executable instructions, and the computer-executable instructions are loaded and executed by a processor to execute a task processing method as described above in this embodiment.

[0139] This embodiment also provides a computer program product or a computer program. The computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the methods provided in various alternative implementations of the above task processing.

[0140] This embodiment also provides an electronic device including a processor and a memory. The memory stores a computer program, and the computer program is adapted to be loaded and executed by the processor to execute a task processing method as described above in this embodiment.

[0141] The device can be a computer terminal or a mobile terminal, and the device can also participate in constituting the device or system provided in the embodiments of the present application. For example Figure 11As shown, the server 11 may include one or more processors 1102 (shown as 1102a, 1102b, ……, 1102n in the figure), a processor 1102 may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA, a memory 1104 for storing data, and a transmission device 1106 for communication functions. In addition, it may further include: a display, an input / output interface (I / O interface), a network interface, a power supply, and / or a camera. Those of ordinary skill in the art can understand that Figure 11 the structure shown is only schematic and does not limit the structure of the above electronic device. For example, the server 11 may further include more or fewer components than those Figure 11 shown in, or have a different configuration from that Figure 11 shown.

[0142] It should be noted that the above one or more processors 1102 and / or other data processing circuits are generally referred to as "data processing circuits" herein. The data processing circuit may be embodied in software, hardware, firmware, or any combination thereof, in whole or in part. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any one of the other elements in the server 11.

[0143] The memory 1104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the method in the embodiments of the present application. The processor 1102 executes various functional applications and data processing by running the software programs and modules stored in the memory 1104, that is, implements the above method for generating a timing behavior capture frame based on a self-attention network. The memory 1104 may include a high-speed random access memory, and may further include a non-volatile memory, such as one or more magnetic storage devices, a flash memory, or other non-volatile solid-state memories. In some instances, the memory 1104 may further include a memory remotely disposed relative to the processor 1102, and these remote memories may be connected to the server 11 through a network. Examples of the above network include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0144] The transmission device 1106 is used to receive or send data via a network. Specific examples of the above network may include a wireless network provided by a communication provider of the server 11. In one instance, the transmission device 1106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station and thus communicate with the Internet. In one instance, the transmission device 1106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0145] The display can be, for example, a touch-screen liquid crystal display (LCD), which enables a user to interact with the user interface of the server 11.

[0146] This specification provides method operation steps such as in the embodiments or flowcharts, but may include more or fewer operation steps based on routine or non-creative labor. The steps and sequences listed in the embodiments are only one way among many sequences of steps to be executed and do not represent the only execution sequence. When the actual system or interrupt product is executed, it can be executed in the order shown in the embodiments or the drawings or in parallel (e.g., in an environment of parallel processors or multi-threaded processing).

[0147] The structure shown in this embodiment is only a part of the structure related to the solution of this application and does not constitute a limitation on the devices to which the solution of this application is applied. The specific devices may include more or fewer components than those shown, or combine certain components, or have different arrangements of components. It should be understood that the methods, devices, etc. disclosed in this embodiment can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of modules is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, indirect couplings or communication connections of devices or unit modules.

[0148] Based on such an understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in various embodiments of this application. And the foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0149] Those skilled in the art can further realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this specification can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0150] In the above, the above embodiments are only used to illustrate the technical solutions of this application, rather than to limit it; although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of this application.

Claims

1. A task processing method, characterized in that, The method includes: Obtaining behavior association information corresponding to a target object; the behavior association information includes text conversations, links or attachments in a context, voice conversations, and meeting records in enterprise knowledge data; Performing context analysis on the behavior association information to obtain context key information and an execution object; performing standardized conversion on the context key information and the execution object according to a preset standardized category, and obtaining preset mapping information based on the mapping relationship between the context key information and the execution object after the standardized conversion; the preset mapping information represents the mapping relationship between multiple context key information and at least one execution object corresponding to each of them, and the preset standardized category includes timestamp information, sorting information, constraint information, media information, and object information; Obtaining a task instruction corresponding to a task to be processed; Performing semantic analysis on the task instruction to determine task execution content and target context key information; Based on the target context key information, obtaining a target execution object corresponding to the task execution content from the preset mapping information; Based on the target execution object, executing the task to be processed.

2. The task processing method according to claim 1, wherein, The performing semantic analysis on the task instruction to determine task execution content and target context key information includes: Obtaining text feature information corresponding to the task instruction; Performing intention recognition on the text feature information to obtain the task execution content; Performing context analysis on the text feature information to determine the target context key information.

3. The task processing method according to claim 1, wherein The obtaining a target execution object corresponding to the task execution content from the preset mapping information based on the target context key information includes: Extracting an object identifier corresponding to the target execution object from the task execution content; Based on the preset mapping information, determining at least one execution object that matches the target context key information; Based on the object identifier, determining the target execution object from the at least one matching execution object.

4. The task processing method according to any one of claims 1 to 3, characterized in that, The method further includes: In the case where the task execution content is a query content and the target execution object includes multiple query results, sorting the multiple query results based on a preset priority to obtain a query result sequence; The executing the task to be processed based on the target execution object includes: Executing the task to be processed based on the query result sequence.

5. The task processing method according to any one of claims 1 to 3, characterized in that, The method further includes: In the case where the task execution content is an action content and the action content corresponds to multiple actions, obtaining the execution order of the multiple actions corresponding to the action content; The executing the task to be processed based on the target execution object includes: Executing the task to be processed based on the execution order and the target execution object.

6. The task processing method according to claim 1, characterized in that The method further includes: Obtaining organizational relationship information corresponding to a target user; Updating the preset mapping information based on the organizational relationship information.

7. A task processing device, characterized in that, The apparatus includes: a behavior association obtaining module, a behavior association analysis module, a mapping information obtaining module, a task instruction obtaining module, a task instruction analysis module, a target task searching module, and a target task executing module; The behavior association acquisition module is used to acquire behavior association information corresponding to a target object; the behavior association information includes text conversations, links or attachments in the context, voice conversations, and meeting records in enterprise knowledge data; The behavior association analysis module is used to perform context analysis on the behavior association information to obtain context key information and execution objects; according to a preset standardization category, perform standardization conversion on the context key information and the execution objects, and the preset standardization category includes timestamp information, sorting information, constraint information, media information, and object information; The mapping information acquisition module is used to obtain preset mapping information based on the mapping relationship between the context key information and the execution objects after standardization conversion; the preset mapping information represents the mapping relationship between multiple context key information and at least one corresponding execution object; The task instruction acquisition module is used to acquire a task instruction corresponding to a task to be processed; The task instruction analysis module is used to perform semantic analysis on the task instruction to determine task execution content and target context key information; The target task search module is used to obtain a target execution object corresponding to the task execution content from the preset mapping information based on the target context key information; The target task execution module is used to execute the task to be processed based on the target execution object.

8. An electronic device, characterized in that, The electronic device includes a processor and a memory, and at least one instruction or at least one program segment is stored in the memory, and the at least one instruction or the at least one program segment is loaded and executed by the processor to implement a task processing method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The storage medium includes a processor and a memory, and at least one instruction or at least one program segment is stored in the memory, and the at least one instruction or the at least one program segment is loaded and executed by the processor to implement a task processing method according to any one of claims 1-6.

Images