A method, apparatus, device and medium for processing an active task

Abstract

The application discloses a kind of active task processing method, device, equipment and medium.The method comprises: the present application embodiment can create active opportunity according to current fact set, persistence duty, yet to be audited suggestion and recent audit decision;The persistence duty indicates that the work that system has undertaken;The active opportunity is converted into to-be-audited suggestion, and the audit decision of user to the to-be-audited suggestion is obtained;According to the to-be-audited suggestion and to-be-audited suggestion audit decision, update persistence duty, and according to the interrelated execution instance, execution qualification, persistent execution carrier and running context of updated persistence duty;Based on the interrelated execution instance, execution qualification, persistent execution carrier and running context, call the model in intelligent agent system to execute task and obtain result report.The present application embodiment can improve the efficiency and accuracy of intelligent agent system task processing.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method, apparatus, device, and medium for processing proactive tasks. Background Technology

[0002] Intelligent agent systems typically rely on real-time user input as the primary trigger. After a user sends a message, the intelligent agent system invokes the model to read the current dialogue and provide a response. When there are no new messages, the system generally stops monitoring the workspace and does not proactively discover new issues across messages, channels, or execution cycles. This reactive structure is suitable for one-time question-and-answer sessions but cannot handle continuous monitoring, periodic refactoring, risk detection, or long-term follow-up.

[0003] Existing technologies sometimes use traditional scheduled tasks to execute pre-written actions. Developers or users predefine fixed periods and callback procedures, and the system runs directly upon expiration. This approach can solve the problem of "when a known task will be executed," but it cannot address "what new situations have arisen in the current workspace, whether it is worthwhile to create a task, why the task is worthwhile, and whether it should be undertaken by an agent." Furthermore, the information faced by proactive systems typically comes from multiple people, multiple channels, and long-term history, potentially leading to uncertainty in model judgments. If the system lacks user review boundaries, it may create long-term plans based on misinterpreted information, notify irrelevant personnel, or invoke operations that alter external states. Simply requiring "caution" in the prompt text cannot establish stable technical constraints, nor can it retain user acceptance or rejection decisions for subsequent analysis. In other words, with minimal human intervention, tasks planned and executed autonomously by agent systems using existing technologies have low accuracy and often fail to meet user expectations. With significant human intervention, the agent needs to wait for human feedback and confirmation at each stage, resulting in low efficiency. Summary of the Invention

[0004] This invention provides a method, apparatus, device, and medium for processing active tasks, thereby improving the efficiency and accuracy of task processing in intelligent agent systems.

[0005] According to one aspect of the present invention, a method for processing active tasks is provided, comprising: Proactive opportunities are created based on the current set of facts, ongoing responsibilities, pending review recommendations, and recent review decisions; the ongoing responsibilities indicate the work already undertaken by the system. The proactive opportunities are converted into suggestions for review, and the user's review decision on the suggestions for review is obtained; The persistent responsibilities are updated based on the proposed review and the review decision of the proposed review, and interrelated execution instances, execution qualifications, persistent execution carriers and runtime contexts are generated based on the updated persistent responsibilities. Based on the interrelated execution instances, execution qualifications, persistent execution carriers, and runtime contexts, the model in the intelligent agent system is invoked to execute tasks and obtain result reports.

[0006] According to another aspect of the present invention, an active task processing apparatus is provided, comprising: The opportunity creation module is used to create proactive opportunities based on the current set of facts, ongoing responsibilities, pending review recommendations, and recent review decisions; the ongoing responsibilities indicate the work already undertaken by the system. The suggestion review module is used to convert the proactive opportunity into a suggestion to be reviewed, and to obtain the user's review decision on the suggestion to be reviewed; The task preparation module is used to update the persistent responsibilities based on the proposed review and the review decision of the proposed review, and to generate interrelated execution instances, execution qualifications, persistent execution carriers and runtime contexts based on the updated persistent responsibilities. The task execution module is used to call the model in the intelligent agent system to execute tasks and obtain result reports based on the interrelated execution instances, execution qualifications, persistent execution carriers and running contexts.

[0007] According to another aspect of the present invention, a computer program product is provided, the computer program product comprising a computer program that, when executed by a processor, implements the active task processing method according to any embodiment of the present invention.

[0008] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the active task processing method described in any embodiment of the present invention.

[0009] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the active task processing method described in any embodiment of the present invention.

[0010] In this embodiment of the invention, the system automatically organizes the current facts to generate proactive tasks. Users only need to perform a simple review of the suggestions to be reviewed generated based on the proactive tasks, which greatly reduces human intervention. At the same time, the subsequent processing of proactive tasks is continued through multiple constraints to ensure that proactive tasks can be accurately processed with minimal human intervention, which greatly improves the efficiency and accuracy of task processing in the intelligent agent system.

[0011] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart of a method for processing an active task according to an embodiment of the present invention; Figure 2A This is a flowchart of a method for processing an active task according to another embodiment of the present invention; Figure 2B This is a schematic diagram of an active task processing flow according to another embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an active task processing device according to another embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device that implements an embodiment of the present invention. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0015] It should be noted that the terms "first," "second," etc., used in this invention 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 so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0016] Figure 1 This is a flowchart illustrating a proactive task processing method according to an embodiment of the present invention. This embodiment is applicable to proactive task discovery, generation, and closed-loop processing methods and systems based on multi-source fact analysis. It enables an intelligent agent to identify worthwhile tasks from continuously arriving information even without real-time user messages, while ensuring that each suggestion has a source, each ongoing responsibility is confirmed by the user, each execution is reversible, each completion result is verifiable, and each review does not bypass user authorization. This method can be executed by a proactive task processing device, which can be implemented in hardware and / or software. This device can be configured in an electronic device with corresponding data processing capabilities, such as an intelligent agent system. Figure 1 As shown, the method includes: S110. Create proactive opportunities based on the current set of facts, ongoing responsibilities, pending audit recommendations, and recent audit decisions.

[0017] S120. Convert the proactive opportunity into a suggestion to be reviewed, and obtain the user's review decision on the suggestion to be reviewed.

[0018] S130. Update the continuing responsibilities based on the proposed review and the review decision of the proposed review, and generate interrelated execution instances, execution qualifications, persistent execution carriers and runtime contexts based on the updated continuing responsibilities.

[0019] S140. Based on the interrelated execution instances, execution qualifications, persistent execution carriers, and runtime contexts, the model in the intelligent agent system is invoked to execute tasks and obtain result reports.

[0020] The continuous responsibilities indicate the work already undertaken by the system; the current fact set describes the current state of workspace facts (workspace facts are stable information compiled from one or more sources, including facts, resources, explicit decisions, and user preferences. It is the current state of an object-oriented system, not a message flow, to-do list, or execution log); pending review suggestions describe review suggestions that have not yet been decided; and recent review decisions describe review suggestions that users have recently accepted or rejected.

[0021] Specifically, records from sources such as messages, business events, manual input, historical execution results, and user feedback are retrieved, deduplicated, cleaned, and formatted to form standardized record batches. Using these standardized record batches, combined with existing facts and necessary historical data, stable objects are identified, and a set of current facts with evidentiary citations is formed.

[0022] By leveraging the current set of facts and reading existing ongoing responsibilities, pending audit recommendations, and recent audit decisions, one or more structured proactive opportunities can be generated.

[0023] The system establishes a source evidence record for proactive opportunities, and then saves the review recommendations generated based on these opportunities along with the evidence within the same transaction boundary. Review recommendations retain the opportunity identifier, triggering basis, execution plan, and optional execution parameters, allowing each recommendation to trace its source.

[0024] The system receives review requests from the user interface, message cards, or controlled approval entry points. Users can accept, reject, or edit pending review suggestions before accepting them. Based on this, the system generates a review decision that includes a suggestion identifier, decision status, decision time, and decision-making body.

[0025] Establish or adjust ongoing responsibilities using accepted audit decisions and develop an implementation plan consistent with those responsibilities; rejected, withdrawn, or substituted pending audit decisions serve as the basis for identifying subsequent opportunities.

[0026] When the execution plan expires or a controlled event is triggered, an execution instance and a persistent execution carrier are created or reused to obtain time-limited execution eligibility and form a runtime context associated with that instance.

[0027] Based on the generated execution instance, execution qualification, persistent execution carrier, and runtime context, the model in the intelligent agent system is invoked to execute the task, the actual results are registered, and a completion report containing a result summary or a failure report containing the reasons for failure is submitted as the result report.

[0028] In this embodiment of the invention, the system automatically organizes the current facts to generate proactive tasks. Users only need to perform a simple review of the suggestions to be reviewed generated based on the proactive tasks, which greatly reduces human intervention. At the same time, the subsequent processing of proactive tasks is continued through multiple constraints to ensure that proactive tasks can be accurately processed with minimal human intervention, which greatly improves the efficiency and accuracy of task processing in the intelligent agent system.

[0029] Based on the above embodiments, the proactive opportunity includes opportunity type, brief title, specific delivery target, triggering basis, execution plan and evidence reference; the opportunity type includes tasks that can be proactively undertaken, new ongoing responsibilities, and adjusted ongoing responsibilities.

[0030] The triggering criteria explain why action should be taken at this time; the evidence citation indicates the source of the judgment; the execution plan explains how the system will work after user acceptance; and the evidence citation is stable, related information that traces back from workspace facts, proactive opportunities, pending review suggestions, or execution results to the source record or outcome object. In addition to the information mentioned above, proactive opportunities may also include target responsibility identifiers and content related to responsibility changes.

[0031] Specifically, the system first performs cross-record semantic analysis on the current set of facts, ongoing responsibilities, pending review suggestions, and recent review decisions to determine whether the facts form matters worthy of proactive action and to formulate an implementation plan. Then, it performs deterministic validation, checking opportunity types, required fields, plan formats, evidence citations, and their relationship with existing responsibilities, outputting structured proactive opportunities. When there are no tasks worthy of user attention in a regular batch, an empty set is allowed to be output, i.e., no proactive opportunities are output. The opportunity types and their definitions (usage scenarios, additional validations) for proactive opportunities are shown in Table 1 below: Table 1 Structured proactive opportunities are used to generate suggestions for review, but at this stage, no plans are directly created, nor are operations that change external states invoked. In other words, proactive opportunities are merely internal candidate work packages. Their content should be sufficient for users to understand "what to do, why to do it now, the basis for it, and what to do after acceptance," but periodic plans should not be directly created or operations that change external states should be invoked at this stage.

[0032] In addition, the identification of proactive opportunities also adopts the following exclusion rules: immediate responses are part of the current session processing and do not constitute proactive opportunities awaiting review; system maintenance, such as fixing its own data gaps, clearing internal queues, or verifying its own output, is internal maintenance and is not submitted to users; in collaborative spaces without ongoing responsibilities, long-term work is not established based on weak inferences unless the source record contains an explicit request; matters that have recently been rejected by users and are not supported by new facts are not raised again.

[0033] When a candidate involves adjusting an existing responsibility, the system first reads the number of times that responsibility has been executed, the completion result, the history of any obstacles, work records, and relevant user feedback. If the responsibility has just been adjusted but no new execution result has yet been generated, there is a lack of basis for judging the effect of the adjustment, and the system should leave it unchanged. This historical data reading prerequisite can be verified by the program: candidates who have not completed historical data reading should not be registered as responsibility adjustment opportunities.

[0034] Based on the above embodiments, optionally, updating the continuing responsibility according to the proposed review and the review decision of the proposed review includes: If the suggestion to be reviewed is a new suggestion and the review decision is "accepted", then a new ongoing responsibility shall be established based on the content of the suggestion to be reviewed. If the suggestion to be reviewed is an adjustment suggestion and the review decision is to accept it, then the corresponding continuing responsibilities will be adjusted according to the content of the suggestion to be reviewed.

[0035] Specifically, for newly added suggestions awaiting review, when the system decides that the suggestion has been accepted, it verifies the cycle plan, time zone, and execution instructions, determines the responsible party and the corresponding collaboration channel, establishes new ongoing responsibilities, and simultaneously generates an execution plan.

[0036] For adjustment suggestions pending review, the system will still verify the target responsibility when the review decision is accepted; however, only the persistent fields explicitly listed in the suggestion, such as cycle, execution description, responsible entity, or enabled status, will be applied; fields not listed, such as title and description, will retain their original values.

[0037] In cases where a recommendation is rejected, the rejection decision is saved without establishing or modifying ongoing responsibilities; the subsequent identification module reads this decision and, unless substantial new evidence emerges, does not resubmit it.

[0038] The different review decisions, their follow-up processing, and status constraints are shown in Table 2 below: Table 2 Prior to the above embodiments, optionally, before creating the proactive opportunity based on the current set of facts, sources of ongoing responsibility, pending audit recommendations, and recent audit decisions, the method further includes: The source records of external input are structured and organized to obtain the standardized record batch of the current time window; The historical fact set is updated based on the normalized record batches of the current time window to obtain the current fact set.

[0039] Source records are original information obtained from message channels, business systems, historical execution, user feedback, or manual entry. They must include at least the source, time, and stable references for future review. Standardized records are analyzable records formed after standardizing the format of source records, eliminating duplicates, cleaning the text, and supplementing necessary metadata.

[0040] Specifically, the system determines the collaboration space to which each externally input source record belongs (a collaboration space refers to a data isolation range with independent members, message sources, continuous responsibilities, fact items, and execution records. Fact gathering, suggestion review, and proactive execution are all carried out within a single collaboration space), and saves the source identifier, transmission type, account or system identifier, external session identifier, record generation time, and access time. When multiple source records arrive within the same analysis window, they form a batch to be processed; when the same source record arrives again, duplicates are removed based on the stability identifier.

[0041] The system extracts analyzable text from source records, removes irrelevant transport wrappers, standardizes time, encoding, and citation formats, and retains the location information to return to the original record. For citations where the original content cannot be obtained, no corresponding normalized record is generated in this batch, nor is the processed position of that source advanced. Using the generated normalized record batches, combined with existing facts and necessary history, stable objects are identified, and a set of current facts with evidence citations is formed.

[0042] Prior to the above embodiments, optionally, updating the historical fact set according to the normalized record batch of the current time window to obtain the current fact set includes: For each normalized record in the batch of normalized records, determine the stable objects in the normalized records; Based on the target fact type to which the stable object belongs, determine the appropriate content to be stored in each normalized record; The historical fact set is updated based on the appropriate content to be preserved in each normalized record to obtain the current fact set.

[0043] The types of facts include facts, resources, decisions, or preferences.

[0044] Specifically, the system first identifies stable objects that can be reused across batches from the normalized records, such as project goals, long-used documents, established decisions, or long-term user preferences.

[0045] The stable object is categorized into one of four fact types. Temporary events, open tasks, system operation information, and the current analysis process are not included in long-term facts. Each fact type has corresponding suitable and unsuitable content for preservation. Based on the target fact type to which the stable object belongs, and the suitable and unsuitable content for preservation of the target fact type, the suitable content for preservation in the normalized record is selected.

[0046] Read the existing facts and necessary history from the historical fact set, determine whether to keep it unchanged, update the current state, or delete the expired existing facts and necessary history based on the appropriate content to be saved for each normalized record, and update the historical fact set accordingly to obtain the current fact set with version, and mark the previous version of the current fact set as replaced.

[0047] For example, the suitable and unsuitable content for saving different event types is shown in Table 3 below: Table 3 By compressing an infinitely growing sequence of messages into a finite, updatable, and traceable set of current facts, subsequent opportunity identification does not require reinterpreting the entire history each time, nor does it lose the source evidence due to the summary.

[0048] Figure 2A This is a flowchart illustrating a method for processing active tasks according to another embodiment of the present invention. This embodiment is an optimization and improvement upon the above embodiment. Figure 2A As shown, the method includes: S210. Create proactive opportunities based on the current set of facts, ongoing responsibilities, pending audit recommendations, and recent audit decisions.

[0049] S220. Convert the proactive opportunity into a suggestion to be reviewed, and obtain the user's review decision on the suggestion to be reviewed.

[0050] S230. Update the continuing responsibility according to the proposed review and the review decision of the proposed review; when the updated continuing responsibility meets the triggering conditions, generate or reuse a unique execution instance; obtain the limited-term execution qualification of the execution instance through assignment.

[0051] S240. If the execution instance does not have an associated persistent execution carrier, then a persistent execution carrier for the execution instance is generated based on the execution context of the execution instance; if the execution instance has an associated persistent execution carrier, then the persistent execution carrier and its execution context are reused.

[0052] Specifically, when a plan expires or a controlled event meets the triggering conditions, the system generates an execution request. The continuous responsibility stores the title, description, cycle, time zone, execution instructions, responsible party, affiliated collaboration channel, and activation status. The execution request itself does not directly indicate that processing has begun.

[0053] Each trigger first searches for execution instances still in a pending state within the same collaboration space and the same persistent responsibility scope. If one exists, the earlier instance is reused without being queued again; if the responsibility already has an instance in execution, the new cycle trigger does not overlap with the second round of valid execution. If no reusable instance exists, the system creates a new execution instance and saves the trigger plan, trigger time, source suggestion, and evidence relationship.

[0054] Before assigning tasks, the system checks that persistent responsibilities still exist and are not archived, that the relevant collaboration channels are still valid, that the responsible entity can be resolved, and whether other valid active executions already exist in the same collaboration space. Once these conditions are met, the instance is changed from pending execution to execution through a persistent update with state preconditions. In the event of multiple processes competing simultaneously, only one update will succeed; unsuccessful processes will reread their current state without creating a new execution vehicle.

[0055] After obtaining execution eligibility through assignment, the system searches for the persistent execution carrier associated with that instance. If it already exists, the original record is used to avoid repeatedly creating external work after a service restart; if it does not exist, a new record is created, containing the current responsibility description, execution subject, result recipient, execution instance association, and running strategy, serving as the running context. After the running context is established, it is written back to the execution instance, thus maintaining a stable relationship between the execution instance, execution carrier, and the specific execution.

[0056] The system grants execution instances a time-limited execution qualification to bridge the time gap between persistent state and actual operation. While the qualification is still valid, other instances cannot enter the same mutually exclusive scope; when the qualification expires and the original operation has lost its effective association, the system reverts the instance to pending execution, registration blocked, or cancellation based on the execution instance, execution carrier, and running status.

[0057] S250. Based on the interrelated execution instances, execution qualifications, persistent execution carriers, and runtime contexts, the model in the intelligent agent system is invoked to execute tasks and obtain result reports.

[0058] Based on the above embodiments, optionally, the report type of the result report includes a completion report or a blocked report; after obtaining the result report by calling the model in the intelligent agent system to execute the task based on the interrelated execution instances, execution qualifications, persistent execution carriers, and runtime contexts, it further includes: Based on the report type of the result report, the existence of evidence in the content of the result report is verified; If the evidence existence verification passes, a work record indicating successful task execution or a record indicating failed task execution will be generated based on the result report. The fact update suggestion is generated based on the work record, or the responsibility adjustment suggestion is generated based on the obstructed record or the responsibility adjustment suggestion; the fact update suggestion is used to update the current fact set, and the responsibility adjustment suggestion is used to adjust the existing ongoing responsibilities.

[0059] Specifically, for task execution results, the system only accepts two types of business conclusions: "completed" and "obstructed".

[0060] For completed reports, the requirements for the existence verification of evidence are as follows: the completed report must contain a non-empty results summary and at least one piece of results evidence that belongs to the current collaboration space and can be read again.

[0061] For obstructed reports, the requirements for the existence verification of evidence are: the obstructed report must have an understandable reason and a stable category of obstruction.

[0062] When verification fails, a completion status without basis is not recorded. The existence of a runtime context, the model stopping output, the execution carrier being shut down, or the execution duration reaching its limit all indicate changes in the processing procedure. Only completion reports that pass both evidence existence verification and more basic field verification are registered as completed. For completion reports, a work record is created, storing associated ongoing responsibilities, execution instances, runtime context, result summary, and achievement evidence. For obstructed reports, an obstruction record is created, storing the reason for the obstruction, the type of obstruction, and the number of times similar problems occurred. Completion is confirmed only when both the result summary and achievement evidence exist, and the evidence identifier must actually exist; therefore, it is possible to distinguish between "a plan was generated," "the run ended," and "truly verifiable results were produced."

[0063] The system writes completed execution instances and their work records or blocked records to the "Pending Review" window. For work records in the "Pending Review" window, the system generates fact update suggestions, which will be referenced when updating the current fact set. For blocked records in the "Pending Review" window, the system generates responsibility adjustment suggestions. These suggestions must be reviewed by the user again; ongoing responsibilities cannot be directly modified without user approval.

[0064] This invention integrates the completion results, obstacle categories, user feedback, and historical responsibility adjustments through a retrospective analysis; it updates the facts with stable conclusions and reformulates suggestions for review based on behavioral changes, enabling the system to improve without losing control.

[0065] For example, Figure 2B This is a schematic diagram of a full-processing workflow for an active task provided in this application.

[0066] Based on this process, this application, compared to existing technologies, can: • Improve the accuracy of proactive task discovery. Fact-checking first eliminates format differences and duplicate records, then merges multiple mentions of the same object into the current status; opportunity identification reads existing responsibilities and historical decisions, resulting in fewer outdated, duplicate, or unfounded suggestions compared to generating to-dos for only a single batch of messages.

[0067] • Reduce noise in user review processes. The system allows for no output when there is no opportunity, identifies duplicate suggestions by standardizing titles or target responsibilities, replaces old suggestions with new ones, and sets a configurable upper limit for the queue of suggestions to be reviewed, so that users see a small number of work suggestions with deliverables and justification.

[0068] • Reduce the risk of unauthorized external operations. Separate proactive opportunities from ongoing responsibilities; user acceptance is a necessary condition for establishing execution permissions; retrospectives should not directly modify long-term behaviors, and the entire process from discovery to adjustment should go through the same review boundary.

[0069] • Improve continuity across batches and restarts. Facts, recommendations, responsibilities, execution examples, evidence of results, and work records are all persisted, and the model does not need to remember the results of the last analysis; after the service restarts, it can continue to read the current facts, pending status, and original execution carrier.

[0070] • Reduce duplicate execution and side effects. The trigger phase reuses the instance to be executed, and the dispatch phase obtains unique execution qualification through state preconditions. If an execution carrier already exists, it is restored along the original record, reducing duplicate external writes caused by multi-process competition, duplicate scheduling and restarts.

[0071] • Improve the credibility of completion determination. Completion is confirmed only when both the results summary and the evidence of achievements exist, and the evidence identifier must actually exist; thus, it is possible to distinguish between "a plan has been generated", "the operation has ended", and "truly verifiable results have been produced".

[0072] • Form a continuous improvement loop. The review integrates the completed results, categories of obstacles, user feedback, and history of responsibility adjustments; stabilize the conclusions, update the facts, and reformulate the suggestions for review based on behavior changes, so that the system can be improved without losing control.

[0073] Figure 3 This is a schematic diagram of a processing device for an active task, provided in another embodiment of the present invention. Figure 3 As shown, the device includes: Opportunity creation module 310 is used to create proactive opportunities based on the current set of facts, ongoing responsibilities, pending review suggestions, and recent review decisions; the ongoing responsibilities indicate the work already undertaken by the system. The suggestion review module 320 is used to convert the proactive opportunity into a suggestion to be reviewed, and to obtain the user's review decision on the suggestion to be reviewed; The task preparation module 330 is used to update the persistent responsibilities based on the proposed suggestions for review and the review decision of the proposed suggestions for review, and to generate interrelated execution instances, execution qualifications, persistent execution carriers and runtime contexts based on the updated persistent responsibilities. The task execution module 340 is used to call the model in the intelligent agent system to execute tasks and obtain result reports based on the interrelated execution instances, execution qualifications, persistent execution carriers and running contexts.

[0074] The active task processing apparatus provided in the embodiments of the present invention can execute the active task processing method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0075] Optionally, the runtime context includes a description of responsibilities, an execution subject, a result receiving subject, associated execution instances, and a runtime strategy. The task preparation module 330 is specifically used for: When the updated persistent responsibility meets the triggering conditions, a unique execution instance is generated or reused; The execution instance is granted a limited-term execution qualification through assignment; If the execution instance does not have an associated persistent execution carrier, then a persistent execution carrier for the execution instance is generated based on the execution context of the execution instance; If the execution instance has an associated persistent execution carrier, then the persistent execution carrier and its runtime context are reused.

[0076] Optionally, the result report type includes a completion report or an obstruction report, and the device further includes a task review module for: Based on the report type of the result report, the existence of evidence in the content of the result report is verified; If the evidence existence verification passes, a work record indicating successful task execution or a record indicating failed task execution will be generated based on the result report. The fact update suggestion is generated based on the work record, or the responsibility adjustment suggestion is generated based on the obstructed record or the responsibility adjustment suggestion; the fact update suggestion is used to update the current fact set, and the responsibility adjustment suggestion is used to adjust the existing ongoing responsibilities.

[0077] Optionally, the device further includes: The record organization module is used to organize the source records of external input in a structured way to obtain the standardized record batch of the current time window; The fact update module is used to update the historical fact set according to the normalized record batch of the current time window to obtain the current fact set.

[0078] Optionally, the fact update module is specifically used for: For each normalized record in the batch of normalized records, determine the stable objects in the normalized records; Based on the target fact type to which the stable object belongs, determine the appropriate content to be stored in each normalized record; wherein, the fact type includes facts, resources, decisions, or preferences; The historical fact set is updated based on the appropriate content to be preserved in each normalized record to obtain the current fact set.

[0079] Optionally, the proactive opportunity includes opportunity type, brief title, specific delivery target, triggering basis, execution plan and evidence reference; the opportunity type includes tasks that can be proactively undertaken, new ongoing responsibilities and adjusted ongoing responsibilities.

[0080] Optionally, the task preparation module 330 is specifically used for: If the suggestion to be reviewed is a new suggestion and the review decision is "accepted", then a new ongoing responsibility shall be established based on the content of the suggestion to be reviewed. If the suggestion to be reviewed is an adjustment suggestion and the review decision is to accept it, then the corresponding continuing responsibilities will be adjusted according to the content of the suggestion to be reviewed.

[0081] The active task processing apparatus described in further detail can also execute the active task processing method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the method execution.

[0082] Figure 4 A schematic diagram of an electronic device 40 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0083] like Figure 4 As shown, the electronic device 40 includes at least one processor 41 and a memory, such as a read-only memory (ROM) 42 or a random access memory (RAM) 43, communicatively connected to the at least one processor 41. The memory stores computer programs executable by the at least one processor. The processor 41 can perform various appropriate actions and processes based on the computer program stored in the ROM 42 or loaded from storage unit 48 into the RAM 43. The RAM 43 may also store various programs and data required for the operation of the electronic device 40. The processor 41, ROM 42, and RAM 43 are interconnected via a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0084] Multiple components in electronic device 40 are connected to I / O interface 45, including: input unit 46, such as keyboard, mouse, etc.; output unit 47, such as various types of monitors, speakers, etc.; storage unit 48, such as disk, optical disk, etc.; and communication unit 49, such as network card, modem, wireless transceiver, etc. Communication unit 49 allows electronic device 40 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0085] Processor 41 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 41 performs the various methods and processes described above, such as active task processing methods.

[0086] In some embodiments, the active task processing method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 48. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 40 via ROM 42 and / or communication unit 49. When the computer program is loaded into RAM 43 and executed by processor 41, one or more steps of the active task processing method described above may be performed. Alternatively, in other embodiments, processor 41 may be configured to perform the active task processing method by any other suitable means (e.g., by means of firmware).

[0087] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0088] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0089] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0090] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0091] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0092] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0093] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0094] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for processing active tasks, characterized in that, The method includes: Proactive opportunities are created based on the current set of facts, ongoing responsibilities, pending review recommendations, and recent review decisions; the ongoing responsibilities indicate the work already undertaken by the system. The proactive opportunities are converted into suggestions for review, and the user's review decision on the suggestions for review is obtained; The persistent responsibilities are updated based on the proposed review and the review decision of the proposed review, and interrelated execution instances, execution qualifications, persistent execution carriers and runtime contexts are generated based on the updated persistent responsibilities. Based on the interrelated execution instances, execution qualifications, persistent execution carriers, and runtime contexts, the model in the intelligent agent system is invoked to execute tasks and obtain result reports.

2. The method according to claim 1, characterized in that, The runtime context includes a description of responsibilities, the executing entity, the result receiving entity, the associated execution instance, and the runtime strategy; The process of generating interrelated execution instances, execution qualifications, persistent execution carriers, and runtime contexts based on the updated persistent responsibilities includes: When the updated persistent responsibility meets the triggering conditions, a unique execution instance is generated or reused; The execution instance is granted a limited-term execution qualification through assignment; If the execution instance does not have an associated persistent execution carrier, then a persistent execution carrier for the execution instance is generated based on the execution context of the execution instance; If the execution instance has an associated persistent execution carrier, then the persistent execution carrier and its runtime context are reused.

3. The method according to claim 2, wherein the report type of the result report includes a completion report or a blocked report; after obtaining the result report by calling the model in the intelligent agent system to execute the task based on the interrelated execution instances, execution qualifications, persistent execution carriers, and runtime contexts, it further includes: Based on the report type of the result report, the existence of evidence in the content of the result report is verified; If the evidence existence verification passes, a work record indicating successful task execution or a record indicating failed task execution will be generated based on the result report. The fact update suggestion is generated based on the work record, or the responsibility adjustment suggestion is generated based on the obstructed record or the responsibility adjustment suggestion; the fact update suggestion is used to update the current fact set, and the responsibility adjustment suggestion is used to adjust the existing ongoing responsibilities.

4. The method according to claim 1, characterized in that, Before creating a proactive opportunity based on the current set of facts, ongoing responsibilities, pending review recommendations, and recent review decisions, the following is also included: The source records of external input are structured and organized to obtain the standardized record batch of the current time window; The historical fact set is updated based on the normalized record batches of the current time window to obtain the current fact set.

5. The method according to claim 4, characterized in that, The step of updating the historical fact set according to the normalized record batch of the current time window to obtain the current fact set includes: For each normalized record in the batch of normalized records, determine the stable objects in the normalized records; Based on the target fact type to which the stable object belongs, determine the appropriate content to be stored in each normalized record; wherein, the fact type includes facts, resources, decisions, or preferences; The historical fact set is updated based on the appropriate content to be preserved in each normalized record to obtain the current fact set.

6. The method according to claims 1-5, characterized in that, The proactive opportunities include opportunity type, brief title, specific delivery target, triggering basis, execution plan and evidence reference; the opportunity type includes tasks that can be proactively undertaken, new ongoing responsibilities, and adjustments to ongoing responsibilities.

7. The method according to claim 6, characterized in that, The update of continuing responsibilities based on the proposed review and the review decision of the proposed review includes: If the suggestion to be reviewed is a new suggestion and the review decision is "accepted", then a new ongoing responsibility shall be established based on the content of the suggestion to be reviewed. If the suggestion to be reviewed is an adjustment suggestion and the review decision is to accept it, then the corresponding continuing responsibilities will be adjusted according to the content of the suggestion to be reviewed.

8. A processing device for active tasks, characterized in that, The device includes: The opportunity creation module is used to create proactive opportunities based on the current set of facts, ongoing responsibilities, pending review recommendations, and recent review decisions; the ongoing responsibilities indicate the work already undertaken by the system. The suggestion review module is used to convert the proactive opportunity into a suggestion to be reviewed, and to obtain the user's review decision on the suggestion to be reviewed; The task preparation module is used to update the persistent responsibilities based on the proposed review and the review decision of the proposed review, and to generate interrelated execution instances, execution qualifications, persistent execution carriers and runtime contexts based on the updated persistent responsibilities. The task execution module is used to call the model in the intelligent agent system to execute tasks and obtain result reports based on the interrelated execution instances, execution qualifications, persistent execution carriers and running contexts.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and a memory communicatively connected to said at least one processor; The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the active task processing method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the method for processing the active task according to any one of claims 1-7.

Images