Method and system for collaborative allocation of mixed teaching tasks of art practice courses

Abstract

The present application relates to the technical field of teaching management, and discloses a kind of artistic practice course's mixed teaching task collaborative allocation method and system, the method includes: the multi-source task sequence and teaching state data of target process are collected, and task chain is constructed based on execution constraint and dependency relationship. Available resources and role state are combined into state snapshot, according to the demand of first task, candidate role set is filtered, the role with minimum load is selected as allocation target and is set as load balancing anchor point, and cascaded successor task obtains collaborative allocation sequence. Continuously receive the attendance signal and completion signal of offline classroom and online platform, and are associated and bound according to timestamp and corresponding task node, to generate execution event track. According to the state transition mark in track, update node state, form collaborative allocation closed loop record;The present application can improve the efficiency of artistic practice course's mixed teaching task collaborative allocation.

Description

Technical Field

[0001] This invention relates to the field of teaching management technology, and in particular to a method and system for collaborative allocation of blended teaching tasks in art practice courses. Background Technology

[0002] In the implementation of blended learning in existing art practice courses, the multi-source task data generated by offline classrooms and online platforms are independent of each other. There is a lack of unified association identifiers between task names, resource requirements, and student-submitted practice works, questions, and other behavioral records, making it impossible to accurately determine the execution order and dependencies of each task. At the same time, the reservation records of venue equipment and the interaction status of the teaching platform are stored in different systems, without forming a unified resource view. When a task needs to call upon both offline classrooms and online roles simultaneously, existing technology cannot comprehensively assess the availability of resources and the load of roles at the same time point, which can easily lead to resource conflicts or allocation imbalances.

[0003] Existing technologies typically employ simple polling or role-based allocation during the task allocation phase, failing to comprehensively consider factors such as the dependencies between preceding and subsequent tasks in the task chain, the number of currently assigned but incomplete tasks for each role, the total remaining time, and historical completion rates. During the execution tracking phase, the attendance signals from offline classes and the completion signals from online platforms cannot be automatically linked to the corresponding task nodes in the collaborative allocation sequence. This results in delayed or misaligned status updates, incomplete resource release records, and an inability to form a closed-loop record from task allocation to execution to resource recovery. Consequently, the overall traceability and load balancing effectiveness of the teaching process are difficult to guarantee. Summary of the Invention

[0004] This invention provides a method and system for collaborative allocation of blended teaching tasks in art practice courses, the main purpose of which is to solve the problem of low efficiency in collaborative allocation of blended teaching tasks in art practice courses.

[0005] To achieve the above objectives, the present invention provides a method for collaborative allocation of blended teaching tasks in art practice courses, comprising: Collect multi-source task sequences and teaching status data of the target process; The task chain of the target process is constructed based on the execution constraints of the multi-source task sequence and the dependencies between the preceding and subsequent tasks. The available resources and role status of the teaching status data are combined into a status snapshot, and the corresponding candidate role set is selected from the status snapshot based on the resource requirements of the first task in the task chain. The role with the smallest role load value in the candidate role set is taken as the allocation target, and the subsequent tasks of the task chain are cascaded with the allocation target as the load balancing anchor point to obtain the collaborative allocation sequence of the target process. Based on the execution progress of the collaborative allocation sequence, the system continuously receives the attendance and completion signals from offline classrooms and online platforms during the target process, and associates and binds the attendance and completion signals with the corresponding task nodes in the collaborative allocation sequence according to their timestamps to obtain the execution event trajectory of the target process. Based on the state transition markers in the execution event trajectory, update the execution state of the corresponding task node in the collaborative allocation sequence to obtain the collaborative allocation closed-loop record of the target process.

[0006] In a preferred embodiment, the acquisition of multi-source task sequences and teaching status data of the target process includes: The names of the practical tasks, the types of teaching resources required, and the estimated duration of the target process are combined to form the original task description set of the target process, and a corresponding task identifier is assigned to the original task description set. The student's submitted practice tasks, question messages, and work files are collected at fixed time intervals during the target process to obtain the student task trace of the target process. The student task trace is then timestamped with the task identifier to obtain the task description information of the target process. Extract reservation records from the venue equipment reservation data during the target process. The reservation records include venue identifier, start time, and end time. The interaction records and member list of the teaching platform during the target process are used as the platform interaction state set of the target process; The task description information is sorted by creation timestamp to obtain the multi-source task sequence of the target process, and the reservation record and the platform interaction status set are combined into the teaching status data of the target process.

[0007] In a preferred embodiment, constructing the task chain of the target process based on the execution constraints of the multi-source task sequence and the dependencies between preceding and subsequent tasks includes: Based on the structured markers of the task description field in the multi-source task sequence, the explicit preceding task identifiers in the multi-source task sequence are located, and the preceding task list of the target process is obtained. Based on the temporal relationship of task creation timestamps in the multi-source task sequence, the implicit predecessor task identifiers that the tasks with missing explicit predecessor task identifiers depend on are added to the predecessor task list. Using the current task in the multi-source task sequence as a node, add directed edges from the node of the preceding task to the node of the current task for the preceding tasks in the preceding task list, to obtain the directed graph of the target process. Based on the starting node of the directed graph, the successor node sequence is recursively extracted along the directed edges. The chain copy is split at the branch node of the successor node sequence until the end node of the successor node sequence, so as to obtain the original task chain of the target process. Merge the original task chains with the same prefix to obtain the task chain merge set of the target process, record branch information at the branch nodes of the task chain merge set, assign chain identifiers and node numbers to the task chain merge set, and obtain the task chain of the target process.

[0008] In a preferred embodiment, the step of combining the available resources and role status of the teaching status data into a status snapshot, and filtering the corresponding candidate role set from the status snapshot based on the resource requirements of the first task in the task chain, includes: Starting from the current moment of the target process, the occupied time slot sequence of the reservation records in the teaching status data is scanned in chronological order. The interval between adjacent occupied time slots in the occupied time slot sequence is taken as the free time slot. The start and end times of the free time slots are collected into a list of free time slots of the target process according to the location identifier. The virtual classroom identifier, number of online students, and activity level markers of the teaching status data are used as the classroom status list for the target process; Using the list of idle time slots and the list of classroom statuses as attribute values, and the current clock time of the target process as the generation timestamp, a state snapshot of the target process is constructed. Based on the type of art practice resources required for the first task in the task chain, the list of execution roles, and the skill level requirements, resources that match the elements and are compatible with the time period are located from the state snapshot to obtain the candidate role set for the target process.

[0009] In a preferred embodiment, the step of selecting the role with the smallest role load value in the candidate role set as the allocation target, and using the allocation target as the load balancing anchor point to cascade the subsequent tasks of the task chain to obtain the collaborative allocation sequence of the target process includes: The load value of the first role in the candidate role set is taken as the initial minimum value, and the first role is taken as the temporary target of the target process; By sequentially comparing the load values ​​of subsequent roles in the candidate role set with the initial minimum value, the temporary target is updated to obtain the first task allocation anchor point of the target process; The role corresponding to the allocation target is set as the load balancing anchor point, and the dynamic load adjustment value of the available role in the state snapshot during the target process is calculated based on the order of subsequent tasks in the task chain: The dynamic load adjustment value is compared with the load value of the load balancing anchor point, and the role with the smallest value that is not greater than the load value of the load balancing anchor point is selected as the subsequent task allocation target of the target process. The first task allocation anchor point and the subsequent task allocation targets are arranged in the order of the task chain to obtain the collaborative allocation sequence of the target process.

[0010] In a preferred embodiment, the formula for calculating the dynamic load adjustment value includes: in, The dynamic load adjustment value, The number of unfinished tasks currently assigned to the character. The time urgency sensitivity coefficient This represents the total remaining time for any unfinished tasks by the character. The total estimated duration of tasks assigned to the character. The historical completion rate impact coefficient. This represents the number of tasks the character has completed. The total number of tasks assigned to the character. This is the nonlinear amplification factor. This is the coupling adjustment coefficient.

[0011] In a preferred embodiment, the execution progress based on the collaborative allocation sequence continuously receives attendance and completion signals from offline classrooms and online platforms during the target process, and associates and binds the attendance and completion signals with corresponding task nodes in the collaborative allocation sequence according to timestamps to obtain the execution event trajectory of the target process, including: Assign a status identifier to the task nodes in the collaborative allocation sequence. The status identifier includes sleep, ready, and active. Set the initial state of the task node to sleep. Receive execution signals from the offline classroom and the online platform, wherein the execution signals include a type identifier, a source identifier, and a timestamp identifier; Based on the current state and resource requirements of the task nodes in the collaborative allocation sequence, the signal with the type identifier "start" is matched to the task node whose state identifier is "ready" and whose resource requirement matches the source identifier; the signal with the type identifier "end" is matched to the task node whose state identifier is "active" and whose resource usage corresponds to the source identifier, thus obtaining the signal node binding pair of the target process. Based on the interval positioning of the signal node binding pair and the timestamp identifier within the execution time window of the corresponding node, the signal node binding pair is converted into the state transition event tuple of the target process according to the type identifier; The execution event trajectory of the target process is obtained by sorting the state transition event tuples by timestamp.

[0012] In a preferred embodiment, the step of converting the signal node binding pair into a state transition event tuple of the target process based on the interval positioning of the signal node binding pair and the timestamp identifier within the corresponding node execution time window includes: Signals falling within the execution time window are established as temporary signal node binding pairs, and signals falling outside the execution time window are recorded as abnormal events of the target process. According to the type identifier, corresponding state transition instructions are generated respectively, wherein the signal with the type identifier of start type generates the activation node instruction of the target process, and the signal with the type identifier of end type generates the completion node instruction of the target process; The state transition instruction, the task node identifier, the timestamp identifier, and the source and target states of the state identifier are encapsulated into an event tuple for the target process.

[0013] In a preferred embodiment, updating the execution state of the corresponding task node in the collaborative allocation sequence based on the state transition markers in the execution event trajectory to obtain the collaborative allocation closed-loop record of the target process includes: The state of the corresponding node in the collaborative allocation sequence is updated according to the type of state transition instruction in the execution event trajectory, and the resources occupied by the corresponding node are released to obtain the resource release record of the target process; The final state of the nodes in the collaborative allocation sequence, the resource release record, and the signal node binding pair of the target process are merged to obtain the collaborative allocation closed-loop record of the target process.

[0014] To address the aforementioned problems, the present invention also provides a collaborative task allocation system for blended teaching in art practice courses, the system comprising: The data acquisition module collects multi-source task sequences and teaching status data for the target process; The task chain module constructs the task chain of the target process based on the execution constraints of the multi-source task sequence and the dependency relationship between the preceding and subsequent tasks. The candidate role set module combines the available resources and role status of the teaching status data into a status snapshot, and filters the corresponding candidate role set from the status snapshot based on the resource requirements of the first task in the task chain. The collaborative allocation sequence module takes the role with the smallest role load value in the candidate role set as the allocation target, and uses the allocation target as the load balancing anchor point to cascade the subsequent tasks of the task chain to obtain the collaborative allocation sequence of the target process. The execution event trajectory module continuously receives the attendance and completion signals of offline classrooms and online platforms during the target process based on the execution progress of the collaborative allocation sequence, and associates and binds the attendance and completion signals with the corresponding task nodes in the collaborative allocation sequence according to the timestamp to obtain the execution event trajectory of the target process. The collaborative allocation closed-loop record module updates the execution status of the corresponding task node in the collaborative allocation sequence based on the state transition markers in the execution event trajectory, thereby obtaining the collaborative allocation closed-loop record of the target process.

[0015] Compared with the prior art, the present invention has the following beneficial effects:

[0016] 1. This technical solution significantly improves the accuracy of task execution sequence identification and the comprehensiveness of resource matching by incorporating explicit and implicit pre-task identifiers from multi-source task sequences into dependency construction and combining idle periods in teaching status data with virtual classroom status into a status snapshot. Based on the resource requirements of the first task in the task chain, a candidate role set is selected from the status snapshot. Then, the role with the lowest load value is used as the load balancing anchor point to cascade subsequent tasks. This ensures that the role selection of each task node in the entire collaborative allocation sequence is constrained by the load level of the anchor point, thereby improving the balance of task allocation among multiple roles in a hybrid teaching scenario.

[0017] 2. This technical solution continuously receives attendance and completion signals from offline classrooms and online platforms, and associates and binds them with the corresponding task nodes in the collaborative allocation sequence according to timestamps, generating an execution event trajectory and achieving precise tracking of task state changes. Based on the state change markers in the execution event trajectory, the node state is updated and the occupied resources are released synchronously. Finally, the final node state, resource release record, and signal node binding are merged into a collaborative allocation closed-loop record, thus forming a complete data closed loop from task allocation to execution triggering to resource recovery, improving the traceability of the teaching process and the efficiency of resource utilization. Attached Figure Description

[0018] Figure 1 A flowchart illustrating a method for collaborative allocation of blended teaching tasks in an art practice course, as provided in an embodiment of the present invention.

[0019] Figure 2 A functional block diagram of a hybrid teaching task collaborative allocation system for an art practice course provided in an embodiment of the present invention;

[0020] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0022] This application provides a method for collaboratively allocating blended learning tasks in art practice courses. The executing entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method for collaboratively allocating blended learning tasks in art practice courses can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0023] Reference Figure 1 The diagram shown is a flowchart illustrating a method for collaboratively allocating blended teaching tasks in an art practice course according to an embodiment of the present invention. In this embodiment, the method for collaboratively allocating blended teaching tasks in an art practice course includes: In this embodiment of the invention, the process of collecting multi-source task sequences and teaching status data of the target process is specifically used for: The names of the practical tasks, the types of teaching resources required, and the estimated duration of the target process are combined to form the original task description set of the target process, and a corresponding task identifier is assigned to the original task description set. The student's submitted practice tasks, question messages, and work files are collected at fixed time intervals during the target process to obtain the student task trace of the target process. The student task trace is then timestamped with the task identifier to obtain the task description information of the target process. Extract reservation records from the venue equipment reservation data during the target process. The reservation records include venue identifier, start time, and end time. The interaction records and member list of the teaching platform during the target process are used as the platform interaction state set of the target process; The task description information is sorted by creation timestamp to obtain the multi-source task sequence of the target process, and the reservation record and the platform interaction status set are combined into the teaching status data of the target process.

[0024] Specifically, the name of each practical task is extracted one by one from the course syllabus of the target process or the task list pre-set by the teacher. At the same time, all types of teaching resources required for the execution of the task are obtained, including canvas, oil paints, palette, easel and multimedia presentation equipment. The estimated duration of the task from start to finish is recorded and taken as an integer value in minutes.

[0025] Specifically, throughout the entire teaching cycle, a data collection operation is performed at fixed intervals, covering scanned copies of paper submissions from offline classes, file upload directories on the online teaching platform, and message records in instant messaging groups. From the collected data, three types of student activity records are selected: completed exercise task files, text messages of questions sent by students, and uploaded work files by students. These three types of records are collectively referred to as the student task traces.

[0026] Specifically, all reservation data is read from the classroom management system, laboratory reservation platform, or equipment borrowing registration form associated with the target process. This data is usually stored in tabular form, with each row representing a reservation operation. For each row of reservation data, three core fields are extracted: the first field is the venue identifier, which is a unique number pre-assigned to each art studio, each pottery wheel, or each virtual classroom and other scenarios; the second field is the start time, recorded as the specific year, month, day, hour, and minute, indicating the start time when the venue or equipment is reserved; the third field is the end time, also recorded as the specific year, month, day, hour, and minute, indicating the end time of the reservation.

[0027] Specifically, log in to the online teaching platform used in the target process and export two types of data from the platform's backend management interface. The first type of data is interaction records, including text messages in all chat rooms on the platform, the posting and replying of forum posts, the exchange of homework annotations, and the submission results of voting questionnaires. Each interaction record is accompanied by the sender's identifier and the sending time. The second type of data is a member list, which includes the list of all students currently enrolled in the course, the list of teaching assistants, and the list of instructors. Each member's information includes a unique user identifier, role type, and online status.

[0028] Specifically, the task description information generated in the previous step is retrieved. This information contains the creation timestamp of each student task and the corresponding task identifier. Based on the value of the creation timestamp, all task description information entries are sorted in ascending order, with the earliest timestamp at the beginning and the latest timestamp at the end. If two task description information entries have the same timestamp during the sorting process, their order is determined by the lexicographical order of the task identifiers. The ordered list obtained after sorting is called the multi-source task sequence, where each element is a task description information entry with a timestamp and an identifier.

[0029] Furthermore, the above three items—the name of the practical task, the list of required teaching resource types, and the estimated duration—are concatenated in a fixed order to form a complete descriptive text, which is called the original task description set. Each task corresponds to an independent descriptive text. Subsequently, a globally unique task identifier is generated for each original task description set. This identifier is formed by combining an incrementing number with the first letter of the task name, thereby ensuring that the student task traces collected later can be accurately associated with the original task through this identifier.

[0030] Furthermore, for each student task trace, the system timestamp at the time of its generation and the practice task name implicitly referenced in its content are extracted. Then, the timestamp is compared one by one with the previously assigned task identifiers to find the task identifiers whose task names match and whose timestamps are within the expected duration range of the task. The identifiers are then appended to the corresponding student task traces. After the above alignment operation, each student task trace is accompanied by the correct task identifier and the original timestamp. These student task traces with attached identifiers and timestamps together constitute the task description information.

[0031] Furthermore, after extracting the above three fields completely from the original reservation record, without performing any format conversion or calculation, they are directly combined into a structured reservation record entry according to the order of venue identifier, start time, and end time; all the extracted reservation record entries are gathered together in chronological order as one of the original materials for subsequently constructing teaching status data.

[0032] Furthermore, the aforementioned interaction records and member lists are packaged directly into a data set without any filtering or aggregation. This set is referred to as the platform interaction status set. The interaction records in this set reflect the real-time communication between teachers and students during the teaching process, while the member list provides basic profiles of task executors that can be assigned. Together, they constitute all the information related to online activities in the teaching status data.

[0033] Furthermore, all previously extracted reservation record entries are simultaneously merged with the platform interaction status set. The merging method involves placing the reservation record list and the platform interaction status set as two parallel components in the same data structure without cross-calculation or deduplication. This merged overall data structure is referred to as the teaching status data. Thus, the multi-source task sequence records the order in which tasks appear over time, and the teaching status data records the usage arrangements of venue equipment and the real-time status of the online platform. Both provide a complete data foundation for subsequent task chain construction and role selection.

[0034] In summary, by standardizing the core elements of practical tasks into structured descriptions and assigning a unique identifier to each task, a clear reference benchmark is provided for the subsequent collection and alignment of task traces. This avoids data confusion caused by ambiguity in task names or inconsistencies in resource descriptions and ensures that multi-source data can be accurately associated with the correct task entities.

[0035] In summary, by automatically collecting various behavioral data generated by students in online and offline activities at regular intervals, and aligning these scattered student task traces with pre-assigned task identifiers using timestamps, dynamic capture and attribution mapping of student task execution processes are achieved. This transforms the originally isolated student activity records into descriptive information with task semantics, providing a data foundation for the subsequent construction of multi-source task sequences that reflect the actual task execution order.

[0036] In summary, by directly obtaining structured reservation records from the venue and equipment management system and retaining the venue identifier, start time, and end time as key fields, the system accurately reflects the time occupancy of physical resources such as offline classrooms and equipment. This provides precise resource availability data for the subsequent generation of idle time slot lists and status snapshots, avoiding task allocation failures caused by resource conflicts.

[0037] In summary, all interactive messages generated in the online teaching platform are packaged with the profiles of currently online members into a unified state set, which fully records the real-time communication status and available human resources in the virtual classroom. This allows subsequent task allocation to fully consider the online status and activity level of online roles, thereby supporting the collaborative scheduling of offline and online resources in a blended teaching scenario.

[0038] In summary, by sorting the task description information by time, the disordered task traces are transformed into a multi-source task sequence arranged chronologically, which truly reflects the emergence order of tasks in the actual teaching process. At the same time, the appointment records and platform interaction status are aggregated into unified teaching status data, providing a complete resource view that includes both offline venue time periods and online classroom status for task chain construction and role selection, laying the data foundation for load balancing distribution.

[0039] In this embodiment of the invention, the construction of the task chain for the target process based on the execution constraints of the multi-source task sequence and the dependencies between preceding and subsequent tasks is specifically used for: Based on the structured markers of the task description field in the multi-source task sequence, the explicit preceding task identifiers in the multi-source task sequence are located, and the preceding task list of the target process is obtained. Based on the temporal relationship of task creation timestamps in the multi-source task sequence, the implicit predecessor task identifiers that the tasks with missing explicit predecessor task identifiers depend on are added to the predecessor task list. Using the current task in the multi-source task sequence as a node, add directed edges from the node of the preceding task to the node of the current task for the preceding tasks in the preceding task list, to obtain the directed graph of the target process. Based on the starting node of the directed graph, the successor node sequence is recursively extracted along the directed edges. The chain copy is split at the branch node of the successor node sequence until the end node of the successor node sequence, so as to obtain the original task chain of the target process. Merge the original task chains with the same prefix to obtain the task chain merge set of the target process, record branch information at the branch nodes of the task chain merge set, assign chain identifiers and node numbers to the task chain merge set, and obtain the task chain of the target process.

[0040] Specifically, each task description in the multi-source task sequence is traversed, and pre-agreed structured tags are searched from the task description field of that information. These structured tags begin with a specific prefix and are used to explicitly declare which other tasks must be completed before the current task can be executed.

[0041] Specifically, the creation timestamps of all task description information in the multi-source task sequence are scanned. Tasks that already contain explicit preceding task identifiers in the preceding task list are not processed further. For tasks lacking explicit preceding task identifiers, the order of their creation timestamps with the creation timestamps of other tasks in the multi-source task sequence is checked.

[0042] Specifically, each task in the multi-source task sequence is considered as a node in a directed graph, and the name of the node is the task identifier of that task. Each dependency record in the list of preceding tasks is extracted; this record contains a preceding task identifier and a current task identifier. Starting from the node representing the preceding task, a directed edge is added pointing to the node representing the current task.

[0043] Specifically, identify all nodes with an in-degree of zero in the directed graph. These nodes have no preceding tasks pointing to them and are therefore the starting nodes of the directed graph. Starting from each starting node, recursively visit its successor nodes along the direction of the directed edges.

[0044] Specifically, all the original task chains are extracted, and their prefixes are compared. Starting from the first node of the first original task chain, each node is compared with the corresponding node of the next original task chain. If two original task chains have several identical nodes at the beginning, these identical nodes are merged into a shared prefix, while the subsequent differences are retained as branches. The merge operation is performed pairwise on all original task chains until there are no more prefixes that can be merged. The resulting data structure is called the task chain merge set.

[0045] Furthermore, for each task description, if its task description field contains such an explicit preceding task tag, all task identifiers referenced in the tag are parsed out, and these identifiers are associated with the identifier of the current task itself, forming a preceding relationship entry. All entries are then aggregated to obtain the preceding task list, where each item indicates an explicitly declared preceding task that a particular task depends on.

[0046] Furthermore, if the timestamp of one task is later than that of another task, and there is a natural sequential dependency between the two tasks in terms of teaching content, the task identifier of the earlier task is used as an implicit predecessor task identifier and added to the predecessor task list, forming a dependency record from the earlier task to the later task. This process iterates through all tasks, ensuring that all predecessor dependencies of each task are completely recorded in the predecessor task list.

[0047] Furthermore, the above operation is repeated for all dependent records. After all edges have been added, a data structure consisting of nodes and directed edges is obtained, which is called the directed graph. There are no isolated nodes in this directed graph because each task appears to the left or right of at least one dependent record, and the directions of the edges strictly follow the order in which the tasks are executed.

[0048] Furthermore, during the recursive process, whenever a node with multiple outgoing edges is encountered, the current path is copied at that branch point with the same number of replicas as the number of successor nodes. Each replica continues recursively down the path along a successor node. This recursion continues until a terminal node with no outgoing edges is reached. Each complete path from the starting node to the terminal node, along with all nodes along the path and the replica splitting at each branch point, is recorded as a primary task chain. All primary task chains generated from the starting node together constitute the set of primary task chains.

[0049] Furthermore, at each branch node in the task chain merge set, the identifiers of all subsequent nodes branching off from that node and the conditions under which the branching occurred are recorded. A unique chain identifier is assigned to each complete path in the task chain merge set, and an incremental node number is assigned to each node on that path in order from the start to the end. After the above assignments are completed, each path in the task chain merge set has a chain identifier and a node number; this data structure is then referred to as the task chain.

[0050] In summary, by parsing the pre-defined structured tags in the task description field, the system automatically identifies the explicitly declared dependencies between tasks, avoiding the tediousness and errors of manually sorting out task dependencies, and providing a reliable source of dependency edges for building accurate task chains in the future.

[0051] In summary, the system automatically completes implicit dependencies that do not explicitly declare prerequisite tasks but objectively have temporal dependencies by using the order of timestamps. This ensures that the prerequisite task list covers all necessary inter-task constraints and prevents incorrect task execution order due to missing dependencies.

[0052] In summary, the dependencies between tasks are transformed into edges in a directed graph, which intuitively and completely expresses the predecessor and successor logic of task execution in a graph structure, providing a computer-processable formal model for subsequent recursive extraction of task chains.

[0053] In summary, by recursively traversing the directed graph and splitting copies at branch nodes, all possible task execution paths are generated exhaustively, ensuring that no original task chain that conforms to the dependency relationship is missed. This provides complete basic material for merging tasks with the same prefix and forming standardized task chains in the future.

[0054] In summary, merging original task chains with the same prefix into a compact representation with a branching structure significantly reduces data redundancy in the task chains. At the same time, by recording branch information, assigning chain identifiers, and node numbers, each path has the ability to be uniquely addressed, which facilitates subsequent sequential role allocation and load balancing.

[0055] In this embodiment of the invention, the step of combining the available resources and role status of the teaching status data into a status snapshot, and filtering the corresponding candidate role set from the status snapshot based on the resource requirements of the first task in the task chain, is specifically used for: Starting from the current moment of the target process, the occupied time slot sequence of the reservation records in the teaching status data is scanned in chronological order. The interval between adjacent occupied time slots in the occupied time slot sequence is taken as the free time slot. The start and end times of the free time slots are collected into a list of free time slots of the target process according to the location identifier. The virtual classroom identifier, number of online students, and activity level markers of the teaching status data are used as the classroom status list for the target process; Using the list of idle time slots and the list of classroom statuses as attribute values, and the current clock time of the target process as the generation timestamp, a state snapshot of the target process is constructed. Based on the type of art practice resources required for the first task in the task chain, the list of execution roles, and the skill level requirements, resources that match the elements and are compatible with the time period are located from the state snapshot to obtain the candidate role set for the target process.

[0056] Specifically, all reservation records are extracted from the teaching status data. Each reservation record includes a location identifier, start time, and end time. These reservation records are arranged in chronological order of their start times to form a sequence of occupied time slots. The current time of the target process is obtained as the scanning starting point, and each occupied time slot in the sequence is traversed in chronological order from earliest to latest. During the traversal, the time interval between the end time of the previous occupied time slot and the start time of the current occupied time slot is recorded. If this interval is greater than zero, it is defined as an idle time slot, with the start time of the idle time slot being the end time of the previous occupied time slot and the end time being the start time of the current occupied time slot. For time slots prior to the first occupied time slot, the interval from the current time to the start time of the first occupied time slot is also considered an idle time slot.

[0057] Specifically, the platform interaction status set, which contains real-time operational data of the online teaching platform, is read from the teaching status data. Virtual classroom identifiers, unique numbers for each online classroom, are extracted from the platform interaction status set. The number of students currently online is counted from the member list of the platform interaction status set, yielding a specific numerical value for the number of online students.

[0058] Specifically, an empty data structure is created as a container for the state snapshot. The list of idle time slots generated in the previous step is stored in this container as an attribute value, named "Idle Time Slot List". The list of classroom statuses generated in the previous step is also stored in this container as another attribute value, named "Classroom Status List". The system clock reading at the current moment of the target process is obtained, accurate to the second, and this reading is stored in the container as a generation timestamp, named "Generation Timestamp".

[0059] Specifically, the first task, or the initial task, is retrieved from the task chain, and its resource requirement description is read. This resource requirement description contains three parts: the first part is the type of art practice resource; the second part is the list of execution roles; and the third part is the skill level requirement. The list of available time slots in the state snapshot is traversed, and for each available time slot under each venue identifier, it is checked whether the type of art practice resource that the venue can provide completely matches the type of art practice resource required by the initial task.

[0060] Furthermore, for the time period following the last occupied time period, the interval between its end time and the end time of the day's teaching is also considered as a free time period. The start and end times of each extracted free time period are categorized and aggregated according to their respective venue identifiers; that is, all free time periods under the same venue identifier are grouped together. The resulting data structure is called the free time period list, which records all available free time periods after the current time for each venue, grouped by venue identifier.

[0061] Furthermore, the message sending frequency and number of active users within the most recent five minutes are calculated from the interaction records of the platform's interaction state set, and the activity level is marked into different levels according to a preset threshold. The extracted virtual classroom identifier, number of online students, and activity level markers are combined into an entry in a fixed order, with each virtual classroom corresponding to an independent entry. The data structure formed by all entries is called the classroom status list. This list reflects the real-time availability of online teaching resources during the target process.

[0062] Furthermore, the three attribute values ​​mentioned above are encapsulated in the same data structure, which is the state snapshot. The state snapshot fully records the distribution of all available resources in the target process at the time of the generated timestamp, including idle time information of offline venues and real-time status information of online classrooms, providing a unified resource view for subsequent candidate role selection.

[0063] Furthermore, the classroom status list in the state snapshot is traversed simultaneously. For each virtual classroom, it is checked whether the role types that the classroom can support are included in the list of execution roles. For each resource entry that satisfies both resource type matching and role type matching, the time slot compatibility of the resource within the expected duration of the first task is further checked; that is, the execution time of the first task must fall entirely within the idle time slot of the resource or the available time window of the virtual classroom. All resource entries that pass the above element matching and time slot compatibility checks are collected, and the corresponding executable roles are collected to form a data structure called the candidate role set.

[0064] In summary, by scanning the occupied time slots in the reservation records and extracting the intervals between adjacent occupancy, a structured list of available time slots grouped by location is automatically generated. This provides a precise offline resource availability window for subsequent task allocation, avoiding allocation failures caused by resource time conflicts.

[0065] In summary, integrating the virtual classroom identifiers, real-time online student counts, and activity markers of online teaching platforms into a unified classroom status list allows for the quantification of online resource availability and interaction intensity, providing data support for selecting online roles with sufficient participation in blended learning scenarios.

[0066] In summary, by encapsulating the idle time of offline venues with the real-time status of online classrooms and the current moment into an immutable state snapshot, the consistency and integrity of the resource view on which subsequent role selection is based at any given point in time are ensured, preventing allocation deviations due to dynamic changes in status.

[0067] In summary, by using the resource requirements of the first task as a filter, resource type matching, role type matching, and time period compatibility checks are performed simultaneously in the status snapshot, accurately selecting all candidate roles that can undertake the first task, and providing a limited and reasonable range of candidate roles for subsequent load balancing.

[0068] In this embodiment of the invention, the step of selecting the role with the smallest role load value in the candidate role set as the allocation target, and using the allocation target as the load balancing anchor point to cascade the subsequent tasks of the task chain to obtain the collaborative allocation sequence of the target process, is specifically used for: The load value of the first role in the candidate role set is taken as the initial minimum value, and the first role is taken as the temporary target of the target process; By sequentially comparing the load values ​​of subsequent roles in the candidate role set with the initial minimum value, the temporary target is updated to obtain the first task allocation anchor point of the target process; The role corresponding to the allocation target is set as the load balancing anchor point, and the dynamic load adjustment value of the available role in the state snapshot during the target process is calculated based on the order of subsequent tasks in the task chain: The dynamic load adjustment value is compared with the load value of the load balancing anchor point, and the role with the smallest value that is not greater than the load value of the load balancing anchor point is selected as the subsequent task allocation target of the target process. The first task allocation anchor point and the subsequent task allocation targets are arranged in the order of the task chain to obtain the collaborative allocation sequence of the target process.

[0069] Specifically, the role ranked first in the candidate role set is selected, and the load value of the role is read. The load value represents the number of tasks that the role has been assigned but not yet completed. This value is recorded as the initial minimum value, and the first role is recorded as the temporary target of the target process.

[0070] Specifically, iterate through all subsequent roles starting from the second role in the candidate role set. For each subsequent role, read its load value and compare it with the currently recorded initial minimum value. If the load value of the subsequent role is less than the initial minimum value, update the initial minimum value to the load value of the subsequent role and update the temporary target to the subsequent role. If the load value of the subsequent role is greater than or equal to the initial minimum value, keep the initial minimum value and the temporary target unchanged. After completing the comparison of all subsequent roles, the role corresponding to the currently recorded temporary target is the first task allocation anchor point of the target process.

[0071] Specifically, the role corresponding to the first task allocation anchor point is set as the load balancing anchor point. The load value of the load balancing anchor point is used as the comparison benchmark for subsequent task allocation. All subsequent tasks after the first task are taken out from the task chain and processed one by one according to their order of appearance in the task chain. For each subsequent task, all available roles in the state snapshot are traversed and the dynamic load adjustment value of each available role is calculated.

[0072] Specifically, for each subsequent task, the dynamic load adjustment values ​​of all available roles calculated in the previous step are compared one by one with the load value of the load balancing anchor point, and all roles whose dynamic load adjustment values ​​are less than or equal to the load value of the load balancing anchor point are selected.

[0073] Specifically, the first task allocation anchor point is used as the first element of the collaborative allocation sequence, and then the subsequent task allocation target corresponding to each subsequent task is added to the collaborative allocation sequence in sequence according to the order of the subsequent tasks in the task chain.

[0074] Furthermore, the calculation process for this dynamic load adjustment value is as follows: First, obtain the number of currently assigned but unfinished tasks for the role; then, obtain the sum of the remaining time for all unfinished tasks for the role and the total estimated time for assigned tasks; simultaneously, obtain the number of completed tasks for the role and the total number of assigned tasks; and then perform comprehensive calculations on these values ​​according to preset time urgency sensitivity coefficients, historical completion rate influence coefficients, nonlinear urgency amplification coefficients, and coupling adjustment coefficients to finally obtain a comprehensive load assessment value representing the role's performance under the current state after undertaking new tasks.

[0075] Furthermore, the role with the smallest dynamic load adjustment value is selected from these selected roles as the allocation target for the subsequent task. If there is no role with a dynamic load adjustment value less than or equal to the load balancer anchor value, the role with the smallest dynamic load adjustment value is selected as the allocation target.

[0076] Furthermore, the first task assignment anchor point and each subsequent task assignment target are arranged into an ordered list according to the execution order of the task chain. This list is the collaborative assignment sequence of the target process, and each element in the sequence specifies which role should execute the corresponding task in the task chain.

[0077] In summary, by setting an initial minimum value and a temporary target, a clear benchmark is established for subsequent traversal comparisons, enabling the selection process for the role with the minimum load value to be executed in an orderly manner starting from the first role in the candidate role set, thus avoiding comparison chaos caused by the lack of a starting reference.

[0078] In summary, by linearly traversing and comparing the load values ​​of all roles in the candidate role set, the role with the smallest load value is finally locked as the anchor point for the first task assignment. This ensures that the first task is assigned to the role with the lightest current burden, achieving load optimization for the first task assignment and laying the foundation for load balancing of subsequent tasks.

[0079] In summary, the allocation result of the first task is used as the load balancing anchor point, so that the allocation of subsequent tasks always refers to the load level of this anchor point. At the same time, by calculating dynamic load adjustment values, the current burden, time urgency and historical completion rate of each available role are comprehensively evaluated, so that the allocation of subsequent tasks takes into account both real-time status and long-term performance.

[0080] In summary, by limiting the allocation target of each subsequent task to a range where the dynamic load adjustment value does not exceed the load balancing anchor point, the load balance among roles is maintained throughout the entire task chain execution, preventing subsequent tasks from causing any role's load to exceed the load level of the first task's anchor point.

[0081] In summary, the allocation results of the first task and all subsequent tasks are combined into an ordered sequence according to the execution order of the task chain. Each task node is assigned a clear responsibility role, so that the subsequent execution signal reception, status update and resource release can be carried out in an orderly manner according to this sequence, forming a traceable allocation link from allocation to execution to closed-loop recording.

[0082] In this embodiment of the invention, the calculation formula for the dynamic load adjustment value is specifically used for: in, The dynamic load adjustment value, The number of unfinished tasks currently assigned to the character. The time urgency sensitivity coefficient This represents the total remaining time for any unfinished tasks by the character. The total estimated duration of tasks assigned to the character. The historical completion rate impact coefficient. This represents the number of tasks the character has completed. The total number of tasks assigned to the character. This is the nonlinear amplification factor. This is the coupling adjustment coefficient.

[0083] Specifically, the number of currently assigned but incomplete tasks for a role comes from the cumulative number of tasks assigned to that role during the collaborative allocation sequence generation process that have not yet received a completion signal. This value increases each time a new task is assigned and decreases each time a task completion signal is received. The total remaining time of a role's incomplete tasks is obtained by calculating and adding the remaining execution time of each of the role's currently incomplete tasks. The remaining time of each incomplete task is equal to the estimated total time of that task minus the time elapsed from the start time to the current time. The total estimated time of a role's assigned tasks is obtained by summing the estimated times of all tasks assigned to that role, regardless of whether these tasks have been completed. The number of tasks a role has completed is obtained from the execution event trajectory by counting the number of task nodes for which the role is the allocation target and whose state transitions to completed. The total number of tasks a role has been assigned is obtained from the collaborative allocation sequence by counting the number of all task nodes for which the role is designated as the allocation target. The time urgency sensitivity coefficient, historical completion rate influence coefficient, nonlinear urgency amplification coefficient, and coupling adjustment coefficient are all fixed configuration values ​​preset by the teaching administrator according to the characteristics of the course before the start of the target process. Among them, the time urgency sensitivity coefficient controls the influence intensity of the remaining time percentage on the load adjustment, the historical completion rate influence coefficient controls the influence intensity of the completed task percentage on the load adjustment, the nonlinear urgency amplification coefficient is used to enhance the secondary influence of the remaining time percentage, and the coupling adjustment coefficient is used to balance the interactive influence between the completion rate and urgency.

[0084] Furthermore, the overall meaning of this formula is to calculate a comprehensive load assessment value for a role when undertaking new tasks. This value is used to select the optimal allocation target from the candidate role set. The formula first uses the number of currently assigned but uncompleted tasks as the base load value, which directly reflects the number of tasks the role is currently handling. Based on this, the formula introduces a remaining time percentage factor. This factor is calculated by dividing the sum of the remaining time of the role's uncompleted tasks by the total estimated time of the role's assigned tasks. It reflects the time urgency of the role's current backlog of tasks; a larger remaining time percentage indicates greater time pressure on the role, requiring an increase in the load adjustment value to avoid being assigned new tasks. The formula also introduces a historical completion rate factor. This factor measures the role's execution efficiency by calculating the ratio of the number of completed tasks to the total number of tasks assigned to the role. A lower completion rate indicates insufficient execution capacity, requiring an increase in the load adjustment value to limit the role from receiving new tasks. The formula also introduces a squared term for the remaining time percentage. This squared term amplifies larger remaining time percentages, giving roles with particularly high time pressure a higher load adjustment value, making them less likely to be selected as allocation targets for new tasks. Finally, the formula introduces a coupling adjustment term, which multiplies the historical completion rate by the percentage of remaining time. This term is used to capture the synergistic effect between the two factors. When a role has both a low completion rate and a high percentage of remaining time, the coupling term will generate a large positive value, further increasing the load adjustment value and ensuring that the role will not be assigned new tasks.

[0085] In general, as the number of currently assigned but unfinished tasks increases, the dynamic load adjustment value exhibits a linear growth trend. This means that characters with a large number of unfinished tasks receive higher load adjustment values, reducing their probability of being selected when assigning new tasks. As the ratio of the total remaining time of a character's unfinished tasks to the total estimated time of their assigned tasks increases, the dynamic load adjustment value shows an accelerated growth trend. This is because the formula includes both a linear and a squared term for this ratio. When the ratio is small, the squared term has a weak effect; when the ratio exceeds half, the squared term begins to significantly increase the load adjustment value. This accelerated growth mechanism ensures that characters under high time pressure will not continue to receive new tasks. As the ratio of the number of completed tasks to the total number of assigned tasks increases, the dynamic load adjustment value shows a downward trend. This is because a higher historical completion rate indicates stronger performance capabilities. The historical completion rate factor in the formula exists in a negative correlation; characters with high completion rates receive lower load adjustment values, thus making them more likely to be selected as targets for new task assignments. When a role simultaneously has a low completion rate and a high percentage of remaining time, the coupling adjustment term will generate a large positive increment, causing the dynamic load adjustment value to jump. This coupling trend prevents roles with both unfavorable factors from being over-assigned tasks. When a role currently has zero assigned unfinished tasks, the dynamic load adjustment value is entirely determined by the percentage of remaining time, the historical completion rate, and their coupling terms. In this case, even if a role has a historically low completion rate, as long as it currently has no unfinished tasks, the load adjustment value may still remain at a low level, allowing the role to receive new tasks.

[0086] In this embodiment of the invention, when the execution progress based on the collaborative allocation sequence continuously receives the attendance and completion signals from offline classrooms and online platforms during the target process, and associates and binds the attendance and completion signals with the corresponding task nodes in the collaborative allocation sequence according to their timestamps to obtain the execution event trajectory of the target process, it is specifically used for: Assign a status identifier to the task nodes in the collaborative allocation sequence. The status identifier includes sleep, ready, and active. Set the initial state of the task node to sleep. Receive execution signals from the offline classroom and the online platform, wherein the execution signals include a type identifier, a source identifier, and a timestamp identifier; Based on the current state and resource requirements of the task nodes in the collaborative allocation sequence, the signal with the type identifier "start" is matched to the task node whose state identifier is "ready" and whose resource requirement matches the source identifier; the signal with the type identifier "end" is matched to the task node whose state identifier is "active" and whose resource usage corresponds to the source identifier, thus obtaining the signal node binding pair of the target process. Based on the interval positioning of the signal node binding pair and the timestamp identifier within the execution time window of the corresponding node, the signal node binding pair is converted into the state transition event tuple of the target process according to the type identifier; The execution event trajectory of the target process is obtained by sorting the state transition event tuples by timestamp.

[0087] Specifically, each task node in the collaborative allocation sequence is traversed, and a status identifier field is assigned to each task node. This field can take one of three states: dormant, ready, or active. The allocation rules are as follows: all task nodes are initially set to a dormant state, indicating that the task is not yet ready for execution. When all the prerequisite tasks for a task node have been completed and the resources required by the task node are ready, the status identifier of the task node is changed from dormant to ready.

[0088] Specifically, the system continuously monitors signals from the hardware devices in the offline classroom and message pushes from the online platform. Signal sources from the offline classroom include attendance terminals, device on / off sensors, and start / end buttons manually triggered by the teacher. Signal sources from the online platform include student-side task start clicks, work submission buttons, and teaching assistant task completion confirmations.

[0089] Specifically, each execution signal is retrieved from the signal queue, and different matching operations are performed according to its type identifier. For signals with the type identifier "Start", all task nodes with the status identifier "Ready" in the cooperative allocation sequence are traversed, and it is checked whether the source identifier list in the resource requirement field of the task node contains the source identifier of the current signal. If it does, a one-to-one binding relationship is established between the signal and the task node, forming a signal node binding pair.

[0090] Specifically, for each binding relationship in the signal node binding pair, the execution time window of the task node in the binding pair is obtained. This execution time window consists of the expected start time and expected end time of the task node. The timestamp identifier of the signal in the binding pair is obtained, and it is determined whether the timestamp falls within the execution time window of the corresponding node. For start-type signals, their timestamp must fall after the start time and before the end time of the execution time window; for end-type signals, their timestamp must also fall after the start time and before the end time of the execution time window.

[0091] Specifically, all the state transition event tuples obtained in the previous step are sorted in ascending order according to the value of their timestamp identifiers, with the event tuples with earlier timestamps being placed at the front, and event tuples with the same timestamps arranged in lexicographical order according to the task node identifiers.

[0092] Furthermore, when a task node in the ready state is formally assigned to a role to begin execution, the status flag of that task node is changed from ready to active. After the above allocation is completed, the initial status flag of all task nodes in the collaborative allocation sequence is uniformly set to dormant, at which point each task node carries an initial dormant flag.

[0093] Furthermore, each received execution signal contains three essential fields: a type identifier to distinguish whether the signal is a start type or an end type (start type indicates the task has begun execution, end type indicates the task has been completed); a source identifier to indicate the unique code of the issuing device or account, such as a student's user identifier or a device's serial number; and a timestamp to record the moment the signal was received by the system, accurate to the second. All received execution signals are temporarily stored in a signal queue in chronological order of receipt.

[0094] Furthermore, for signals with the type identifier "End", all task nodes with the status identifier "Active" in the collaborative allocation sequence are traversed. The source identifier in the resource usage field of each task node is checked to see if it is completely consistent with the source identifier of the current signal. If they are consistent, a one-to-one binding relationship is established between the signal and the task node, forming a signal node binding pair. All binding relationships are aggregated to obtain the signal node binding pair set, which records which task node each execution signal is associated with.

[0095] Furthermore, signal node binding pairs falling within the execution time window are retained. For start-type signals, they are converted into a state transition event tuple containing four elements: the transition instruction is the activation node instruction, the task node identifier is the task node identifier in the binding pair, the timestamp is the signal's timestamp, and the state transition description is a transition from the ready state to the active state. For end-type signals, they are converted into a state transition event tuple containing four elements: the transition instruction is the completion node instruction, the task node identifier is the task node identifier in the binding pair, the timestamp is the signal's timestamp, and the state transition description is a transition from the active state to the completed state. All the converted state transition event tuples form a temporary set.

[0096] Furthermore, after sorting, an ordered sequence of event tuples is obtained, which is called the execution event trajectory. The execution event trajectory records the entire process of each task node in the target process from ready to activated to completed in chronological order. Each transition is bound to a specific execution signal, forming a traceable event chain.

[0097] In summary, by assigning a clear state identifier to each task node and uniformly starting from the dormant state, an observable state machine model of task execution progress was established. This enables the system to accurately distinguish between three types of task nodes: those that have not yet met the execution conditions, those that are ready to be executed, and those that are currently executing. This provides a state filtering basis for the accurate matching of subsequent execution signals.

[0098] In summary, by uniformly receiving execution signals pushed by offline hardware devices and online platforms, and fully recording the type identifier, source identifier, and timestamp identifier attached to each signal, the system achieves real-time capture of task start and end events in the blended learning scenario, providing standardized data input for subsequently establishing associations between discrete signals and specific task nodes.

[0099] In summary, by handling the matching logic of start and end signals separately through dual condition constraints, the start signal is matched only with task nodes that are in the ready state and have matching resource requirements, and the end signal is matched only with task nodes that are in the active state and have corresponding resource occupancy. This effectively avoids signal mismatch or duplicate matching and ensures that each execution signal uniquely corresponds to a correct task node.

[0100] In summary, by performing interval positioning verification between the timestamp of the signal and the execution time window of the task node, abnormal signals outside the time window are filtered out. The start signal that passes the verification is converted into an activation node instruction and the end signal is converted into a completion node instruction. At the same time, the source state and target state of the transition are encapsulated to form a standardized state transition event tuple, which provides a unified format recording unit for subsequent generation of event trajectories by time sorting.

[0101] In summary, by arranging all state transition event tuples in chronological order according to their timestamps, a complete timeline event sequence is generated. This sequence faithfully records the state transition sequence of each task node in the target process from ready to activated to completed, enabling the entire execution process of the teaching task to be traceable and replayable. It provides an orderly event-driven basis for subsequently updating the node states in the collaborative allocation sequence and generating collaborative allocation closed-loop records.

[0102] In this embodiment of the invention, when the signal node binding pair is converted into a state transition event tuple of the target process based on the interval positioning of the signal node binding pair and the timestamp identifier within the execution time window of the corresponding node, it is specifically used for: Signals falling within the execution time window are established as temporary signal node binding pairs, and signals falling outside the execution time window are recorded as abnormal events of the target process. According to the type identifier, corresponding state transition instructions are generated respectively, wherein the signal with the type identifier of start type generates the activation node instruction of the target process, and the signal with the type identifier of end type generates the completion node instruction of the target process; The state transition instruction, the task node identifier, the timestamp identifier, and the source and target states of the state identifier are encapsulated into an event tuple for the target process.

[0103] Specifically, for each candidate binding relationship in the signal node binding pairs, the timestamp identifier of the signal in the binding pair and the execution time window of the corresponding task node are obtained. The execution time window is jointly defined by the expected start time and expected end time of the task node. The timestamp identifier is compared to see if it is greater than or equal to the start time of the execution time window and less than or equal to the end time of the execution time window. If this condition is met, the candidate binding relationship is confirmed as a valid temporary signal node binding pair and retained for subsequent processing.

[0104] Specifically, each binding pair is retrieved from the temporary binding pairs of the signal nodes, and the type identifier of the execution signal contained therein is read. If the value of the type identifier is a start type, an activation node instruction is generated, which means that the state of the corresponding task node is required to change from ready to active. If the value of the type identifier is an end type, a completion node instruction is generated, which means that the state of the corresponding task node is required to change from active to completed.

[0105] Specifically, for each of the temporary binding pairs of the signal nodes, the corresponding generated state transition instruction, the identifier of the task node in the binding pair, the timestamp identifier of the signal in the binding pair, and the state identifier of the task node before executing the signal and the state identifier that it should change to afterward are extracted.

[0106] Furthermore, if the timestamp is less than the start time of the execution time window or greater than the end time of the execution time window, the signal is considered an abnormal event. All field information of the signal, along with the task node identifier that triggered the abnormality, is recorded in a separate abnormal event record table. This abnormal event record table includes the signal type identifier, source identifier, timestamp identifier, expected associated task node identifier, and abnormal cause description. The abnormal cause description is divided into two types: a timestamp earlier than the start time of the execution time window is called an early arrival signal, and a timestamp later than the end time of the execution time window is called a delayed arrival signal.

[0107] Furthermore, activation node instructions and completion node instructions are collectively referred to as state transition instructions. Each state transition instruction corresponds one-to-one with the execution signal that generated it, and carries the task node identifier bound to that execution signal. State transition instructions do not contain any numerical calculations or conditional judgments; they serve only as explicit action indicators to drive the subsequent event tuple encapsulation process.

[0108] Furthermore, the transition source state is determined based on the signal type identifier: for start-type signals, the source state is ready and the target state is active; for end-type signals, the source state is active and the target state is completed. These four elements are encapsulated into a data structure in a fixed order. This data structure contains four fields: the first field stores the state transition instruction, the second field stores the task node identifier, the third field stores the timestamp identifier, and the fourth field stores the state transition pair consisting of the transition source state and the target state. The resulting data structure after encapsulation is called the event tuple, and each signal node temporarily binds to generate a unique event tuple.

[0109] In summary, by performing validity checks on signals through execution time windows, only signals whose timestamps fall within the node's execution time window are included in the temporary binding pair, while signals outside the time window are recorded separately as abnormal events. This filters out invalid signals that arrive early or late, and provides an independent recording channel for subsequent anomaly investigation.

[0110] In summary, by converting the type identifier of the original execution signal into a unified state transition instruction within the system, the start type is mapped to the activation node instruction and the end type is mapped to the completion node instruction, thus eliminating the heterogeneity of the signal format and enabling subsequent state update operations to be executed uniformly in an instruction-driven manner.

[0111] In summary, state transition instructions, task node identifiers, timestamp identifiers, and transition pairs from the source state to the target state are packaged into a self-contained event tuple, enabling the event record to have complete semantic information. This facilitates the subsequent generation of execution event trajectories by sorting by timestamp and supports independent auditing and backtracking of each state transition step.

[0112] In this embodiment of the invention, when updating the execution state of the corresponding task node in the collaborative allocation sequence based on the state transition markers in the execution event trajectory to obtain the collaborative allocation closed-loop record of the target process, it is specifically used for: The state of the corresponding node in the collaborative allocation sequence is updated according to the type of state transition instruction in the execution event trajectory, and the resources occupied by the corresponding node are released to obtain the resource release record of the target process; The final state of the nodes in the collaborative allocation sequence, the resource release record, and the signal node binding pair of the target process are merged to obtain the collaborative allocation closed-loop record of the target process.

[0113] Specifically, if the state transition instruction is an activation node instruction, the task node corresponding to the identifier is found in the collaborative allocation sequence, and its current state identifier is changed from ready to active. No resources are released at this time because the activation node instruction only indicates that the task has started execution; the resources are still occupied by the node. If the state transition instruction is a completion node instruction, the task node corresponding to the identifier is found in the collaborative allocation sequence, and its current state identifier is changed from active to completed. Simultaneously, all resource entries occupied by the task node from the state snapshot at the start of execution are identified, including location resources, equipment resources, and role / human resources. The unique identifiers of these resource entries are removed from the occupied state table, and the time of removal is recorded as the resource release time.

[0114] Specifically, the final state identifier of each task node is read from the collaborative allocation sequence. This final state identifier is the state obtained after updating all state transition instructions in the execution event trajectory, and may take the values ​​of hibernation, ready, active, or completed. All resource release entries are extracted from the resource release record, and each entry contains a task node identifier, a resource identifier, and a release time.

[0115] Furthermore, each released resource is recorded as a resource release entry in the format of task node identifier, resource identifier, and release time. All resource release entries are arranged in chronological order of release time, forming a data structure called the resource release record. This resource release record completely records all resource information released after each task node in the collaborative allocation sequence is completed.

[0116] Furthermore, all binding relationships are extracted from the signal node binding pairs of the target process. Each binding relationship contains the corresponding information between the execution signal and the task node. The above three data sets are then merged by creating a new empty data structure with the task node identifier as the primary key. The final state identifier of each task node, the entry in the resource release record corresponding to that task node, and the entry in the signal node binding pair corresponding to that task node are all stored in different fields under that task node identifier. The data structure obtained after merging is called the collaborative allocation closed-loop record. This collaborative allocation closed-loop record contains complete execution closed-loop information for each task node in the collaborative allocation sequence, including the node's initial state, received execution signals, state transition process, final state, and the release status of the occupied resources, realizing a fully traceable record from task allocation to execution to resource recovery.

[0117] In summary, by executing activation and completion node instructions in the event trajectory to drive the status updates of corresponding task nodes in the collaborative allocation sequence, the system ensures that the dormant, ready, activated, and completed states of task nodes remain strictly consistent with the actual execution progress. When a completion node instruction triggers a status update to complete, the location resources, equipment resources, and human resources occupied by that node are released simultaneously, preventing resources from being occupied by completed tasks for extended periods without being reallocated. Each resource release operation is recorded as a resource release log, categorized by task node identifier, resource identifier, and release time, providing a precise available resource time window for subsequent task allocation and preventing resource conflicts and leaks.

[0118] In summary, the final state of each task node in the collaborative allocation sequence after all state transitions, all resource entries released by that node in the resource release record, and the association between that node and the execution signal in the signal node binding pair are merged into a single data structure, forming a complete closed-loop record with the task node identifier as the primary key. This collaborative allocation closed-loop record simultaneously includes the task allocation result, the source of the execution signal, the state transition sequence, and resource release information, achieving end-to-end traceability from task allocation to execution triggering to resource reclamation. Teaching administrators can use this closed-loop record to evaluate the execution efficiency of each task node, conduct backtracking analysis on allocation deviations caused by abnormal signals, and provide historical reference for the generation of subsequent batches of collaborative allocation sequences.

[0119] Compared with the prior art, the present invention has the following beneficial effects:

[0120] 1. This technical solution significantly improves the accuracy of task execution sequence identification and the comprehensiveness of resource matching by incorporating explicit and implicit pre-task identifiers from multi-source task sequences into dependency construction and combining idle periods in teaching status data with virtual classroom status into a status snapshot. Based on the resource requirements of the first task in the task chain, a candidate role set is selected from the status snapshot. Then, the role with the lowest load value is used as the load balancing anchor point to cascade subsequent tasks. This ensures that the role selection of each task node in the entire collaborative allocation sequence is constrained by the load level of the anchor point, thereby improving the balance of task allocation among multiple roles in a hybrid teaching scenario.

[0121] 2. This technical solution continuously receives attendance and completion signals from offline classrooms and online platforms, and associates and binds them with the corresponding task nodes in the collaborative allocation sequence according to timestamps, generating an execution event trajectory and achieving precise tracking of task state changes. Based on the state change markers in the execution event trajectory, the node state is updated and the occupied resources are released synchronously. Finally, the final node state, resource release record, and signal node binding are merged into a collaborative allocation closed-loop record, thus forming a complete data closed loop from task allocation to execution triggering to resource recovery, improving the traceability of the teaching process and the efficiency of resource utilization.

[0122] like Figure 2 The diagram shown is a functional block diagram of a collaborative allocation system for blended teaching tasks in an art practice course provided by an embodiment of the present invention.

[0123] The art practice course blended learning task collaborative allocation system 100 of this invention can be installed in an electronic device. Depending on the functions implemented, the art practice course blended learning task collaborative allocation system 100 may include a data acquisition module 101, a task chain module 102, a candidate role set module 103, a collaborative allocation sequence module 104, an execution event trajectory module 105, and a collaborative allocation closed-loop recording module 106. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and which are stored in the memory of the electronic device.

[0124] In this embodiment, the functions of each module / unit are as follows: The data acquisition module collects multi-source task sequences and teaching status data for the target process; The task chain module constructs the task chain of the target process based on the execution constraints of the multi-source task sequence and the dependency relationship between the preceding and subsequent tasks. The candidate role set module combines the available resources and role status of the teaching status data into a status snapshot, and filters the corresponding candidate role set from the status snapshot based on the resource requirements of the first task in the task chain. The collaborative allocation sequence module takes the role with the smallest role load value in the candidate role set as the allocation target, and uses the allocation target as the load balancing anchor point to cascade the subsequent tasks of the task chain to obtain the collaborative allocation sequence of the target process. The execution event trajectory module continuously receives the attendance and completion signals of offline classrooms and online platforms during the target process based on the execution progress of the collaborative allocation sequence, and associates and binds the attendance and completion signals with the corresponding task nodes in the collaborative allocation sequence according to the timestamp to obtain the execution event trajectory of the target process. The collaborative allocation closed-loop record module updates the execution status of the corresponding task node in the collaborative allocation sequence based on the state transition markers in the execution event trajectory, thereby obtaining the collaborative allocation closed-loop record of the target process.

[0125] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0126] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0127] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0128] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0129] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0130] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for collaboratively allocating blended teaching tasks in an art practice course, characterized in that, The method includes: Collect multi-source task sequences and teaching status data of the target process; The task chain of the target process is constructed based on the execution constraints of the multi-source task sequence and the dependencies between the preceding and subsequent tasks. The available resources and role status of the teaching status data are combined into a status snapshot, and the corresponding candidate role set is selected from the status snapshot based on the resource requirements of the first task in the task chain. The role with the smallest role load value in the candidate role set is taken as the allocation target, and the subsequent tasks of the task chain are cascaded with the allocation target as the load balancing anchor point to obtain the collaborative allocation sequence of the target process. Based on the execution progress of the collaborative allocation sequence, the system continuously receives the attendance and completion signals from offline classrooms and online platforms during the target process, and associates and binds the attendance and completion signals with the corresponding task nodes in the collaborative allocation sequence according to their timestamps to obtain the execution event trajectory of the target process. Based on the state transition markers in the execution event trajectory, update the execution state of the corresponding task node in the collaborative allocation sequence to obtain the collaborative allocation closed-loop record of the target process.

2. The method for collaborative allocation of blended teaching tasks in art practice courses as described in claim 1, characterized in that, The multi-source task sequence and teaching status data collected during the target acquisition process include: The names of the practical tasks, the types of teaching resources required, and the estimated duration of the target process are combined to form the original task description set of the target process, and a corresponding task identifier is assigned to the original task description set. The student's submitted practice tasks, question messages, and work files are collected at fixed time intervals during the target process to obtain the student task trace of the target process. The student task trace is then timestamped with the task identifier to obtain the task description information of the target process. Extract reservation records from the venue equipment reservation data during the target process. The reservation records include venue identifier, start time, and end time. The interaction records and member list of the teaching platform during the target process are used as the platform interaction state set of the target process; The task description information is sorted by creation timestamp to obtain the multi-source task sequence of the target process, and the reservation record and the platform interaction status set are combined into the teaching status data of the target process.

3. The method for collaborative allocation of blended teaching tasks in art practice courses as described in claim 1, characterized in that, The task chain for constructing the target process based on the execution constraints of the multi-source task sequence and the dependencies between preceding and subsequent tasks includes: Based on the structured markers of the task description field in the multi-source task sequence, the explicit preceding task identifiers in the multi-source task sequence are located, and the preceding task list of the target process is obtained. Based on the temporal relationship of task creation timestamps in the multi-source task sequence, the implicit predecessor task identifiers that the tasks with missing explicit predecessor task identifiers depend on are added to the predecessor task list. Using the current task in the multi-source task sequence as a node, add directed edges from the node of the preceding task to the node of the current task for the preceding tasks in the preceding task list, to obtain the directed graph of the target process. Based on the starting node of the directed graph, the successor node sequence is recursively extracted along the directed edges. The chain copy is split at the branch node of the successor node sequence until the end node of the successor node sequence, so as to obtain the original task chain of the target process. Merge the original task chains with the same prefix to obtain the task chain merge set of the target process, record branch information at the branch nodes of the task chain merge set, assign chain identifiers and node numbers to the task chain merge set, and obtain the task chain of the target process.

4. The method for collaborative allocation of blended teaching tasks in art practice courses as described in claim 1, characterized in that, The step of combining the available resources and role status of the teaching status data into a status snapshot, and filtering the corresponding candidate role set from the status snapshot based on the resource requirements of the first task in the task chain, includes: Starting from the current moment of the target process, the occupied time slot sequence of the reservation records in the teaching status data is scanned in chronological order. The interval between adjacent occupied time slots in the occupied time slot sequence is taken as the free time slot. The start and end times of the free time slots are collected into a list of free time slots of the target process according to the location identifier. The virtual classroom identifier, number of online students, and activity level markers of the teaching status data are used as the classroom status list for the target process; Using the list of idle time slots and the list of classroom statuses as attribute values, and the current clock time of the target process as the generation timestamp, a state snapshot of the target process is constructed. Based on the type of art practice resources required for the first task in the task chain, the list of execution roles, and the skill level requirements, resources that match the elements and are compatible with the time period are located from the state snapshot to obtain the candidate role set for the target process.

5. The method for collaborative allocation of blended teaching tasks in art practice courses as described in claim 1, characterized in that, The step of selecting the role with the smallest role load value in the candidate role set as the allocation target, and using the allocation target as the load balancing anchor point, cascading the subsequent tasks of the task chain to obtain the collaborative allocation sequence of the target process includes: The load value of the first role in the candidate role set is taken as the initial minimum value, and the first role is taken as the temporary target of the target process; By sequentially comparing the load values ​​of subsequent roles in the candidate role set with the initial minimum value, the temporary target is updated to obtain the first task allocation anchor point of the target process; The role corresponding to the allocation target is set as the load balancing anchor point, and the dynamic load adjustment value of the available role in the state snapshot during the target process is calculated based on the order of subsequent tasks in the task chain: The dynamic load adjustment value is compared with the load value of the load balancing anchor point, and the role with the smallest value that is not greater than the load value of the load balancing anchor point is selected as the subsequent task allocation target of the target process. The first task allocation anchor point and the subsequent task allocation targets are arranged in the order of the task chain to obtain the collaborative allocation sequence of the target process.

6. The method for collaborative allocation of blended teaching tasks in art practice courses as described in claim 5, characterized in that, The calculation formula for the dynamic load adjustment value includes: in, The dynamic load adjustment value, The number of unfinished tasks currently assigned to the character. The time urgency sensitivity coefficient This represents the total remaining time for any unfinished tasks by the character. The total estimated duration of tasks assigned to the character. The historical completion rate impact coefficient. This represents the number of tasks the character has completed. The total number of tasks assigned to the character. This is the nonlinear amplification factor. This is the coupling adjustment coefficient.

7. The method for collaborative allocation of blended teaching tasks in art practice courses as described in claim 1, characterized in that, The execution progress based on the collaborative allocation sequence continuously receives attendance and completion signals from offline classrooms and online platforms during the target process, and associates and binds the attendance and completion signals with corresponding task nodes in the collaborative allocation sequence according to timestamps to obtain the execution event trajectory of the target process, including: Assign a status identifier to the task nodes in the collaborative allocation sequence. The status identifier includes sleep, ready, and active. Set the initial state of the task node to sleep. Receive execution signals from the offline classroom and the online platform, wherein the execution signals include a type identifier, a source identifier, and a timestamp identifier; Based on the current state and resource requirements of the task nodes in the collaborative allocation sequence, the signal with the type identifier "start" is matched to the task node whose state identifier is "ready" and whose resource requirement matches the source identifier; the signal with the type identifier "end" is matched to the task node whose state identifier is "active" and whose resource usage corresponds to the source identifier, thus obtaining the signal node binding pair of the target process. Based on the interval positioning of the signal node binding pair and the timestamp identifier within the execution time window of the corresponding node, the signal node binding pair is converted into the state transition event tuple of the target process according to the type identifier; The execution event trajectory of the target process is obtained by sorting the state transition event tuples by timestamp.

8. The method for collaborative allocation of blended teaching tasks in art practice courses as described in claim 7, characterized in that, The step of locating the signal node binding pair and the timestamp identifier within the interval of the corresponding node's execution time window, and converting the signal node binding pair into a state transition event tuple of the target process according to the type identifier, includes: Signals falling within the execution time window are established as temporary signal node binding pairs, and signals falling outside the execution time window are recorded as abnormal events of the target process. According to the type identifier, corresponding state transition instructions are generated respectively, wherein the signal with the type identifier of start type generates the activation node instruction of the target process, and the signal with the type identifier of end type generates the completion node instruction of the target process; The state transition instruction, the task node identifier, the timestamp identifier, and the source and target states of the state identifier are encapsulated into an event tuple for the target process.

9. The method for collaborative allocation of blended teaching tasks in art practice courses as described in claim 1, characterized in that, The step of updating the execution state of the corresponding task node in the collaborative allocation sequence based on the state transition markers in the execution event trajectory to obtain the collaborative allocation closed-loop record of the target process includes: The state of the corresponding node in the collaborative allocation sequence is updated according to the type of state transition instruction in the execution event trajectory, and the resources occupied by the corresponding node are released to obtain the resource release record of the target process; The final state of the nodes in the collaborative allocation sequence, the resource release record, and the signal node binding pair of the target process are merged to obtain the collaborative allocation closed-loop record of the target process.

10. A system for collaboratively allocating blended teaching tasks in an art practice course, used to implement the method for collaboratively allocating blended teaching tasks in an art practice course as described in any one of claims 1-9, characterized in that, The system includes: The data acquisition module collects multi-source task sequences and teaching status data for the target process; The task chain module constructs the task chain of the target process based on the execution constraints of the multi-source task sequence and the dependency relationship between the preceding and subsequent tasks. The candidate role set module combines the available resources and role status of the teaching status data into a status snapshot, and filters the corresponding candidate role set from the status snapshot based on the resource requirements of the first task in the task chain. The collaborative allocation sequence module takes the role with the smallest role load value in the candidate role set as the allocation target, and uses the allocation target as the load balancing anchor point to cascade the subsequent tasks of the task chain to obtain the collaborative allocation sequence of the target process. The execution event trajectory module continuously receives the attendance and completion signals of offline classrooms and online platforms during the target process based on the execution progress of the collaborative allocation sequence, and associates and binds the attendance and completion signals with the corresponding task nodes in the collaborative allocation sequence according to the timestamp to obtain the execution event trajectory of the target process. The collaborative allocation closed-loop record module updates the execution status of the corresponding task node in the collaborative allocation sequence based on the state transition markers in the execution event trajectory, thereby obtaining the collaborative allocation closed-loop record of the target process.

Images