Piano grading teaching management method and system based on learning profile

Abstract

The application discloses a piano grading teaching management method and system based on learning situation portrait, and particularly relates to the technical field of piano teaching management, which comprises collecting the on-site sound, rotation sequence, evaluation music and current learning situation portrait of a student before entering the classroom, dividing the on-site sound according to the starting and ending time of sound emission by an edge computing node and marking the source position, and outputting a sound segment table before evaluation; collecting the on-site sound before formal evaluation and identifying the exposure segment corresponding to the evaluation music, determining the auditory exposure pollution in combination with the class waiting path, the entering path and the time sequence connection relationship, and then processing the pollution evaluation and the direct evaluation separately and writing only the direct evaluation result into the learning situation portrait and the grading management link.

Description

Technical Field

[0001] This invention relates to the field of piano teaching management technology, and more specifically, to a piano graded teaching management method and system based on student learning profiles. Background Technology

[0002] In the technology related to piano graded teaching management, the existing approach mainly revolves around how to determine the grade based on the student's performance in the assessment process, and arrange for promotion, retention, course advancement and after-class training accordingly. In actual application, teachers often combine classroom observation and assessment results, or use electronic keyboards, sound pickup devices, teaching terminals, etc. to collect students' pitch, rhythm, wrong notes and overall performance, and then write the corresponding results into the teaching management system as the basis for subsequent management. Taking a training institution with multiple practice rooms for continuous teaching as an example, students of the same level often need to wait in the waiting area and take turns entering the assessment between adjacent classrooms. The connection time between students is very short. Teachers need to continuously demonstrate, correct mistakes, clap and prompt sentences before and after the previous student finishes. At the same time, the scene is also limited by the limited sound insulation between classrooms and the need for the assessment results to be fed back on the spot and used immediately for class adjustment and progress arrangement. In this situation, the following phenomenon tends to occur repeatedly in practice: some students have already heard the corresponding piece, rhythm framework, fingering method, or the teacher's correction tips before the formal assessment. Subsequently, they perform a result that does not correspond to their true stable level in a short period of time. The existing management method usually makes a grade judgment directly based on this result, which causes the students to show a significant decline in performance after subsequent retests, changing pieces, or leaving the on-site prompts. This leads to over-grading, inaccurate class grouping, and repeated regression in the teaching pace. The root cause of this problem is that the existing scheme only focuses on the performance results presented by students during the assessment, but does not include the auditory exposure before the assessment in the management scope, thus failing to identify and isolate the interference of such short-term auditory influence on the validity of the grade judgment. The technical problem this application aims to solve is: how to identify and control the auditory exposure influence formed before the formal assessment of students in the process of piano graded teaching management based on student learning profiles, so as to avoid the performance results affected by this influence being directly used for grade determination and subsequent teaching management. Summary of the Invention

[0003] To overcome the aforementioned deficiencies in the prior art, embodiments of the present invention provide a piano graded teaching management method and system based on student learning profiles. This method collects on-site sound before formal assessments and identifies exposure segments corresponding to the assessment pieces. It then determines auditory exposure contamination by combining waiting paths, entry paths, and temporal sequence relationships. Finally, it streams the contamination assessment and direct assessment, and only writes the direct assessment results into the student learning profile and graded management link, thereby solving the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a piano graded teaching management method based on student learning profiles, comprising: S1. Collect the on-site sound, rotation order, test songs and current learning status profile of the students to be tested before they enter the venue. The on-site sound is segmented by the edge computing node according to the start and end times of the sound and the source location is marked. Output the sound segment table before the test. S2. Based on the pre-evaluation sound segment table and the evaluation track, perform segment-by-segment correspondence, and determine the exposed segment according to at least one of the following: consistent pitch order, consistent beat placement order, consistent starting fingering command, and consistent error correction language. Record the source location, occurrence time, and corresponding track position, and output the exposure record table. S3. Perform contamination determination based on the exposure record table and rotation sequence. If there is an exposure segment before the formal assessment, mark this assessment as a contamination assessment and generate an isolation retest sequence. If there is no exposure segment, mark this assessment as a direct assessment and output the assessment status table. S4. Obtain the performance data of the student under test in the order of direct assessment or isolated retest. The edge computing node binds the performance data with the assessment status table. During the contaminated assessment, the performance data is retained as a practice record. During the direct assessment, the pitch completion, rhythm completion and paragraph completion are extracted, and the valid assessment results are output. S5. Based on the effective assessment results and the current learning situation profile, perform updates. When conducting direct assessments, generate level identification, class grouping adjustments, and after-school training arrangements. When only corresponding practice records are recorded, maintain the original level and write a retest mark, and output the results of graded teaching management.

[0005] In a preferred embodiment, S1 includes: S11. Read the live sound during the preset time period before the students enter the room according to the rotation order of the students to be tested, and divide the continuous live sound into multiple sound segments according to the start time and end time of the sound, and output the sound segment sequence. S12. Extract the start time, duration, and arrival time difference from the sound source to the edge computing node for each sound segment in the sound segment sequence, and perform corresponding judgments based on the waiting area location, assessment area location, and teacher location, and output the source location marker corresponding to each sound segment. S13. Write the sound segment sequence, the source location mark corresponding to each sound segment, the rotation order of the students to be tested, the test pieces, and the current learning situation profile into the same pre-test sound segment table in the order in which the sound segments appear, and output the pre-test sound segment table to be read in subsequent steps.

[0006] In a preferred embodiment, S2 includes: S21. Establish a segment-by-segment pairing table between each sound segment in the pre-evaluation sound segment table and the segment sequence of the evaluation piece according to the order of their appearance. Extract the pitch order field, beat placement field, command text field, and error correction text field for each sound segment. Extract the standard pitch order field, standard beat placement field, standard starting fingering command field, and standard error correction term field for each piece segment. Output a candidate correspondence table of sound segment and piece segment. S22. For each candidate corresponding item in the candidate correspondence table of sound segment-track segment, perform pitch order verification, beat placement verification, starting fingering command verification, and error correction term verification respectively. Write the corresponding verification mark when any verification is true, write a strong correspondence mark when two or more verifications are true at the same time, write a continuation correspondence mark when only a single verification is true and adjacent sound segments point to the same track segment, and write a removal mark when none of the verifications are true. Output the corresponding mark table.

[0007] In a preferred embodiment, S2 further includes: S23. Based on the corresponding tag table, perform a timing continuity check on sound segments with the same source location and consecutive occurrence times. Write a continuation exposure tag for candidate corresponding items that are consecutive and point to the same track location. Perform conflict check on candidate corresponding items with different source locations but pointing to the same track location. If the number of verification tags for the later sound segment is greater than that for the earlier sound segment, retain the later sound segment and write an overwrite tag. If the number of verification tags is the same and the source location is closer to the entrance path of the student to be tested, retain the corresponding sound segment and write a priority tag. Output the exposure candidate table. S24. Based on the exposure candidate table, summarize the sound segments with strong corresponding markers, successive corresponding markers, continuous exposure markers, overlay markers, or priority markers. Write the source location, occurrence time, corresponding track location, establishment marker type, and preceding and following successive relationship of the corresponding sound segment into the exposure record table. Write the candidate corresponding items not written into the exposure record table into the item to be reviewed table and output the exposure record table.

[0008] In a preferred embodiment, S3 includes: S31. Based on the source location, occurrence time, corresponding song position and sequential relationship of each exposure segment in the exposure record table, construct a position correspondence matrix according to the same position relationship between the source location and the waiting path and entry path of the student to be tested, construct a time difference matrix according to the difference in occurrence time of each exposure segment, and construct a song continuation matrix according to the same and adjacent relationships of the corresponding song positions of each exposure segment. Then, perform a position-by-position multiplication of the position correspondence matrix, time difference matrix and song continuation matrix to output the pollution correlation matrix. S32. Extract non-zero association items from the pollution association matrix according to row and column positions. Take the exposure segment corresponding to each non-zero association item as a node and construct a pollution continuation graph with the conditions that the source position is continuous, the occurrence time is progressive, and the corresponding track position is the same or adjacent. Then perform spectral decomposition on the adjacency matrix corresponding to the pollution continuation graph, extract connected node groups, and expand them into candidate pollution chains according to the occurrence time order within the node group. Output the candidate pollution chain set.

[0009] In a preferred embodiment, S3 further includes: S33. For each candidate contamination chain in the candidate contamination chain set, the source location sequence of each exposed segment is aligned with the waiting path sequence and entry path sequence of the students to be tested. The occurrence time sequence of each exposed segment is summed by difference. The adjacent displacement of the corresponding track position sequence of each exposed segment is accumulated to form path fit amount, temporal continuity amount and track continuity amount. The path fit amount, temporal continuity amount and track continuity amount are then combined to form a contamination feature vector. Orthogonal projection is performed on the contamination feature vector to output the contamination projection value corresponding to each candidate contamination chain.

[0010] In a preferred embodiment, S3 further includes: S34. Sort the pollution projection values ​​of each candidate pollution chain in descending order. When the pollution projection values ​​are the same, retain the candidate pollution chain whose last exposure time is closer to the formal evaluation time. When the last exposure time is the same, retain the candidate pollution chain with a shorter path length from the source location to the entry location. When the path length is the same, retain the candidate pollution chain with a larger number of corresponding track locations covered. Write the retention results into the pollution locking table and write the unretained results into the auxiliary pollution table. S35. Generate an assessment status table based on the pollution lock table. When the pollution lock table is not empty, mark this assessment as a pollution assessment and adjust the students to be tested to subsequent assessment positions that do not overlap with the source positions corresponding to the pollution lock table to form an isolated retest order. When the pollution lock table is empty, mark this assessment as a direct assessment.

[0011] In a preferred embodiment, S4 includes: S41. Read the performance data and evaluation status table generated by the student under the direct evaluation or isolated re-evaluation order, divide the performance data into performance segments according to the start and end times of the notes, establish a performance segment sequence according to the order in which the performance segments appear, and output the performance segment table. S42. Perform segment-by-segment pairing between the performance segment table and the standard repertoire segment sequence corresponding to the evaluation repertoire. Calculate the pitch deviation, rhythm deviation, and segment coverage for each performance segment. Write the pitch deviation, rhythm deviation, and segment coverage into the performance evaluation table in the order of the performance segments and output the performance evaluation table. S43. Based on the evaluation status table, perform status binding on the performance evaluation table. When the evaluation status table corresponds to the pollution evaluation, write the pitch deviation, rhythm deviation and section coverage of each performance segment into the practice record table. When the evaluation status table corresponds to the direct evaluation, perform summary calculation on the pitch deviation, rhythm deviation and section coverage of each performance segment, and output the valid evaluation table. S44. Based on the valid evaluation form, generate pitch completion status, rhythm completion status, and paragraph completion status respectively. Write the pitch completion status, rhythm completion status, and paragraph completion status, along with the student identification, the evaluation piece identification, and the evaluation order, into the valid evaluation result table, and output the valid evaluation results.

[0012] In a preferred embodiment, S5 includes: S51. Read the valid assessment results and the current learning profile. When the valid assessment results correspond to the direct assessment, write the pitch completion status, rhythm completion status and paragraph completion status into the pitch item, rhythm item and paragraph item corresponding to the current learning profile, respectively. Generate a profile change table according to the field difference before and after this assessment, and output the updated learning profile and profile change table. S52. Based on the updated learning profile and profile change table, perform item-by-item correspondence for the historical profile sequence of students of the same level. When the pitch, rhythm and paragraph items all fall into the same level profile range, generate the level determination result. Then, generate the class adjustment result according to the correspondence between the level determination result and the current class profile range. Subsequently, allocate after-school training content according to the profile item with the largest difference in the profile change table and output the graded processing table. S53. Based on the valid assessment results, the status is divided. When the valid assessment results only correspond to the practice record, the original level is maintained and a retest mark is written to the current learning profile. When the valid assessment results correspond to the direct assessment, the level determination results, class adjustment results and after-class training content in the graded teaching management result table are written to the graded teaching management result table, and the graded teaching management results are output.

[0013] The piano graded teaching management system based on student learning profiles includes a data acquisition and segmentation module, an identification module, a judgment module, a result extraction module, and a graded management module. The data acquisition and segmentation module is used to collect the on-site sound, rotation order, test songs, and current learning status profile of the students before they enter the room. The edge computing nodes segment the on-site sound according to the start and end times of the sound and mark the source location, and output a sound segment table before the test. The identification module is used to perform segment-by-segment correspondence with the test track based on the pre-test sound segment list. It determines the exposed segment and records the source location, occurrence time and corresponding track position according to at least one of the following: consistent pitch order, consistent beat placement order, consistent starting fingering command and consistent error correction language. It then outputs the exposure record table. The judgment module is used to perform contamination judgment based on the exposure record table and the rotation order. If there is an exposure segment before the formal assessment, the current assessment is marked as a contamination assessment and an isolation retest order is generated. If there is no exposure segment, the current assessment is marked as a direct assessment and the assessment status table is output. The results extraction module is used to obtain the performance data formed by the student under the direct assessment or isolated retest sequence. The edge computing node binds the performance data with the assessment status table. During the contaminated assessment, the performance data is retained as a practice record. During the direct assessment, the pitch completion, rhythm completion and paragraph completion are extracted, and the valid assessment results are output. The graded management module updates based on valid assessment results and current student profiles. During direct assessments, it generates grade identification, class grouping adjustments, and after-school training arrangements. When only corresponding practice records are recorded, it maintains the original grade and adds a retest flag, outputting the graded teaching management results.

[0014] The technical effects and advantages of this invention are as follows: 1. This solution collects on-site sound before the formal assessment, identifies the exposure segment corresponding to the assessment piece and performs a contamination judgment, and then transfers the affected assessment to an isolation retest. This can relatively suppress the interference of short-term auditory exposure on the validity of the grade judgment and reduce the situation where contaminated performance results are directly written into the subsequent teaching management chain. 2. The sounds from the waiting area, assessment area, and teacher's location are uniformly divided into sound segments with the time of occurrence and the location of the source, and each segment is matched with a piece of music. This can transform the originally difficult-to-trace on-site auditory contact into a recordable and comparable exposure record, thereby relatively improving the feasibility of exposure identification before assessment. 3. By constructing a location correspondence matrix, a temporal difference matrix, and a track continuation matrix, and combining the contamination continuation diagram, candidate contamination chains, and contamination projection values ​​to achieve unique retention, discrete exposure information can be integrated into continuous contamination relationships, which relatively improves the coherence of contamination determination and the consistency of result attribution. 4. The implementation status of pollution assessment and direct assessment is separated. Only practice records are retained during pollution assessment, and valid assessment results are generated during direct assessment. This can reduce the occurrence of practice results and pollution results being included in the update of student profile, thereby relatively improving the stability of the basis for subsequent level determination and class adjustment. 5. The pitch completion, rhythm completion, and paragraph completion results from the effective assessment results are written into the current learning profile. Combined with the historical profile sequence of students of the same level, the level determination results and class grouping adjustment results are generated. This can make the leveling process correspond to the actual assessment results in a closed loop, which can relatively alleviate the problems of over-leveling, inaccurate class grouping, and regression in teaching pace. 6. Assigning after-class training content based on the portrait item with the largest difference in the portrait change table, and writing a retest mark when only corresponding practice records are included, can ensure that the after-class training arrangement is consistent with the effective shortcomings of this study, and at the same time avoids premature level adjustment when the assessment basis is insufficient, thus relatively improving the pertinence of teaching management. Attached Figure Description

[0015] Fig. 1 This is a flowchart outlining the method steps of the present invention; Fig. 2 This is a schematic diagram of the system module structure of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Refer to the instruction manual appendix Figs. 1-2 The present invention provides a piano graded teaching management method based on student learning profiles, comprising: S1. Collect the on-site sound, rotation order, test songs and current learning status profile of the students to be tested before they enter the venue. The on-site sound is segmented by the edge computing node according to the start and end times of the sound and the source location is marked. Output the sound segment table before the test. In this implementation, the purpose of S1 is to organize the continuous on-site sounds before the students enter the testing room into a calculable, locatable, and associative pre-test sound segment table, providing a unified time, location, and task benchmark for subsequent exposure segment identification. The processing logic is as follows: first, the effective collection range before the students enter the testing room is defined according to the rotation order; then, the on-site sounds within this range are segmented and their locations assigned; finally, the sound segment results, rotation order, testing repertoire, and current student profile are written into the same data table according to a unified field structure. This implementation process includes the following steps: The purpose of S11 is to form a sequence of sound segments. The input includes the rotation order of the students to be tested, as well as the sound streams from the waiting area, the assessment area, and the teacher's near-end sound stream. The rotation order includes at least the student's identifier, the planned entry time, and the planned assessment position number. The starting point of the preset time period before entry is determined by the preset configuration duration backward from the planned entry time, and the ending point is the actual time when the student enters the assessment area. The preset configuration duration is pre-written into the system configuration table. The processing action is as follows: first, the sound streams from the waiting area, the assessment area, and the teacher's near-end sound stream are unified onto the same timeline, and then the on-site scene within the preset time period before entry is captured. The sound is then segmented into continuous effective sound intervals. The sound start time is determined when the sound energy is continuously higher than the silence recognition baseline, and the sound end time is determined when the sound energy falls back to the silence recognition baseline and continues to reach the preset silence interval. When the silence interval between two consecutive sounds from the same source location is shorter than the preset silence interval, the consecutive sounds are merged into the same sound segment. The output is a sequence of sound segments ordered by their order of appearance, with the sound segment number, sound start time, and sound end time written in for S12 to read. If there is no effective on-site sound within the preset time period before entry, an empty sound segment sequence is output and a null value is written in. The purpose of S12 is to determine the source location of each sound segment. The input includes the sound segment sequence, the waiting area location table, the assessment area location table, the teacher location table, and the unified timestamp record of each edge computing node. Among them, the waiting area location table, the assessment area location table, and the teacher location table are established with fixed location numbers during system deployment. The processing action is as follows: extract the sound start time, duration, and arrival order difference from the sound source to the edge computing node for each sound segment. The arrival order difference is obtained by comparing the first hit timestamp of the same sound segment on different edge computing nodes. Then, the corresponding judgment is performed by combining the waiting area location, assessment area location, and teacher location. When the teacher's near-end node hits first and covers the main duration of the sound segment, the sound segment is determined to be the teacher's location sound segment. When the waiting area node is hit first and the corresponding waiting area location number is on the waiting path of the student to be tested, the sound segment is determined to be the waiting area location sound segment; when the assessment area node is hit first and the corresponding assessment position number is consistent with the planned assessment position number, the sound segment is determined to be the assessment area location sound segment; when multiple positions meet the determination conditions at the same time, the position with the higher coverage ratio of the duration is retained as the source position, and when the coverage ratios are the same, the position number that is closer to the entry path of the student to be tested is retained; the output is the source position mark corresponding to each sound segment, and the sound segment number, source position number, source position category, sound start time, duration, and arrival order difference are written for S13 to read; if a unique attribution cannot be formed, the corresponding sound segment is written to the source item to be reviewed and recorded as an undetermined position; The purpose of S13 is to generate a pre-assessment audio segment table that can be directly read in subsequent steps. The input includes the audio segment sequence, the source location markers for each audio segment, the rotation order of the student to be tested, the assessment piece, and the current learning profile. The assessment piece must include at least a piece identifier and a sequence of audio segments, and the current learning profile must include at least the student identifier, current level, pitch, tempo, and section. The processing involves writing the audio segment sequence and source location markers one-to-one according to the order in which the audio segments appear, and simultaneously writing the rotation order, assessment piece, and current learning profile fields to form a unified record. Each record must include at least the student's current level, pitch, tempo, and section. The system generates a pre-assessment sound segment table containing the student identifier, sound segment number, sound start time, sound end time, source location number, source location category, rotation sequence number, planned entry time, assessment song identifier, song segment sequence index, and current student profile index. For sound segments with undetermined positions, the system retains the time field and the original sound segment index, and writes the undetermined identifier into the source location field. For empty sound segment sequences, the system generates an empty pre-assessment sound segment table containing only the student identifier, rotation sequence, assessment song identifier, current student profile index, and null identifier. The output is the pre-assessment sound segment table, which is written to the local cache of the edge computing node and the synchronization area of ​​the teaching management terminal for S2 to read. Under the above processing, S1 converts the pre-entry on-site sound into a unified pre-assessment sound segment table, enabling subsequent steps to complete the segment-by-segment correspondence between sound segments and assessment pieces within the same data table, and ensuring consistency in time sequence, location attribution, and task binding. In practical applications: when teaching in two practice rooms in continuous rotation, the system traces back the waiting period according to the planned entry time of the student to be tested, extracts the on-site sound from the waiting area, assessment area, and teacher side, divides it into multiple sound segments, and assigns each sound segment to the waiting area location, assessment area location, or teacher location based on the first hit timestamp and continuous coverage ratio of each edge computing node. Then, along with the rotation order, assessment pieces, and current student profile, it is written into the pre-assessment sound segment table for direct reading by subsequent exposure segment identification.

[0018] S2. Based on the pre-evaluation sound segment table and the evaluation track, perform segment-by-segment correspondence, and determine the exposed segment according to at least one of the following: consistent pitch order, consistent beat placement order, consistent starting fingering command, and consistent error correction language. Record the source location, occurrence time, and corresponding track position, and output the exposure record table. In this implementation, the purpose of S2 is to identify sound content that directly corresponds to the test track from the pre-test sound segment table, and to organize this sound content into an exposure record table that can be directly read for subsequent pollution determination. The processing logic is as follows: first, candidate pairings are established between the sound segments and the track segments of the test track; then, item-by-item verification is performed based on pitch order, beat placement, starting fingering commands, and error correction terms; subsequently, continuity identification and conflict resolution are completed by combining temporal continuity and source location relationships; finally, the retained results are written into the exposure record table, and the results that are not retained but still require subsequent inspection are written into the pending review item table. This implementation process includes the following steps: The purpose of S21 is to establish a candidate correspondence between sound segments and music segments, providing a unified input carrier for subsequent item-by-item verification. The input consists of the pre-evaluation sound segment table and the evaluation music. The evaluation music is pre-divided into a sequence of music segments according to fixed rules. Each music segment is written with its music segment number, segment sequence number, and beat range in sequence. The fixed rules are determined by the teaching management terminal based on the boundaries of musical measures and teaching phrases when the music is entered into the music catalog. The processing action is to read each sound segment from the pre-evaluation sound segment table in the order of their appearance. For each sound segment, extract the pitch order field, beat placement field, command text field, and error correction text field. The pitch order field consists of the pitch numbers identified sequentially within the sound segment. The beat placement field consists of the time positions corresponding to strong beats, weak beats, and beat divisions. The command text field consists of the starting fingering command code values, such as starting with a certain number of fingers, starting with the right hand, and starting with the left hand. The error correction text field consists of error correction code values, such as slow down, start over, lift the finger, and connect the fingers. Both the command code values ​​and the error correction code values ​​are obtained by mapping from a pre-established teaching command code table. Simultaneously, for each track segment, the standard pitch order field, standard beat placement field, standard starting fingering command field, and standard error correction terminology field are extracted. Then, each sound segment and each track segment are paired to form a candidate correspondence table, which is a sound segment-track segment candidate correspondence table. The output is the sound segment-track segment candidate correspondence table, and the fields written include at least the sound segment number, track segment number, pitch order field, beat placement field, command text field, error correction text field, and corresponding standard field, for S22 to read. If a sound segment does not have a command text field or error correction text field extracted, a null value is written into the corresponding field, but the candidate correspondence item is still retained for subsequent pitch order verification and beat placement verification. The purpose of S22 is to perform item-by-item correspondence verification on candidate corresponding items and form corresponding markers that can be retained or eliminated later. The input is a candidate correspondence table of sound segment to track segment. The processing actions are: to perform pitch order verification, beat placement verification, starting fingering command verification, and error correction term verification on each candidate corresponding item. Among them, pitch order verification adopts a bit-by-bit correspondence method, comparing the pitch number sequence of the sound segment with the standard pitch number sequence of the corresponding track segment in order. It is determined to be valid when the first item is consistent and the subsequent order is not reversed. Beat placement verification adopts a placement correspondence method, mapping the strong beat, weak beat, and beat time position of the sound segment to the standard beat position of the corresponding track segment in order. It is determined to be valid when the beat position order is consistent. Starting fingering command verification adopts a first-item command comparison method. It is determined to be valid when the first command code value of the sound segment is consistent with the standard starting fingering command field. Error correction term verification adopts a code value hit method. It is determined to be valid when the error correction term code value set of the sound segment has a same position hit or consecutive hits with the standard error correction term field. Write corresponding verification flags for each of the four types of verifications; write corresponding verification flags when any verification is true; write strong correspondence flags when two or more verifications are true at the same time; write continuation correspondence flags when only a single verification is true and the preceding or following sound segment after sorting by occurrence time points to the same track segment; write removal flags when none of the four types of verifications are true; the output is a corresponding flag table, and the written fields include at least the sound segment number, track segment number, four types of verification results, corresponding verification flags, strong correspondence flags, continuation correspondence flags, and removal flags, for S23 to read; if a candidate corresponding item cannot complete the corresponding verification only because the password text field or the error correction text field is missing, then the incomplete verification item is written to the empty item flag, and the remaining completed verification items still participate in the flag generation; The purpose of S23 is to complete the identification of continued exposure and conflict resolution based on the corresponding tag table, forming a unique candidate exposure result that is retained or pending review. The input includes the corresponding tag table, the source location field in the pre-assessment sound segment table, and the entrance path of the student to be tested. The entrance path is formed by sequentially expanding from the waiting area location number to the planned assessment position number. The processing actions are as follows: First, perform a temporal continuity check on sound segments with the same source location and consecutive occurrence times. Consecutive occurrence times mean that the start time of the next sound segment after sorting by occurrence time is adjacent to the end time of the previous sound segment, and there are no other source location sound segments inserted in between. Write the continued exposure tag for the candidate corresponding items that are consecutive and point to the same piece of music position. Next, conflict checks are performed on candidate corresponding items with different source locations but pointing to the same track location: when the number of verification tags for the later sound segment is greater than that for the earlier sound segment, the later sound segment is retained and a cover tag is written; when the number of verification tags is the same, the number of path steps from the source location to the student's entry path is calculated, and the sound segment with the smaller number of path steps is retained and a priority tag is written; when the number of path steps is the same, the sound segment with the later occurrence time is retained; when the occurrence time is still the same, the sound segment with the smaller sound segment number is retained; the output is an exposed candidate table, and the written fields include at least the sound segment number, source location, occurrence time, corresponding track location, retention tag, continuation exposure tag, cover tag, priority tag, and preceding and following relationship, for S24 to read; for candidate corresponding items that are not retained but have at least one valid verification, they are not directly deleted, but written to the candidate area to be reviewed, for subsequent formation of the item to be reviewed table; The purpose of S24 is to organize the exposure candidate table into an exposure record table that can be directly used for subsequent pollution determination, and to retain the review entry for the results that are not selected; the input is the exposure candidate table; the processing action is to summarize the sound segments with strong corresponding markers, consecutive corresponding markers, continued exposure markers, overlay markers or priority markers, and write the source position, occurrence time, corresponding track position, establishment marker type and preceding and following relationship of the corresponding sound segment into the exposure record table; Specifically, the establishment marker type is written item by item according to the actual establishment status of strong correspondence marker, successive correspondence marker, continuation exposure marker, coverage marker, and priority marker. The successive relationship is written with the preceding and following sound segment numbers. If the same sound segment has multiple establishment marker types, all of them are retained and written. For candidate correspondence items that are not written to the exposure record table but enter the candidate area for review, they are written to the candidate item table for review. The candidate item table for review includes at least the sound segment number, track segment number, reason for not retaining, and established verification item. The output is the exposure record table, and the candidate item table for review is output synchronously. The exposure record table is written to the local cache of the edge computing node and the subsequent S3 reading area, and the candidate item table for review is written to the manual review area or the subsequent rule update area. If the exposure candidate table is empty, an empty exposure record table is generated and a null value identifier is written. Through the above processing, S2 completed the conversion from the pre-evaluation sound segment table to the exposure record table, so that the subsequent pollution judgment no longer directly faces the original sound segments, but faces the exposure results that have been matched segment by segment, verified item by item, identified continuity and resolved conflicts, thereby ensuring that the subsequent pollution judgment has a clear basis for the corresponding music, the continuity of time and the source location. In practical application: Before the student enters the formal assessment, the sound segment in the waiting area contains the main melody fragment of the piece played by the previous student, and the sound segment in the teacher's position contains commands such as "start with the second finger of the right hand" and "slow down and start again." The system first establishes candidate correspondences between these sound segments and each segment of the piece in this assessment, and then performs item-by-item verification on pitch order, beat placement, starting fingering commands, and error correction terms. When two sound segments in the waiting area appear consecutively and both point to the same piece, the system writes a continuation exposure marker. When the sound segment in the teacher's position and the sound segment in the assessment area both point to the same piece, the system performs conflict resolution based on the number of verification markers and the number of steps from the source position to the entry path, retaining a unique result and writing it into the exposure record table, and writing the unretained result into the pending review item table. Subsequently, S3 can directly determine whether this formal assessment is affected by auditory exposure based on the exposure record table.

[0019] S3. Perform contamination determination based on the exposure record table and rotation sequence. If there is an exposure segment before the formal assessment, mark this assessment as a contamination assessment and generate an isolation retest sequence. If there is no exposure segment, mark this assessment as a direct assessment and output the assessment status table. In this implementation, the purpose of S3 is to make a unique determination based on the exposure record table and the rotation order to determine whether the auditory exposure before the formal assessment has formed an effective contamination for this assessment, and to generate a direct assessment or isolation retest order accordingly. The processing logic is as follows: first, the relationship between the exposure segment and the waiting path, entry path, and assessment song position of the student to be tested is converted into a unified matrix; then, related items with simultaneous validity are extracted from the matrix and a contamination continuity diagram is constructed; subsequently, the path fit, temporal continuity, and song continuity are calculated for each candidate contamination chain to form a contamination projection value and complete unique retention; finally, the retention results are written into the contamination locking table and an assessment status table is generated. This implementation process includes the following steps: The purpose of S31 is to unify the time relationships, positional relationships, and repertoire relationships in the exposure record table into a computable matrix relationship, thereby providing a consistent input for subsequent contamination chain extraction. The input quantities are the exposure record table and the rotation order of the students to be tested. The exposure record table includes at least the exposure segment number, source position, occurrence time, corresponding repertoire position, and sequential relationship. The rotation order includes at least the student identifier, planned entry time, waiting path, and entry path. The waiting path is expanded from the waiting area position number in the rotation order, and the entry path is expanded from the waiting area end position number to the planned assessment position number in the actual passage order. The processing action is as follows: first, construct a position correspondence matrix with the exposure segment number as the row and the position numbers in the waiting path and entry path as the column. When the source position of the exposure segment is the same as the corresponding column position number, or the source position is adjacent to the position number in the column, write the position corresponding value in the corresponding element of the position correspondence matrix. Write 1 for the same position, 0.5 for adjacent positions, and 0 for the rest. The determination of adjacent positions is based on the sequential relationship in the fixed position number table. Next, construct a temporal difference matrix using the exposed segment numbers as rows and columns. For any two exposed segments, subtract the time of occurrence of the earlier exposed segment from the time of occurrence of the later exposed segment to form a temporal difference value. Write the difference value when it is greater than 0, and write 0 when it is less than or equal to 0. Then, construct a track continuation matrix using the exposed segment numbers as rows and columns. Write 1 when the corresponding track positions of two exposed segments are the same, write 0.5 when the corresponding track positions differ by 1 in segment number or are adjacent in beat position within the same track segment, and write 0 otherwise. Then, unify the position correspondence matrix, temporal difference matrix, and track continuation matrix to the same dimension and perform a bitwise multiplication. Only retain the positions where all corresponding elements of the three matrices are non-zero to form a contamination correlation matrix. The output is the contamination correlation matrix, which is written to the exposed segment row and column index table for S32 to read. If the exposed record table is empty, an empty contamination correlation matrix is ​​directly generated and a null value is written. Candidate contamination chains will not be extracted in subsequent steps. The purpose of S32 is to extract a set of exposed segments with continuous propagation potential from the pollution correlation matrix and organize them into candidate pollution chains. The inputs are the pollution correlation matrix and the source location, occurrence time, and corresponding track location fields in the exposure record table. The processing actions are as follows: First, extract non-zero correlation items according to the row and column positions of the pollution correlation matrix. Each non-zero correlation item corresponds to a pair of exposed segments with pollution correlation. Using the exposed segments corresponding to each non-zero correlation item as nodes, construct a pollution continuation graph with the conditions of continuous source location, progressive occurrence time, and the same or adjacent corresponding track location. Continuous source location means that the source locations of two exposed segments have the same or adjacent position numbers in the waiting path or entrance path. Progressive occurrence time means that the occurrence time of the later exposed segment is later than the occurrence time of the previous exposed segment. The same or adjacent corresponding track location is determined according to the track continuation rules in S31. After completing the contamination continuation graph, the corresponding adjacency matrix is ​​constructed, and spectral decomposition is performed on the adjacency matrix. Specifically, the characteristic structure of the adjacency matrix is ​​calculated, and connected node groups are extracted according to the block results where non-zero edges are only clustered within the same sub-block. Each connected node group corresponds to a continuable contamination subset. Then, the exposed segments in each connected node group are sorted from earliest to latest according to their occurrence time and expanded into candidate contamination chains. The output is a set of candidate contamination chains, and the written fields include at least the candidate contamination chain number, exposed segment sequence, source location sequence, occurrence time sequence, and corresponding track location sequence for S33 to read. If the contamination association matrix is ​​not empty but has no edges, each individual exposed segment is written as a single-node candidate contamination chain into the candidate contamination chain set. The purpose of S33 is to calculate the pollution intensity of each candidate contamination chain and unify the pollution probability on different paths to a comparable projection value. The inputs are the candidate contamination chain set, the waiting path sequence of the students to be tested, and the entry path sequence. The waiting path sequence and the entry path sequence are both expanded in actual order by fixed position numbers. The processing action is as follows: For each candidate contamination chain in the candidate contamination chain set, the source position sequence of the candidate contamination chain is first aligned with the waiting path sequence and the entry path sequence respectively. The sequence alignment is compared position by position in order. If the source position and the path position number are the same, it is counted as 1 hit. If the source position and the path position number are adjacent, it is counted as 0.5 hits. The rest are counted as 0 hits. Then, all hit values ​​are summed to form the path fit value. The larger value between the waiting path fit value and the entry path fit value is taken as the path fit value of the candidate contamination chain. Next, the occurrence time sequence of each exposed segment is differentially calculated by subtracting the previous occurrence time from the subsequent occurrence time. Positive differences are retained, and all positive differences are summed to form the temporal continuity. If a candidate contamination chain contains only one exposed segment, the temporal continuity is recorded as the interval between the occurrence time of that exposed segment and the planned entry time. Then, the corresponding track position sequence of each exposed segment is accumulated by subtracting the previous item from the subsequent item. When track positions are the same, 0 is accumulated; when track positions are adjacent, 1 is accumulated; when a track position spans multiple segments, the actual number of segments crossed is accumulated to form the track continuity. Finally, the path fit, temporal continuity, and track continuity are calculated. A three-dimensional contamination feature vector is formed, and orthogonal projection is performed using a fixed judgment basis vector. The fixed judgment basis vector is pre-written by the system configuration table and corresponds to the contamination reference direction for high path fit, short temporal interval, and continuous track position. The projection result is the contamination projection value corresponding to the candidate contamination chain. The output is the contamination projection value corresponding to each candidate contamination chain, and the path fit, temporal duration, track duration, and contamination projection value are written into the candidate contamination chain evaluation table for S34 to read. If a candidate contamination chain is missing a source position sequence or a corresponding track position sequence, it is written into the pending review list and does not participate in the automatic sorting in this round. The purpose of S34 is to uniquely retain among multiple candidate contamination chains, forming a direct basis for the generation of subsequent evaluation status. The inputs are the candidate contamination chain evaluation table and the planned formal evaluation time. The processing actions are as follows: First, sort the contamination projection values ​​of each candidate contamination chain in descending order; when the contamination projection values ​​are the same, compare the last exposure time of each candidate contamination chain, and retain the candidate contamination chain whose last exposure time is closer to the formal evaluation time; when the last exposure time is the same, calculate the path length from the last source location of each candidate contamination chain to the entrance location of the student to be tested. The path length is accumulated according to the number of steps in the fixed location number table, and the candidate contamination chain with the shorter path length is retained. When the path length is the same, count the number of corresponding track positions covered by each candidate contamination chain after deduplication, and retain the candidate contamination chain with more corresponding track position coverage; when the corresponding track position coverage is still the same, retain the candidate contamination chain with the smaller candidate contamination chain number; write the finally retained candidate contamination chain into the contamination locking table, and write the unretained candidate contamination chain into the auxiliary contamination table; the output is the contamination locking table and the auxiliary contamination table; the contamination locking table includes at least the main contamination chain number, the main contamination chain exposure segment sequence, the last exposure time, the source location sequence, the corresponding track position sequence, and the contamination projection value, for S35 to read; if the candidate contamination chain evaluation table is empty, generate an empty contamination locking table and write a null value identifier; The purpose of S35 is to uniquely classify this assessment as either a contaminated assessment or a direct assessment based on the contamination lock table, and to generate an executable isolation retest sequence during the contamination assessment. The inputs are the contamination lock table, the rotation order, and the current assessment position queue table. The processing actions are as follows: when the contamination lock table is not empty, mark this assessment as a contaminated assessment and read the source position sequence in the contamination lock table; then, in the current assessment position queue table, scan all subsequent available assessment positions in the rotation order, remove assessment positions that overlap with the source positions corresponding to the contamination lock table, retain the earliest available subsequent assessment position with the smallest assessment position number as the isolation retest position, and write the student to be tested into the isolation retest sequence corresponding to the subsequent assessment position. If all subsequent evaluation slots coincide with the location of the pollution source, the latest waiting time slot that does not coincide with the last source location of the main pollution chain will be retained before entering the originally planned evaluation slot; when the pollution lock table is empty, this evaluation will be marked as a direct evaluation, and the original rotation order and the original planned evaluation slots will remain unchanged; the output is an evaluation status table, and the written fields will include at least the student ID, evaluation status, main pollution chain number, isolation retest slot number, isolation retest order, and formal evaluation time, for S4 to read; if the current evaluation slot queuing table lacks information on subsequent available evaluation slots, this evaluation will be temporarily written to the pending scheduling status and sent back to the teaching management terminal for rescheduling; Through the above processing, S3 transforms the discrete exposure information in the exposure record table into a unique contamination determination result and an executable assessment status arrangement, so that subsequent steps no longer make a simple judgment based solely on the existence of an exposure segment, but instead make a closed-loop judgment on whether the formal assessment is contaminated based on the combined results of location contact, temporal continuity, track continuation, and path alignment, and ensure that the retest path after the contamination assessment can be directly executed. In practical application: Before the student enters assessment position 2, there are three exposure segments in the exposure record table. Two of these exposure segments originate from adjacent positions on the waiting path, appearing consecutively, and corresponding to adjacent beats of the same musical phrase. The other exposure segment originates from the teacher's position but is far from the student's entry path. The system first constructs a position correspondence matrix, a temporal difference matrix, and a musical phrase continuation matrix to form a contamination correlation matrix. Then, it extracts non-zero correlation terms to construct a contamination continuation graph. After spectral decomposition, two connected node groups are obtained. The two waiting path exposure segments are expanded into the first candidate contamination chain, and the teacher's position exposure segment is expanded into the second candidate contamination chain. Subsequently, the path fit, temporal continuation, and musical phrase continuation are calculated to form a contamination projection value. When the contamination projection value of the first candidate contamination chain is higher and the last exposure time is closer to the formal assessment time, the system writes the first candidate contamination chain into the contamination locking table and adjusts the student to assessment position 4, which does not overlap with the waiting path position, for isolation retesting. This avoids the contaminated formal assessment results directly entering the subsequent learning update.

[0020] S4. Obtain the performance data of the student under test in the order of direct assessment or isolated retest. The edge computing node binds the performance data with the assessment status table. During the contaminated assessment, the performance data is retained as a practice record. During the direct assessment, the pitch completion, rhythm completion and paragraph completion are extracted, and the valid assessment results are output. In this implementation, the purpose of S4 is to transform the performance data generated by the student under direct assessment or isolated retest sequence into effective assessment results that can be used for learning progress updates, and to strictly separate contaminated assessment from direct assessment. The processing logic is as follows: First, the performance data for the corresponding assessment round is read according to the assessment status table and divided into performance segments. Then, each performance segment is paired with a standard repertoire segment sequence of the assessment piece, calculating pitch deviation, rhythm deviation, and segment coverage. Subsequently, the results are distributed according to the assessment status. In contaminated assessment, only practice records are retained; in direct assessment, the results are summarized to form an effective evaluation table. Finally, pitch completion, rhythm completion, and segment completion are generated and written into the effective assessment result table. This implementation process includes the following steps: The purpose of S41 is to organize continuous performance data into a performance segment table that can be directly read for subsequent segment-by-segment evaluation. The input consists of performance data and evaluation status table generated by the student under direct evaluation or isolated re-evaluation. The evaluation status table includes at least the student identifier, evaluation status, evaluation order, formal evaluation time, and evaluation repertoire identifier. The performance data consists of the performance audio stream collected during the formal evaluation period or the transcribed note event stream. The processing steps are as follows: first, the performance data for the corresponding evaluation round is read from the local cache of the edge computing node according to the student identifier and evaluation order, and then the data is divided into performance segments according to the start and end times of the notes. When the performance data is a note event stream, the performance segments are directly divided with the start and end times of adjacent note events as boundaries. When the performance data is a performance audio stream, note start and end recognition is performed first, and then performance segments are formed from continuous note intervals. If there are only pauses shorter than the preset pause length in the same continuous performance, the notes before and after the pause will be merged into the same performance segment. The preset pause length is pre-written by the system configuration table and is determined based on the shortest pause length allowed within the same phrase in piano teaching. Then, a performance segment sequence is created according to the order in which the performance segments appear. The output is a performance segment table, and the written fields include at least the performance segment number, performance start time, performance end time, note sequence within the performance segment, and performance segment sequence number, for S42 to read. If the performance data is missing or the evaluation status table cannot match the performance data, an empty performance segment table is generated and an exception flag is written, and it will not be included in the subsequent valid evaluation. The purpose of S42 is to match each performance segment with a standard repertoire segment and generate quantifiable segment-level evaluation results. The input consists of a performance segment table and a sequence of standard repertoire segments corresponding to the evaluation repertoire. The standard repertoire segment sequence is divided according to the boundaries of the musical score and the boundaries of the teaching phrases when the repertoire is entered. Each standard repertoire segment includes at least a standard pitch sequence, a standard beat sequence, and a standard paragraph position range. The processing steps are as follows: First, each performance segment is matched with the standard repertoire segment sequence in the order of the performance segments. The matching rules are: first, standard repertoire segments with the same starting pitch and corresponding starting beat are retained; when a performance segment can correspond to multiple standard repertoire segments, the standard repertoire segment with the most overlapping pitch sequences is retained; when the number of overlaps is the same, the standard repertoire segment with the smaller difference in starting beat position is retained; after the matching is completed, the pitch deviation, beat deviation, and paragraph coverage are calculated for each performance segment. Among them, pitch deviation is the number of inconsistent notes produced after comparing the pitch sequence of the performance segment with the pitch sequence of the corresponding standard repertoire segment in sequence; beat deviation is the cumulative absolute value of the difference between the actual landing time of each note in the performance segment and the landing time of the standard beat; segment coverage is the number of consecutive note positions within the standard repertoire segment covered by the performance segment; the output is the performance evaluation table, and the written fields include at least the performance segment number, the corresponding standard repertoire segment number, pitch deviation, beat deviation, and segment coverage, for S43 to read; if a performance segment cannot be paired with any standard repertoire segment, the performance segment is written to the unpaired performance item, and an empty pairing identifier is written in the performance evaluation table; The purpose of S43 is to divert performance evaluation results into practice records or valid evaluation results based on the evaluation status, avoiding contamination of evaluation results in subsequent learning updates. The inputs are the performance evaluation form and the evaluation status form. The processing actions are as follows: first, the performance evaluation form and the evaluation status form are bound to the evaluation status form according to the student's identifier and the evaluation order; when the evaluation status form corresponds to a contaminated evaluation, the pitch deviation, rhythm deviation, and section coverage of each performance segment, along with the performance segment number, the evaluation piece identifier, and the evaluation order, are written into the practice record form, and a contaminated evaluation identifier is written into the practice record form. A valid evaluation form is not generated. When the evaluation status table corresponds to direct evaluation, the pitch deviation, beat deviation, and segment coverage for each performance segment are summarized and calculated. The pitch deviation is summed across all performance segments to obtain the total pitch deviation, the beat deviation is summed across all performance segments to obtain the total beat deviation, and the segment coverage is accumulated based on the deduplicated standard repertoire segment positions to obtain the total segment coverage. The output is a valid evaluation table or a practice record table. The valid evaluation table includes at least the total pitch deviation, total beat deviation, total segment coverage, and total number of performance segments for S44 to read. If there are unpaired performance items under direct evaluation, the number of unpaired performance items is simultaneously written into the valid evaluation table for consideration when generating completion status later. The purpose of S44 is to convert segment-level or summary-level assessments into completion results that can be directly used for updating learning progress; the inputs are valid assessment forms, standard repertoire segment sequences, and assessment status tables; the processing actions are: when valid assessment forms exist, to generate pitch completion status, rhythm completion status, and segment completion status respectively; The pitch completion is calculated based on the total pitch deviation and the total number of notes in the standard piece, using the correspondence between the total number of notes in the standard piece and the total pitch deviation. The beat completion is calculated based on the total beat deviation and the total number of standard beats, using the distribution of the total beat deviation across the total number of standard beats. The segment completion is calculated based on the total segment coverage and the total number of segments in the standard piece, using the ratio of the total segment coverage to the total number of segments in the standard piece. The pitch completion, beat completion, and segment completion are then written into the valid evaluation result table along with the student identifier, the piece identifier, and the evaluation order. The output is the valid evaluation result. The valid evaluation result table must include at least the student identifier, the piece identifier, the evaluation order, pitch completion, beat completion, and segment completion for S5 to read. If S43 does not generate a valid evaluation table, S44 will not write to the valid evaluation result table, but will only retain the practice record table for later retrieval. Through the above processing, S4 achieves a complete conversion of performance data into valid assessment results and strictly isolates contaminated assessments from direct assessments. This ensures that subsequent updates to the learning profile only read uncontaminated valid assessment results, preventing practice performances or contaminated performances from entering the grade assessment. In practical application: After the student completes the formal assessment in the isolated retest sequence, the edge computing node reads the audio stream of that performance and divides it into multiple performance segments. Then, each performance segment is paired with a standard repertoire segment, and the pitch deviation, beat deviation, and segment coverage are calculated respectively. If the assessment status is a direct assessment, the system summarizes the three types of quantities for all performance segments, generates pitch completion status, beat completion status, and segment completion status, and writes them into the valid assessment result table. If the assessment status is a contaminated assessment, the system only writes the corresponding results into the practice record table and does not generate valid assessment results, thus ensuring that subsequent updates to the learning profile are only based on valid assessment results.

[0021] S5. Based on the effective assessment results and the current learning profile, perform updates. When conducting direct assessments, generate level identification, class grouping adjustments, and after-school training arrangements. When only corresponding to practice records, maintain the original level and write a retest mark, and output the results of graded teaching management. In this implementation, the purpose of S5 is to write the valid assessment results back to the current learning profile, and on this basis, form the level determination, class grouping adjustment, and after-school training arrangement. At the same time, the assessment results corresponding to the practice records are strictly separated from the assessment results corresponding to the direct assessment. The processing logic is as follows: First, update the current learning profile based on the valid assessment results and generate a profile change table. Then, form the level determination results, class grouping adjustment results, and after-school training content based on the updated learning profile and the historical profile sequence of students of the same level. Finally, execute the write-back and distribution according to the assessment result status, and output the graded teaching management results. This implementation process includes the following steps: The purpose of S51 is to update the current learning profile and generate a profile change table that can be directly read by subsequent graded processing. The inputs are valid assessment results and the current learning profile. Valid assessment results include at least the student identifier, assessment piece identifier, assessment order, pitch completion status, rhythm completion status, and paragraph completion status. The current learning profile includes at least the student identifier, current level, pitch item, rhythm item, paragraph item, and last valid update time. The processing actions are as follows: first, read the current learning profile based on the student identifier, then determine whether the valid assessment results correspond to direct assessment. When the valid assessment results correspond to direct assessment, write the pitch completion status, rhythm completion status, and paragraph completion status into the corresponding pitch item, rhythm item, and paragraph item of the current learning profile, respectively, and write the current assessment order into the most recent valid assessment sequence number field and the current assessment time into the most recent valid update time field. Then, by subtracting the previous image field value from the current image field value after the current assessment, the pitch difference, rhythm difference, and paragraph difference are calculated respectively, generating an image change table; the output is an updated student image and an image change table; the image change table includes at least the student's identifier, assessment order, pitch difference, rhythm difference, and paragraph difference, for S52 to read; if any of the three completion statuses are missing in the valid assessment result, the image update is stopped and the result is written to the assessment item table to be reviewed, and it will not proceed to subsequent grading processing; The purpose of S52 is to generate a unique grade determination result, class adjustment result, and after-school training content based on the updated learning profile and profile change table. The input includes the updated learning profile, profile change table, and the historical profile sequence of students at the same grade. The historical profile sequence of students at the same grade is a set of historical valid assessment profiles that are the same as the current grade of the student to be tested. The reading range is pre-defined by the teaching management terminal as the profile records of students at the same grade who have completed direct assessment. The grade profile interval is formed by the value boundaries of the pitch, rhythm, and paragraph items of the historical profile set. The processing actions are as follows: First, the pitch, rhythm, and paragraph items in the updated learning profile are matched item by item with the historical profile sequence of students at the same grade to determine whether all three fields fall within the same grade profile interval. When the pitch, rhythm, and paragraph items all fall within the current grade profile interval, a grade determination result that maintains the current grade is generated. When all three fields fall within the next grade profile interval, a grade determination result that adjusts to the next grade is generated. When the three fields do not fall within the same grade profile interval simultaneously, the original grade is retained and a grade pending confirmation flag is written. Subsequently, class reorganization results are generated based on the correspondence between the grade assessment results and the current class group profile range. The current class group profile range consists of the grade range and profile item range pre-registered by each class group. When the grade assessment result maintains the current grade and the updated learning profile falls within the current class group profile range, a class reorganization result is generated to maintain the original class. When the grade assessment result maintains the current grade but the updated learning profile exceeds the current class group profile range, a class reorganization result is generated to adjust to another class group of the same grade. When the grade assessment result is to adjust to the next grade, a class reorganization result is generated to adjust to a class group of the next grade. Then, after-school training content is allocated according to the profile item with the largest difference in the profile change table. Specifically, when the pitch item has the largest difference, pitch training content is read from the training content table. When the difference in the beat item is the largest, read the beat training content; when the difference in the paragraph item is the largest, read the paragraph training content; when there are two or more differences that are the same and simultaneously the largest, retain the unique training content according to the preset priority order of pitch item, beat item, and paragraph item; the output is a graded processing table; the graded processing table includes at least the student identification, grade determination result, class adjustment result, and after-class training content for S53 to read; if the historical profile sequence of students of the same grade is empty, keep the original grade, keep the original class, and assign after-class training content according to the profile item with the largest difference in the profile change table; The purpose of S53 is to complete the final assignment and write back based on the status of valid assessment results, ensuring that direct assessment results enter the graded teaching management result table, and that practice records only trigger the "to be retested" flag. The inputs are valid assessment results, current student profile, and graded processing table. The processing actions are as follows: First, perform status assignment based on valid assessment results; when valid assessment results only correspond to practice records, keep the original level unchanged, and write the "to be retested" flag to the current student profile, along with the "to be retested" generation time and the corresponding assessment order, without reading the graded processing table or generating level determination results and class adjustment results; when valid assessment results correspond to direct assessment, read the level determination results, class adjustment results, and after-class training content from the graded processing table, and write these results, along with the student flag, assessment piece flag, and assessment order, into the graded teaching management result table, and clear the existing "to be retested" flag in the current student profile. If the current learning profile previously had a "to be retested" flag, this flag and its corresponding testing order will be deleted simultaneously after the current direct assessment results are written. The output is the tiered teaching management results. The tiered teaching management results table should include at least the student flag, assessment piece flag, testing order, level determination result, class adjustment result, after-class training content, and "to be retested" status, for the teaching management end to read. If any result item is missing from the tiered processing table, writing to the tiered teaching management results table will stop and this processing will be written to the tiered pending review table. Through the above processing, S5 completes the closed-loop update from effective assessment results to graded teaching management results, so that the pitch completion, rhythm completion and paragraph completion can be stably written back to the current learning situation profile, and based on the comparison of the profiles of students of the same level, an executable level identification, class adjustment and after-class training arrangement can be formed, while ensuring that the practice records do not directly enter the level processing link. In practical application: After a student receives a direct assessment and a valid assessment result is generated, the system first writes the pitch, rhythm, and paragraph completion status into the current learning profile and calculates the differences in pitch, rhythm, and paragraph before and after the assessment. Then, it reads the historical profile sequence of students at the same level, determines that the updated pitch, rhythm, and paragraph items are all within the profile range of the next level, generates a level assessment result adjusting to the next level, and generates corresponding class adjustment results based on the profile range of the next level class. Since the rhythm difference is the largest, the rhythm training content is read from the training content table and written into the grading processing table. Finally, the level assessment result, class adjustment result, and after-class training content are written into the grading teaching management result table. If the result only corresponds to a practice record, the system does not perform level assessment and class adjustment, but only writes a "to be retested" mark in the current learning profile, waiting for subsequent direct assessment results to overwrite it.

[0022] Furthermore, the present invention also includes a piano graded teaching management system based on student learning profiles, the system comprising a data acquisition and segmentation module, an identification module, a judgment module, a result extraction module, and a graded management module: The data acquisition and segmentation module is used to collect the on-site sound, rotation order, test songs, and current learning status profile of the students before they enter the room. The edge computing nodes segment the on-site sound according to the start and end times of the sound and mark the source location, and output a sound segment table before the test. The identification module is used to perform segment-by-segment correspondence with the test track based on the pre-test sound segment list. It determines the exposed segment and records the source location, occurrence time and corresponding track position according to at least one of the following: consistent pitch order, consistent beat placement order, consistent starting fingering command and consistent error correction language. It then outputs the exposure record table. The judgment module is used to perform contamination judgment based on the exposure record table and the rotation order. If there is an exposure segment before the formal assessment, the current assessment is marked as a contamination assessment and an isolation retest order is generated. If there is no exposure segment, the current assessment is marked as a direct assessment and the assessment status table is output. The results extraction module is used to obtain the performance data formed by the student under the direct assessment or isolated retest sequence. The edge computing node binds the performance data with the assessment status table. During the contaminated assessment, the performance data is retained as a practice record. During the direct assessment, the pitch completion, rhythm completion and paragraph completion are extracted, and the valid assessment results are output. The graded management module updates based on valid assessment results and current student profiles. During direct assessments, it generates grade identification, class grouping adjustments, and after-school training arrangements. When only corresponding practice records are recorded, it maintains the original grade and adds a retest flag, outputting the graded teaching management results.

[0023] Working Principle: This solution doesn't focus on how well the student played, but rather on determining what the student heard before the assessment. The specific process is as follows: First, the system collects audio from the waiting area, assessment area, and teacher's location before the student enters, segmenting the audio into time- and location-specific segments. Then, these segments are compared one by one with segments from the assessment piece to identify the segments directly relevant to the assessment. Finally, considering the student's waiting and entry paths and the order in which these segments appeared, the system determines whether these segments constitute a significant factor in the assessment. If auditory contamination is detected, the assessment will be changed to a retest in isolation; if not, the assessment will proceed directly. Next, the formal performance data will be evaluated in segments, extracting pitch completion, rhythm completion, and section completion. Finally, only valid assessment results that are not contaminated will be written into the student's learning profile, and based on this, grade determination, class adjustments, and after-class training arrangements will be made. In other words, this plan first addresses whether the assessment is valid, and then deals with how well the student has actually learned, avoiding mistaking results influenced by on-site auditory perception for their true level from the outset. For example, in a continuous rotation scenario between two practice rooms at a piano training institution, while student A is waiting for their lesson, the student in front of them is practicing the same piece at an adjacent assessment station, with the teacher repeatedly giving instructions such as "start with the second finger of the right hand" and "slow down and start again." The system first records these sounds from the waiting area, the assessment area, and the teacher, identifies which of these sounds correspond to student A's current assessment piece, and then determines whether these sounds have been effectively exposed along student A's waiting and entry paths. If exposure has been formed, the system does not allow student A to be assessed immediately in their original position, but instead moves them to a later, non-overlapping position for isolated retesting. If no exposure has been formed, the assessment proceeds as planned. After the formal assessment, the system analyzes student A's performance, calculates their pitch, rhythm, and section completion, and only updates student A's learning profile, decides whether to advance to a higher class, change classes, and what to focus on practicing after class if the assessment is a valid direct assessment. In this way, the grading results obtained by the institution are no longer temporary results popped up after hearing the prompts on the spot, but rather results that are closer to the student's true and stable level.

[0024] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A piano graded instruction management method based on student learning profiles, characterized in that, include: S1. Collect the on-site sound, rotation order, test songs and current learning status profile of the students to be tested before they enter the venue. The on-site sound is segmented by the edge computing node according to the start and end times of the sound and the source location is marked. Output the sound segment table before the test. S2. Based on the pre-evaluation sound segment table and the evaluation track, perform segment-by-segment correspondence, and determine the exposed segment according to at least one of the following: consistent pitch order, consistent beat placement order, consistent starting fingering command, and consistent error correction language. Record the source location, occurrence time, and corresponding track position, and output the exposure record table. S3. Perform contamination determination based on the exposure record table and rotation sequence. If there is an exposure segment before the formal assessment, mark this assessment as a contamination assessment and generate an isolation retest sequence. If there is no exposure segment, mark this assessment as a direct assessment and output the assessment status table. S4. Obtain the performance data of the student under test in the order of direct assessment or isolated retest. The edge computing node binds the performance data with the assessment status table. During the contaminated assessment, the performance data is retained as a practice record. During the direct assessment, the pitch completion, rhythm completion and paragraph completion are extracted, and the valid assessment results are output. S5. Based on the effective assessment results and the current learning situation profile, perform updates. When conducting direct assessments, generate level identification, class grouping adjustments, and after-school training arrangements. When only corresponding practice records are recorded, maintain the original level and write a retest mark, and output the results of graded teaching management.

2. The piano graded teaching management method based on student learning profiles according to claim 1, characterized in that: S1 includes: S11. Read the live sound during the preset time period before the students enter the room according to the rotation order of the students to be tested, and divide the continuous live sound into multiple sound segments according to the start time and end time of the sound, and output the sound segment sequence. S12. Extract the start time, duration, and arrival time difference from the sound source to the edge computing node for each sound segment in the sound segment sequence, and perform corresponding judgments based on the waiting area location, assessment area location, and teacher location, and output the source location marker corresponding to each sound segment. S13. Write the sound segment sequence, the source location mark corresponding to each sound segment, the rotation order of the students to be tested, the test pieces, and the current learning situation profile into the same pre-test sound segment table in the order in which the sound segments appear, and output the pre-test sound segment table to be read in subsequent steps.

3. The piano graded teaching management method based on student learning profiles according to claim 2, characterized in that: S2 includes: S21. Establish a segment-by-segment pairing table between each sound segment in the pre-evaluation sound segment table and the segment sequence of the evaluation piece according to the order of their appearance. Extract the pitch order field, beat placement field, command text field, and error correction text field for each sound segment. Extract the standard pitch order field, standard beat placement field, standard starting fingering command field, and standard error correction term field for each piece segment. Output a candidate correspondence table of sound segment and piece segment. S22. For each candidate corresponding item in the candidate correspondence table of sound segment-track segment, perform pitch order verification, beat placement verification, starting fingering command verification, and error correction term verification respectively. Write the corresponding verification mark when any verification is true, write a strong correspondence mark when two or more verifications are true at the same time, write a continuation correspondence mark when only a single verification is true and adjacent sound segments point to the same track segment, and write a removal mark when none of the verifications are true. Output the corresponding mark table.

4. The piano graded teaching management method based on student learning profiles according to claim 3, characterized in that: S2 also includes: S23. Based on the corresponding tag table, perform a timing continuity check on sound segments with the same source location and consecutive occurrence times. Write a continuation exposure tag for candidate corresponding items that are consecutive and point to the same track location. Perform conflict check on candidate corresponding items with different source locations but pointing to the same track location. If the number of verification tags for the later sound segment is greater than that for the earlier sound segment, retain the later sound segment and write an overwrite tag. If the number of verification tags is the same and the source location is closer to the entrance path of the student to be tested, retain the corresponding sound segment and write a priority tag. Output the exposure candidate table. S24. Based on the exposure candidate table, summarize the sound segments with strong corresponding markers, successive corresponding markers, continuous exposure markers, overlay markers, or priority markers. Write the source location, occurrence time, corresponding track location, establishment marker type, and preceding and following successive relationship of the corresponding sound segment into the exposure record table. Write the candidate corresponding items not written into the exposure record table into the item to be reviewed table and output the exposure record table.

5. The piano graded teaching management method based on student learning profiles according to claim 4, characterized in that: S3 includes: S31. Based on the source location, occurrence time, corresponding song position and sequential relationship of each exposure segment in the exposure record table, construct a position correspondence matrix according to the same position relationship between the source location and the waiting path and entry path of the student to be tested, construct a time difference matrix according to the difference in occurrence time of each exposure segment, and construct a song continuation matrix according to the same and adjacent relationships of the corresponding song positions of each exposure segment. Then, perform a position-by-position multiplication of the position correspondence matrix, time difference matrix and song continuation matrix to output the pollution correlation matrix. S32. Extract non-zero association items from the pollution association matrix according to row and column positions. Take the exposure segment corresponding to each non-zero association item as a node and construct a pollution continuation graph with the conditions that the source position is continuous, the occurrence time is progressive, and the corresponding track position is the same or adjacent. Then perform spectral decomposition on the adjacency matrix corresponding to the pollution continuation graph, extract connected node groups, and expand them into candidate pollution chains according to the occurrence time order within the node group. Output the candidate pollution chain set.

6. The piano graded teaching management method based on student learning profiles according to claim 5, characterized in that: S3 also includes: S33. For each candidate contamination chain in the candidate contamination chain set, the source location sequence of each exposed segment is aligned with the waiting path sequence and entry path sequence of the students to be tested. The occurrence time sequence of each exposed segment is summed by difference. The adjacent displacement of the corresponding track position sequence of each exposed segment is accumulated to form path fit amount, temporal continuity amount and track continuity amount. The path fit amount, temporal continuity amount and track continuity amount are then combined to form a contamination feature vector. Orthogonal projection is performed on the contamination feature vector to output the contamination projection value corresponding to each candidate contamination chain.

7. The piano graded teaching management method based on student learning profiles according to claim 6, characterized in that: S3 also includes: S34. Sort the pollution projection values ​​of each candidate pollution chain in descending order. When the pollution projection values ​​are the same, retain the candidate pollution chain whose last exposure time is closer to the formal evaluation time. When the last exposure time is the same, retain the candidate pollution chain with a shorter path length from the source location to the entry location. When the path length is the same, retain the candidate pollution chain with a larger number of corresponding track locations covered. Write the retention results into the pollution locking table and write the unretained results into the auxiliary pollution table. S35. Generate an assessment status table based on the pollution lock table. When the pollution lock table is not empty, mark this assessment as a pollution assessment and adjust the students to be tested to subsequent assessment positions that do not overlap with the source positions corresponding to the pollution lock table to form an isolated retest order. When the pollution lock table is empty, mark this assessment as a direct assessment.

8. The piano graded teaching management method based on student learning profiles according to claim 7, characterized in that: S4 includes: S41. Read the performance data and evaluation status table generated by the student under the direct evaluation or isolated re-evaluation order, divide the performance data into performance segments according to the start and end times of the notes, establish a performance segment sequence according to the order in which the performance segments appear, and output the performance segment table. S42. Perform segment-by-segment pairing between the performance segment table and the standard repertoire segment sequence corresponding to the evaluation repertoire. Calculate the pitch deviation, rhythm deviation, and segment coverage for each performance segment. Write the pitch deviation, rhythm deviation, and segment coverage into the performance evaluation table in the order of the performance segments and output the performance evaluation table. S43. Based on the evaluation status table, perform status binding on the performance evaluation table. When the evaluation status table corresponds to the pollution evaluation, write the pitch deviation, rhythm deviation and section coverage of each performance segment into the practice record table. When the evaluation status table corresponds to the direct evaluation, perform summary calculation on the pitch deviation, rhythm deviation and section coverage of each performance segment, and output the valid evaluation table. S44. Based on the valid evaluation form, generate pitch completion status, rhythm completion status, and paragraph completion status respectively. Write the pitch completion status, rhythm completion status, and paragraph completion status, along with the student identification, the evaluation piece identification, and the evaluation order, into the valid evaluation result table, and output the valid evaluation results.

9. The piano graded teaching management method based on student learning profiles according to claim 8, characterized in that: S5 includes: S51. Read the valid assessment results and the current learning profile. When the valid assessment results correspond to the direct assessment, write the pitch completion status, rhythm completion status and paragraph completion status into the pitch item, rhythm item and paragraph item corresponding to the current learning profile, respectively. Generate a profile change table according to the field difference before and after this assessment, and output the updated learning profile and profile change table. S52. Based on the updated learning profile and profile change table, perform item-by-item correspondence for the historical profile sequence of students of the same level. When the pitch, rhythm and paragraph items all fall into the same level profile range, generate the level determination result. Then, generate the class adjustment result according to the correspondence between the level determination result and the current class profile range. Subsequently, allocate after-school training content according to the profile item with the largest difference in the profile change table and output the graded processing table. S53. Based on the valid assessment results, the status is divided. When the valid assessment results only correspond to the practice record, the original level is maintained and a retest mark is written to the current learning profile. When the valid assessment results correspond to the direct assessment, the level determination results, class adjustment results and after-class training content in the graded teaching management result table are written to the graded teaching management result table, and the graded teaching management results are output.

10. A piano graded teaching management system based on student learning profiles, used to implement the piano graded teaching management method based on student learning profiles as described in any one of claims 1-9, the system comprising a data acquisition and segmentation module, an identification module, a judgment module, a result extraction module, and a graded management module, characterized in that: The data acquisition and segmentation module is used to collect the on-site sound, rotation order, test songs, and current learning status profile of the students before they enter the room. The edge computing nodes segment the on-site sound according to the start and end times of the sound and mark the source location, and output a sound segment table before the test. The identification module is used to perform segment-by-segment correspondence with the test track based on the pre-test sound segment list. It determines the exposed segment and records the source location, occurrence time and corresponding track position according to at least one of the following: consistent pitch order, consistent beat placement order, consistent starting fingering command and consistent error correction language. It then outputs the exposure record table. The judgment module is used to perform contamination judgment based on the exposure record table and the rotation order. If there is an exposure segment before the formal assessment, the current assessment is marked as a contamination assessment and an isolation retest order is generated. If there is no exposure segment, the current assessment is marked as a direct assessment and the assessment status table is output. The results extraction module is used to obtain the performance data formed by the student under the direct assessment or isolated retest sequence. The edge computing node binds the performance data with the assessment status table. During the contaminated assessment, the performance data is retained as a practice record. During the direct assessment, the pitch completion, rhythm completion and paragraph completion are extracted, and the valid assessment results are output. The graded management module updates based on valid assessment results and current student profiles. During direct assessments, it generates grade identification, class grouping adjustments, and after-school training arrangements. When only corresponding practice records are recorded, it maintains the original grade and adds a retest flag, outputting the graded teaching management results.

Images