A method for dynamic evaluation of quality of service and credit score management of home service

Abstract

The application relates to a door-to-door sports service quality dynamic evaluation and credit score management method, and belongs to the technical field of data processing and evaluation, which comprises the following steps: S1, data acquisition, real-time acquisition of physiological sign data, motion trajectory data and first-view audio and video data of a coach in a service process; S2, environment sensing, real-time acquisition of motion feedback data, environment safety data and space-time anchor point data of service start / stop of a student; S3, fusion modeling; S4, multi-dimensional evaluation; S5, score updating; and S6, early warning scheduling. The door-to-door sports service quality dynamic evaluation and credit score management method expands the evaluation dimension of sports service from single'result evaluation' to 'process evaluation' by introducing wearable devices and first-view video; by analyzing physiological indexes and video action flow of the coach, the input degree of the coach in the whole class can be objectively quantified, and the objectivity and accuracy of service quality evaluation are greatly improved.

Description

Technical Field

[0001] This invention relates to the field of data processing and evaluation technology, specifically a method for dynamic evaluation of the quality of on-site sports services and credit score management. Background Technology

[0002] Currently, with the increasing awareness of fitness among the general public, the market for in-home sports coaching services (such as in-home physical training and middle school entrance exam physical education coaching) is rapidly expanding. However, this industry currently faces the following major technical and management challenges:

[0003] First, the service quality assessment mechanism is simplistic and outdated. Traditional assessments rely heavily on user ratings after the service ends. This subjective evaluation is easily influenced by "personal favors" or "malicious negative reviews," and it fails to capture the true teaching details during the service process, such as whether the instructors were dedicated and whether the instruction was scientific and safe.

[0004] Secondly, there is a lack of a credit system. Existing platform credit scores are usually calculated based on simple cumulative consumption or positive review rates, lacking dynamic monitoring of the service process. Some coaches may experience a decline in service quality after initially obtaining high credit scores (i.e., "credit lapse"), and the existing static scoring system is unable to capture and punish such behavior in real time.

[0005] Finally, there are blind spots in safety supervision. In-home sports services take place in diverse scenarios (indoor / outdoor), and currently there is a lack of automated, objective early warning methods based on multimodal data (physiological + video + location) for sudden sports injury risks or negative teaching behaviors by coaches.

[0006] Therefore, there is an urgent need for an intelligent management method that can penetrate the service process, dynamically reflect service quality, and be anchored to the credit system. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a method for dynamic evaluation and credit score management of on-site sports service quality. It has advantages such as dynamic process supervision, proactive intervention in safety risks, and scientific decay of credit scores, solving the problems of strong subjectivity, blind spots in safety supervision, and "credit slack" caused by the solidification of the credit system in traditional on-site sports service evaluation methods.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] A method for dynamic evaluation and credit score management of on-site sports service quality includes the following steps:

[0010] S1. Data Acquisition: Deploy a multimodal data acquisition unit at the service provider end to acquire real-time physiological data, movement trajectory data, and first-person audio and video data of the coach during the service process;

[0011] S2, Environmental Awareness: Deploy smart terminals on the client side to collect real-time data on trainees' motion feedback, environmental safety, and spatiotemporal anchor points for service start / end;

[0012] S3. Fusion Modeling: Align and fuse the data obtained in steps S1 and S2 based on the timeline to construct a "digital twin of the service process";

[0013] S4. Multidimensional evaluation: Invoke the preset evaluation model to score the digital twin of the service process in multiple dimensions, including at least: teaching standardization, interaction enthusiasm, action safety, and time fulfillment.

[0014] S5. Credit score update: Based on the scoring results of step S4 and combined with the decay factor of historical credit scores, the coach's credit score is updated through a dynamic weighting algorithm and stored on the blockchain.

[0015] S6. Early Warning and Dispatch: Prioritize coaches' services and provide risk warnings based on updated credit scores.

[0016] Furthermore, the physiological data in S1 includes heart rate variability (HRV) and blood oxygen saturation, which are used to determine the coach's teaching engagement and fatigue level by analyzing the physiological fluctuations of the coach at a specific teaching moment.

[0017] Furthermore, the evaluation model in S4 is a hybrid model based on Temporal Convolutional Network (TCN). This model uses the commitment statement voice at the start of the service as the starting anchor point and the student reaching the predetermined exercise heart rate as the ending anchor point, dynamically dividing the service stages for segmented scoring.

[0018] Furthermore, the "action safety" dimension in S4 is obtained by parsing first-person perspective videos using computer vision algorithms. Specifically, it includes detecting whether the coach intervened in advance on the trainee's dangerous movement trends (such as loss of balance or joint hyperextension). If no intervention was detected, the safety weight score for the corresponding time period is deducted.

[0019] Furthermore, the dynamic weighting algorithm in S5 is specifically as follows:

[0020]

[0021] Among them, C old For historical credit scores, α is the time decay factor (0 < α < 1), t is the time interval since the last service, and S now The service score is calculated using β as the service weight and B as the reward factor for evidence storage on the blockchain.

[0022] Furthermore, the risk warning in S6 includes: when the coach's real-time heart rate is below the resting heart rate threshold for 5 consecutive minutes and is accompanied by prolonged stillness, the system determines that he is in a "passive bystander" state, triggers a point deduction warning and pushes a reminder to the background.

[0023] Furthermore, it also includes step S7: generating a service quality report and pushing it to the client and the regulatory end. The report highlights the "high-interaction periods" and "high-risk periods" in the service process in the form of a heat map.

[0024] Furthermore, the environmental safety data in S2 includes information on site flatness and altitude changes obtained through client GPS and barometer, which is used to assist in judging the safety of physical education teaching in complex outdoor terrain.

[0025] This invention also provides a dynamic evaluation and credit score management system for the quality of on-site sports services, the system comprising:

[0026] The wearable device module for coaches is used to collect coaches' physiological data, movement trajectories, and first-person audio and video data.

[0027] The client-side APP module is used to collect students' exercise feedback, environmental safety data, and service anchor point information;

[0028] The cloud-based processing and evaluation module is used for data fusion, calling evaluation models to perform multi-dimensional scoring, and executing dynamic weighted integral algorithms;

[0029] It also includes a blockchain evidence storage module, which is used to store core assessment data and credit score change records on the blockchain.

[0030] Compared with existing technologies, this invention provides a method for dynamic evaluation of the quality and credit score management of on-site sports services, which has the following beneficial effects:

[0031] 1. This method for dynamic evaluation and credit score management of in-home sports service quality expands the evaluation dimensions of sports services from a single "outcome evaluation" to "process evaluation" by introducing wearable devices and first-person perspective videos. By analyzing the coach's physiological indicators (such as heart rate variability) and video motion flow, it can objectively quantify the coach's level of engagement throughout the lesson. For example, the system can automatically identify whether the coach is playing on their phone while the student is doing high-intensity interval training (through static state detection), completely eliminating the drawbacks of the traditional scoring model of "not being able to see or explain," and greatly improving the objectivity and accuracy of service quality evaluation.

[0032] 2. This method for dynamic evaluation and credit score management of on-site sports services innovatively introduces a credit score algorithm with a time decay factor. Traditional credit scores only increase or slightly decrease, but in this scheme, if a coach does not accept orders for an extended period or their recent service quality score declines, their historical credit score will automatically decay (α in the formula). t This forces coaches to continuously provide high-quality services in order to maintain a high credit rating, effectively solving the problem of insufficient motivation for later services due to credit accumulation, and building a virtuous cycle of coach ecosystem for the platform.

[0033] 3. This method for dynamic evaluation and credit score management of in-home sports services utilizes computer vision technology to analyze the coach's first-person perspective. This invention enables "pre-incident warning" rather than "post-incident accountability" of sports injury risks. When a student exhibits dangerous tendencies such as distorted movements or instability, the system detects whether the coach has issued the correct intervention instruction or protective action within a preset golden reaction time window (e.g., 1.5 seconds). If the coach fails to respond correctly, the system will immediately provide a local voice prompt and record it as a safety deduction. This mechanism combines the recognition capabilities of artificial intelligence with the on-site execution ability of human coaches, significantly reducing the risk of accidental injuries in in-home sports services.

[0034] 4. This dynamic evaluation and credit score management method for on-site sports services ensures the immutability and traceability of data by storing core service evaluation data and credit score change records on the blockchain. The generated "digital twin of the service process" is not only a scoring basis but also a legally valid service record. For regulatory authorities or insurance companies, these objective and continuous data streams (such as heart rate curves and motion compliance heatmaps) provide a solid data foundation for handling service disputes, formulating industry standards, and developing customized insurance business, thus promoting the standardization, transparency, and intelligence of the on-site sports service industry. Attached Figure Description

[0035] Figure 1 This is a flowchart of a method for dynamic evaluation of the quality and credit score management of on-site sports services according to the present invention.

[0036] Figure 2 This is a system block diagram of a method for dynamic evaluation of the quality of door-to-door sports services and credit score management according to the present invention. Detailed Implementation

[0037] 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.

[0038] Please see Figure 1 This invention provides a method for dynamic evaluation and credit score management of on-site sports service quality, comprising the following steps:

[0039] S1. Data Acquisition: Deploy a multimodal data acquisition unit at the service provider end to acquire real-time physiological data, movement trajectory data, and first-person audio and video data of the coach during the service process;

[0040] S2, Environmental Awareness: Deploy smart terminals on the client side to collect real-time data on trainees' motion feedback, environmental safety, and spatiotemporal anchor points for service start / end;

[0041] S3. Fusion Modeling: Align and fuse the data obtained in steps S1 and S2 based on the timeline to construct a "digital twin of the service process";

[0042] S4. Multidimensional assessment: Call the pre-set assessment model to score the digital twin of the service process in multiple dimensions, including at least: teaching standardization, interaction enthusiasm, action safety, and time commitment.

[0043] S5. Credit score update: Based on the scoring results of step S4 and combined with the decay factor of historical credit scores, the coach's credit score is updated through a dynamic weighting algorithm and stored on the blockchain.

[0044] S6. Early Warning and Dispatch: Prioritize coaches' services and provide risk warnings based on updated credit scores.

[0045] Specifically, the physiological data in S1 includes heart rate variability (HRV) and blood oxygen saturation, which are used to determine a coach's teaching engagement and fatigue by analyzing physiological fluctuations at specific teaching moments.

[0046] Specifically, the risk warnings in S6 include: when a coach's real-time heart rate is below the resting heart rate threshold for 5 consecutive minutes and is accompanied by prolonged stillness, the system determines that he is in a "passive bystander" state, triggers a point deduction warning and pushes a reminder to the backend.

[0047] Specifically, it also includes step S7: generating a service quality report and pushing it to the client and the regulatory end. The report highlights the "high-interaction periods" and "high-risk periods" in the service process in the form of a heat map.

[0048] Specifically, the environmental safety data in S2 includes information on site flatness and elevation changes obtained through client GPS and barometers, which is used to help determine the safety of physical education teaching in complex outdoor terrain.

[0049] The technical effects of the present invention will be further illustrated below through specific embodiments.

[0050] Example 1: System Composition and Data Acquisition Details

[0051] Please see Figure 2 In this embodiment, the on-site sports service quality dynamic evaluation and credit score management system consists of the following core modules:

[0052] Coach-side wearable device module: This module is integrated into a pair of smart AR glasses and a smart sports watch.

[0053] AR glasses: Built-in 8-megapixel wide-angle camera (capturing first-person view video at 30fps), bone conduction microphone (capturing coach's voice commands and ambient sound), and nine-axis IMU sensor (capturing head movement trajectory to help determine the coach's focus). Edge computing chips (such as the Snapdragon XR2 platform) are responsible for initial human skeleton point recognition locally, reducing the bandwidth pressure on cloud transmission.

[0054] Smart sports watch: Built-in PPG photoelectric sensor (collects heart rate, heart rate variability (HRV), and blood oxygen saturation (SpO2), sampling rate 50Hz) and GPS module (collects movement trajectory, sampling rate 1Hz). Connects to AR glasses via Bluetooth, and uploads the aggregated data to the cloud via 4G / 5G network.

[0055] Client-side APP module: Installed on the student's smartphone. It is responsible for collecting heart rate data fed back by the student via Bluetooth heart rate monitor, and for collecting environmental safety data (such as altitude changes and positioning stability in outdoor locations) using the phone's built-in GPS and barometer. The APP also provides a QR code anchoring function to start / end the service.

[0056] Cloud processing and evaluation module: Deployed on a cloud server, it includes a data cleaning and fusion unit, a TCN-based evaluation model unit, and a dynamic weighted integral calculation unit.

[0057] Blockchain evidence storage module: Built on a consortium blockchain (such as FISCO BCOS), it stores the core summary information of each service (such as service start and end time, final score, and hash value of key security events) on the blockchain.

[0058] Example 2: Construction and Evaluation of Digital Twins of Service Processes

[0059] This embodiment describes in detail the process of service startup, data alignment, and model evaluation.

[0060] Service initiation and anchoring:

[0061] After the instructor arrives at the designated location, both parties complete the service confirmation by scanning the QR code on the instructor's AR glasses via the student's app. The system then forces the instructor into a "commitment statement" phase: the instructor must face the AR glasses' camera and clearly read aloud a randomly generated "service commitment statement" (e.g., "The person making the commitment [instructor's name] will pay full attention to the student's safety and provide professional guidance throughout this course"). The start time of this voice segment is precisely recorded as time anchor point A (T0). Simultaneously, the wearable device begins collecting data at full speed.

[0062] Building a "digital twin of the service process":

[0063] The cloud processing module timestamps all data streams based on T0, generating a multi-layered, traceable digital service copy.

[0064] Video stream layer: Compressed H.265 video stream.

[0065] Attitude flow layer: The time series of 3D coordinates of 17 key skeletal points of the trainee extracted from the video.

[0066] Physiological flow layer: the coach's HRV (time domain index SDNN, frequency domain index LF / HF) and heart rate curve.

[0067] Speech Stream Layer: The text stream after speech-to-text (ASR) processing, with sentiment annotation.

[0068] Environmental stratum: GPS track and elevation change curve.

[0069] TCN Model Segmented Scoring (Key Innovation):

[0070] The model first identifies the termination anchor point B in the trainee's heart rate stream: when the trainee's heart rate first enters the preset target fat-burning zone (e.g., (220 - age) × 60% to 80%) and remains there for more than 30 seconds, the system determines that effective training has begun and marks it as B. The entire service process is dynamically segmented by the model into "preparation phase (T0-B)," "core training phase (B to heart rate decline)," and "relaxation phase."

[0071] In terms of teaching standardization: within the "core training phase" slice, the density of technical terms such as "core engagement," "knee stability," and "breathing" in the ASR results is statistically analyzed. If the density is less than 0.5 times / minute, the corresponding score is deducted.

[0072] In terms of motion safety (core innovation):

[0073] The model monitors the trainee's skeletal points in real time through the posture flow layer. For example, when performing a "lunge squat," the system sets the standard knee angle range to 90°-110°.

[0074] Event triggered: At 15 minutes and 23 seconds, the system detected that the trainee's left knee angle was less than 80° and accompanied by a torso forward tilt angle greater than 45°, which was judged as a "high-risk posture" (imminent loss of balance).

[0075] Searching for Coach Response: The system then searches for the coach's first-person video stream, IMU data, and audio stream within a 1.5-second time window.

[0076] Ideal response: The video stream detects the instructor's hand skeletal points rapidly moving towards the student (protective action), and / or the audio stream identifies keywords such as "Careful!" or "Stay calm!" The model marks this event as an "effective safety intervention" and records it as a positive case.

[0077] Negligence Response: The instructor's skeletal points show no obvious displacement in the video stream, and there are no related instructions in the audio stream. The model marks this event as a "safety negligence event" and multiplies the safety score for that time period by a penalty coefficient less than 1 (e.g., 0.6).

[0078] Example 3: Detailed Calculation Case of Dynamic Credit Score Update

[0079] This embodiment illustrates the implementation effect of the dynamic weighting algorithm in detail through the changes in the scores of two groups of coaches.

[0080] Parameter settings:

[0081] The decay factor α = 0.985 (decays by approximately 1.5% per day);

[0082] The service weight β = 0.25 for this application.

[0083] The evidence storage reward factor B = 1 (only obtained after the service is fully uploaded and key data is on-chain);

[0084] Case 1: Li Lei, a coach with a high credit score but who is lazy.

[0085] Status: Li Lei is a gold medal coach with a historical credit score of C. old =950.

[0086] Behavior: Due to personal reasons, Li Lei did not accept any orders for 15 days (t=15). His performance in his first class after returning to work was mediocre, receiving a grade of S. now =80.

[0087] calculate:

[0088] Cnew =(950×0.985 15 )+(80×0.25)+1

[0089] ≈(950×0.80)+20+1

[0090] =760+20+1

[0091] =781

[0092] Results Analysis: Although Li Lei was a top-rated coach in the past, the 15-day gap in his record caused his historical points to shrink to 760. Coupled with his mediocre performance this time, his final points plummeted to 781. System Effect: This serves as a warning to all high-credit coaches that credit needs to be continuously maintained and cannot be relied upon.

[0093] Case Study 2: Han Meimei, a coach with a low credit score but a proactive attitude

[0094] Status: Han Meimei is a new coach with a historical credit score of C. old =600.

[0095] Behavior: She was very diligent, having just performed her service yesterday (t=1), and this time she was extremely attentive, earning a high score of S. now =98.

[0096] calculate:

[0097] C new =(600×0.985 1 )+(98×0.25)+1

[0098] =(600×0.985)+24.5+1

[0099] =591+24.5+1

[0100] =616.5

[0101] Results Analysis: Due to the short time interval, the historical score hardly decayed (only decreasing from 600 to 591), and with the addition of 24.5 points from this high score, the total score steadily increased to 616.5 points. System Effect: Incentivizes new coaches to quickly build credibility through high-frequency, high-quality service.

[0102] Example 4: Comparison of Experimental Data and Results

[0103] To verify the effectiveness of this invention, we collaborated with the "XX Sports" platform to select 50 full-time in-home sports coaches, randomly divided them into a control group and an experimental group, and conducted an A / B test for two months.

[0104] Control group: Using a traditional platform, the overall score is calculated solely based on users' five-star reviews after the course.

[0105] Experimental group: Wearing the smart device of this invention, and using the dynamic evaluation and points management method of this invention.

[0106] The experimental data and results are compared in Table 1.

[0107] Table 1

[0108] Average monthly effective guidance time percentage 68.5% 91.2% +33.1% Intervention rate for high-risk actions Unquantifiable 94.7% - Service complaint and dispute rate 4.8% 1.2% ↓75% High-scoring coach service quality standard deviation 85 points 22 points The standard deviation decreased significantly. Student satisfaction rating 86.3 points 95.7 points +10.9%

[0109] In summary, this invention effectively addresses the regulatory blind spots and credit challenges in the on-demand sports service industry by constructing a "digital twin of the service process" and introducing AI-driven security assessment and dynamic decay credit scoring algorithms. It significantly improves service quality and security and has extremely high industrial application value.

[0110] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0111] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for dynamic evaluation and credit score management of on-site sports service quality, characterized in that, Includes the following steps: S1. Data Acquisition: Deploy a multimodal data acquisition unit at the service provider end to acquire real-time physiological data, movement trajectory data, and first-person audio and video data of the coach during the service process; S2, Environmental Awareness: Deploy smart terminals on the client side to collect real-time data on trainees' motion feedback, environmental safety, and spatiotemporal anchor points for service start / end; S3. Fusion Modeling: Based on the timeline, align and fuse the data obtained in steps S1 and S2 to construct a "digital twin of the service process"; S4. Multidimensional evaluation: Invoke the preset evaluation model to score the digital twin of the service process in multiple dimensions, including at least: teaching standardization, interaction enthusiasm, action safety, and time fulfillment. S5. Credit score update: Based on the scoring results of step S4 and combined with the decay factor of historical credit scores, the coach's credit score is updated through a dynamic weighting algorithm and stored on the blockchain. S6. Early Warning and Dispatch: Prioritize coaches' services and provide risk warnings based on updated credit scores.

2. The method for dynamic evaluation and credit score management of on-site sports service quality according to claim 1, characterized in that, The physiological data in S1 include heart rate variability (HRV) and blood oxygen saturation, which are used to determine the coach's teaching engagement and fatigue by analyzing the physiological fluctuations of the coach at a specific teaching moment.

3. The method for dynamic evaluation and credit score management of on-site sports service quality according to claim 1, characterized in that, The evaluation model in S4 is a hybrid model based on Temporal Convolutional Network (TCN). This model uses the commitment statement at the start of the service as the starting anchor point and the student reaching the predetermined exercise heart rate as the ending anchor point, dynamically dividing the service into segments for scoring.

4. The method for dynamic evaluation and credit score management of on-site sports service quality according to claim 1, characterized in that, The "action safety" dimension in S4 is obtained by analyzing first-person perspective videos using computer vision algorithms. Specifically, it includes detecting whether the coach intervened in advance on the trainee's dangerous movement trends (such as loss of balance or joint hyperextension). If no intervention was detected, the safety weight score for the corresponding time period is deducted.

5. The method for dynamic evaluation and credit score management of on-site sports service quality according to claim 1, characterized in that, The dynamic weighting algorithm in S5 is specifically as follows: Among them, C old For historical credit scores, α is the time decay factor (0 < α < 1), t is the time interval since the last service, and S now The service score is calculated using β as the service weight and B as the reward factor for evidence storage on the blockchain.

6. The method for dynamic evaluation and credit score management of on-site sports service quality according to claim 1, characterized in that, The risk warning in S6 includes: when the coach's real-time heart rate is below the resting heart rate threshold for 5 consecutive minutes and is accompanied by prolonged stillness, the system determines that he is in a "passive bystander" state, triggers a point deduction warning and pushes a reminder to the background.

7. The method for dynamic evaluation and credit score management of on-site sports service quality according to claim 1, characterized in that, It also includes step S7: generating a service quality report and pushing it to the client and the regulatory end. The report highlights the "high-interaction period" and "high-risk period" in the service process in the form of a heat map.

8. The method for dynamic evaluation and credit score management of on-site sports service quality according to claim 1, characterized in that, The environmental safety data in S2 includes information on site flatness and altitude changes obtained through client GPS and barometers, which is used to help determine the safety of physical education teaching in complex outdoor terrain.

9. A dynamic evaluation and credit score management system for the quality of on-site sports services, characterized in that, The system is used to implement the method for dynamic evaluation of the quality and credit score management of on-site sports services according to any one of claims 1-8, the system comprising: The wearable device module for coaches is used to collect coaches' physiological data, movement trajectories, and first-person audio and video data. The client-side APP module is used to collect students' exercise feedback, environmental safety data, and service anchor point information; The cloud-based processing and evaluation module is used for data fusion, calling evaluation models to perform multi-dimensional scoring, and executing dynamic weighted integral algorithms; It also includes a blockchain evidence storage module, which is used to store core assessment data and credit score change records on the blockchain.

Images