A data processing method, device and medium for housekeeping training

Abstract

The application discloses a data processing method and device for housekeeping training and a medium, belongs to the technical field of data processing, and is used for solving the problem of low housekeeping training efficiency. The method comprises the following steps: determining an online training course for collectively training multiple trainees of the same professional category in a fixed time period; determining a professional skill score of each trainee for the online training course according to the online training course and uploaded personal information; obtaining multiple trainee combinations by respectively dividing multiple professional skill scores into pre-set professional skill score intervals; obtaining a video image of each trainee in a pre-set period, and generating an attention score of each trainee for the online training course; determining a training expectation level of each trainee combination according to the attention score and the professional skill score; and regarding a trainee combination lower than a pre-set level as an inactive trainee combination. The efficiency of housekeeping training can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and particularly relates to a data processing method, device and medium for housekeeping training. BACKGROUND

[0002] With the development and needs of the present era, the housekeeping service industry has also been rapidly improved. In view of the problems that the housekeeping market is not standardized, the service standard is not perfect, and the service level of service personnel is not high, the level and quality of employees need to be continuously improved, and therefore, training for users is also an essential link.

[0003] At present, in the process of housekeeping training, online training courses (theory courses) and offline training courses (operation courses) are mainly relied on by trainers to drive trainees. However, when the online training course is carried out, the real state of the user cannot be obtained, so that the training effect of the user cannot be understood in time, resulting in low training efficiency in the process of housekeeping training. SUMMARY

[0004] Embodiments of the present application provide a data processing method, device and medium for housekeeping training, which are used to solve the problem of low training efficiency in the process of housekeeping training.

[0005] Embodiments of the present application adopt the following technical solutions:

[0006] In one aspect, the embodiments of the present application provide a data processing method for housekeeping training, which comprises: determining an online training course for collective housekeeping training of multiple trainees of the same professional category in a fixed time period, to push the online training course to a terminal device of each trainee; determining a professional skill score of each trainee for the online training course according to the online training course and personal information pre-uploaded by each trainee, through a multivariate logistic regression professional skill prediction model; grouping each trainee by dividing multiple professional skill scores into pre-set professional skill score intervals, to obtain multiple trainee combinations; obtaining video images of each trainee in a pre-set period when playing the online training course, generating an attention score of each trainee for the online training course by inputting the video images into a pre-trained neural network model; determining a training expectation level of each trainee combination according to the attention score and the professional skill score; the training expectation level is used to represent the mastery degree of the trainee combination to the training course content in the current period; in multiple trainee combinations, trainee combinations below a pre-set level are taken as inactive trainee combinations; if the online training course is a live course, generating a professional skill analysis diagram of the inactive trainee combinations; pushing the professional skill analysis diagram of the inactive trainee combinations to a trainer terminal device, so that the trainer updates a pre-set teaching progress.

[0007] In one example, the determination of the training expectation level of each trainee combination according to the attention score and the professional skill score specifically comprises: mapping the professional skill score through a first mapping relationship to determine an understanding score of each trainee for the online training course; dividing each trainee in each trainee combination according to a pre-divided attention score interval to obtain multiple cluster combinations; determining a corresponding attention parameter for each cluster combination, the attention parameter comprising age and understanding score; determining a training expectation coefficient of each cluster combination according to the attention parameter and each cluster combination; determining a training expectation weight of each cluster combination through a second mapping relationship for different attention score intervals; the higher the attention score, the greater the training expectation weight; weighting and summing the training expectation coefficient of each cluster combination and the corresponding training expectation weight to obtain a local training expectation coefficient of each trainee combination; mapping the local training expectation coefficient of each trainee combination through a third mapping relationship to determine the training expectation level of each trainee combination.

[0008] In one example, the method further comprises: dividing each of the cluster combinations into a plurality of sub-cluster combinations according to a pre-divided age interval; determining a weight of each of the sub-cluster combinations through a fourth mapping relationship for different age intervals; the higher the age, the lower the weight of each of the sub-cluster combinations; and performing weighted summation on the understanding scores of all the sub-cluster combinations in the same cluster combination and corresponding weights to obtain a local training expectation coefficient of each of the cluster combinations.

[0009] In one example, before the determining the training expectation level of each of the training trainees according to the concentration score and the professional skill score, the method further comprises: determining a concentration score sequence of each of the training trainees for the online training course; determining a concentration level of each of the training trainees when watching the online training course according to a difference between a plurality of concentration scores of each of the training trainees in adjacent periods in the concentration score sequence; if the concentration level is lower than a preset change level, retrieving a compensation degree corresponding to the concentration level according to the concentration score in a pre-constructed compensation mapping table; the concentration level and the compensation degree are negatively correlated in absolute value; and compensating the concentration score according to the compensation degree to update the concentration score.

[0010] In one example, before the generating the concentration score of each of the training trainees for the online training course by inputting the video image into a pre-trained neural network model, the method further comprises: obtaining a plurality of sample video images of sample training trainees of different professional categories; the professional categories include at least one of a mother and baby category, an elderly care category, and a cleaning category; dividing the plurality of sample video images into a plurality of sample groups according to the professional categories; selecting a preset number of training samples from each of the sample groups to obtain a training sample set; training an initial convolutional neural network model with the training sample set as an input sample and the concentration score of the sample training trainee as a sample label until a training stop condition is reached to obtain the neural network model.

[0011] In one example, after the student combination below the preset level is determined as the inactive student combination in the plurality of student combinations, the method further comprises: if the online training course is the recorded course, acquiring a video image of the trainer teaching online in the online training course in a current period, identifying the video image, and extracting voice data of the trainer when teaching online; determining a speech speed value of the trainer when teaching online according to the voice data; if the speech speed value exceeds a preset speech speed threshold, calling a training file bound to the online training course, determining corresponding training content of the video image in the training file; if a difficulty level of the training content is lower than a preset difficulty level, converting text information of the voice data into demonstration playing voice according to the preset speech speed threshold; and confirming whether each training student in the inactive student combination re-plays the video image in the current period; if yes, updating the fixed time period according to a time length of the current period to increase the viewing time length of the corresponding training student.

[0012] In one example, the method further comprises: determining an offline training course to be used for collective home training of a plurality of training students; wherein the offline training course is a practical operation course of the online training course; acquiring a training start time and a training time length of the offline training course; determining a specified training time of the training content in the offline training course according to the training start time and the training time length; determining a position parameter of the inactive training student when performing practical operation in the offline training course; and sending the training content and the position parameter to an intelligent robot when the specified training time is reached, so that the intelligent robot assists in training the practical operation action of each training student in the inactive student combination according to the training content.

[0013] In one example, the intelligent robot generates a standard practical operation action of each training student in the inactive student combination according to the training content; acquires a practical operation video image of each training student in the inactive student combination; identifies the practical operation video image to obtain a practical operation action of each training student in the inactive student combination; compares the practical operation action with the standard practical operation action to determine a matching degree of the practical operation action and the standard practical operation action; and reminds a specified training student when the matching degree is lower than a preset threshold.

[0014] In another aspect, an embodiment of the present application provides a data processing device for housekeeping training, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: determine an online training course to be collectively provided to a plurality of trainees of a same professional category in a fixed time period, and push the online training course to a terminal device of each trainee; determine a professional skill score of each trainee for the online training course according to the online training course and personal information pre-uploaded by each trainee, by using a multivariate logistic regression professional skill prediction model; group each trainee by dividing a plurality of professional skill scores into pre-set professional skill score intervals, to obtain a plurality of trainee combinations; obtain a video image of each trainee in a pre-set period while playing the online training course, and generate a concentration score of each trainee for the online training course by inputting the video image into a pre-trained neural network model; determine a training expectation level of each trainee combination according to the concentration score and the professional skill score; the training expectation level is used to represent a mastery degree of each trainee combination on training course content in a current period; in the plurality of trainee combinations, a trainee combination below a pre-set level is regarded as an inactive trainee combination; if the online training course is a live course, generate a professional skill analysis diagram of the inactive trainee combination; and push the professional skill analysis diagram of the inactive trainee combination to a trainer terminal device, so that the trainer updates a pre-set teaching progress.

[0015] In another aspect, an embodiment of the present application provides a data processing nonvolatile computer storage medium for housekeeping training, which stores computer executable instructions configured to: determine an online training course to be performed on a plurality of trainees of the same professional category in a fixed time period, and push the online training course to a terminal device of each trainee; determine a professional skill score of each trainee for the online training course according to the online training course and personal information uploaded by each trainee in advance through a multivariate logistic regression professional skill prediction model; group each trainee by dividing a plurality of professional skill scores into a pre-set professional skill score interval, to obtain a plurality of trainee combinations; obtain a video image of each trainee in a pre-set period when the online training course is played, and generate a concentration score of each trainee for the online training course by inputting the video image into a pre-trained neural network model; determine a training expectation level of each trainee combination according to the concentration score and the professional skill score; the training expectation level is used to represent a mastery degree of the trainees of each trainee combination on the training course content in the current period; in the plurality of trainee combinations, a trainee combination lower than a pre-set level is taken as an inactive trainee combination; if the online training course is a live course, a professional skill analysis diagram of the inactive trainee combination is generated; and the professional skill analysis diagram of the inactive trainee combination is pushed to a trainer terminal device, so that the trainer updates a pre-set teaching progress.

[0016] The above at least one technical solution adopted by the embodiments of the present application can achieve the following beneficial effects:

[0017] In the online training process, the professional skill score of each trainee for the online training course can be obtained through the multivariate logistic regression professional skill prediction model, the concentration score of each trainee for the online training course can be continuously generated, the training information of each trainee for the training course can be customized individually, the training expectation level of each trainee combination can be determined according to the concentration score and the professional skill score, and the inactive trainee can be extracted in time, so as to assist the trainer to understand the training effect of each trainee in real time and improve the efficiency of housekeeping training. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the present application, some embodiments of the present application will be described in detail below with reference to the accompanying drawings, in which:

[0019] Figure 1 A flowchart of a data processing method for housekeeping training provided by an embodiment of the present application;

[0020] Figure 2 A structural schematic diagram of a data processing device for housekeeping training is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0021] To make the objectives, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be described below in detail with specific embodiments and corresponding drawings. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0022] Some embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0023] Figure 1 A flowchart of a data processing method for housekeeping training is provided for an embodiment of the present application. Some input parameters or intermediate results in the flowchart allow manual intervention adjustment to help improve accuracy.

[0024] The analysis method related to the embodiments of the present application can be implemented for a terminal device or a server, and the present application does not make special limitations thereon. For the convenience of understanding and description, the following embodiments are described in detail with the terminal device as an example.

[0025] It should be noted that the server can be a separate device, or a system composed of multiple devices, i.e., a distributed server, and the present application does not make specific limitations thereon.

[0026] Figure 1 The flowchart in the above embodiment can include the following steps:

[0027] S101: Determine an online training course to be used for collective housekeeping training of multiple trainees of the same professional category in a fixed time period, so as to push the online training course to the terminal device of each trainee.

[0028] It should be noted that the online training course in the present application is not provided free of charge to any user, but is only provided free of charge to internal trainees for learning. Meanwhile, when training the trainees, the trainees are collectively trained in a fixed time period. The online training course can be a live course or a recorded course. In addition, for the recorded course, due to time limitations, the trainees can only play the course quickly during the process of watching the online training course, and cannot play it slowly.

[0029] S102: Determine the professional skill score of each training trainee for the online training course according to the online training course and the personal information uploaded in advance by each training trainee through the multivariate logistic regression professional skill prediction model.

[0030] It should be noted that when constructing the multivariate logistic regression professional skill prediction model, the multivariate logistic regression professional skill prediction model is established by multivariate logistic regression analysis, taking the sample personal information and the sample online training course as the independent variables and the sample professional skill score as the dependent variable. The sample personal information, the sample online training course and the sample professional skill score are in a nonlinear relationship. The standard of the professional skill score can be set according to actual needs, for example, it can be set between 0-100. For example, the process of establishing the multivariate logistic regression professional skill prediction model is as follows: input sample information, data preprocessing, training sample set, selection of kernel factor, selection of penalty factor and other parameters, determination of classification hyperplane, construction of decision function through optimal solution, establishment of prediction model, input of test sample set, output of prediction result, performance evaluation.

[0031] Among them, the personal information includes gender, age, work experience, occupation, education, etc. Because of the different backgrounds of users, the professional skills of each user are different. Among them, the factor that has the greatest impact on professional skill is the work experience of the user for the professional category, and the gender is the factor with the smallest impact.

[0032] For example, the professional categories of housekeeping training usually include mother and baby, old-age care and cleaning, etc., and each professional category includes multiple sub-directions. If the current online training course is part of the old-age care course, the more work experience the user has for this part of the course, the higher the professional skill score of the user.

[0033] S103: Group each training trainee by dividing the multiple professional skill scores into pre-set professional skill score intervals to obtain multiple training trainee combinations.

[0034] S104: When playing the online training course, obtain the video image of each training trainee within a pre-set period, and generate the concentration score of each training trainee for the online training course by inputting the video image into a pre-trained neural network model.

[0035] That is, for the online training course, multiple period video images will be obtained. For example, if the pre-set period is 20 minutes, the terminal playing the online training course will capture the video image within the 20 minutes and send the video image to the server.

[0036] The viewing action of the trainee can reflect the concentration degree of the trainee. For example, if the user is eating, the user is not facing the screen, the user is daydreaming, or the user is sleeping with eyes closed, it means that the trainee is not learning seriously and the concentration degree is low.

[0037] That is, the trained neural network model can perform target detection on the trainee to obtain key point information of the human body features of the trainee, perform semantic recognition on the key point information, and obtain action information of the trainee, so as to generate a corresponding concentration score of the trainee according to the action information.

[0038] It should be noted that when constructing the neural network model, first, a plurality of sample video images of sample trainees of different professional categories are obtained. The professional categories include at least one of the mother and baby category, the elderly care category, and the cleaning category. Then, the plurality of sample video images are divided into a plurality of sample groups according to the professional categories. From each sample group, a predetermined number of training samples are selected to obtain a training sample set. Finally, the initial convolutional neural network model is trained with the training sample set as the input sample and the concentration score of the sample trainee as the sample label until the training stopping condition is reached, and the neural network model is obtained.

[0039] Based on the professional category features of the training samples, the training samples are classified to obtain a plurality of sample groups with different features. Each sample group can represent a certain feature of the original sample. Then, a certain number of training samples representing different features are extracted from different sample groups as model inputs, so as to obtain a training sample set with diversity and representativeness, which can improve the generalization ability of the neural network model.

[0040] S105: According to the concentration score and the professional skill score, determine the training expectation level of each trainee combination; the training expectation level is used to represent the mastery degree of the trainee combination to the training course content in the current period.

[0041] It can be understood that during the process of watching the online training course, the mastery degree of the trainee to the training course knowledge being explained by the trainer is related to the knowledge reserve of the trainee and the concentration degree in the learning process. For example, the higher the professional skill score and the concentration score of the trainee are, the higher the mastery degree of the trainee is.

[0042] S106: Among the plurality of trainee combinations, the trainee combination below the preset level is regarded as an inactive trainee combination.

[0043] It should be noted that in the present application, a plurality of training expectation levels are included, such as a lower level, a low level, a medium level, a higher level, and a high level.

[0044] The preset level is an attribute used to measure the expected training level of a group of trainees. It can be set according to actual needs. For example, if the preset level is medium, then the expected training level will be lower than the preset level.

[0045] Among them, the inactive trainees are those who have a low grasp of the knowledge in the training course being explained, that is, they have not followed the trainer's line of thought.

[0046] S107: If the online training course is a live course, then generate a professional skills analysis chart of the inactive student group.

[0047] It should be noted that the professional skills analysis chart is used to represent the professional skills and personal information of inactive trainees. For example, the professional skills analysis chart includes several dimensions such as the trainee's professional skills score, work experience, gender, and age, so that trainers can more intuitively understand the situation of inactive trainees.

[0048] S108: Push the professional skills analysis chart of the inactive trainee group to the trainer's terminal device so that the trainer can update the pre-set teaching schedule.

[0049] For example, if a trainer finds that among inactive trainees, the percentage of those with high professional skill scores and the percentage of those with low professional skill scores are both relatively high, it indicates that the course being taught in the current cycle is too difficult for most trainees, or that the trainer's teaching method is unsuitable, resulting in a low level of mastery among the trainees. In this case, the trainer can reinforce the relevant course knowledge again at a certain time during the training period.

[0050] It should be noted that, although the embodiments in this application are based on... Figure 1 Steps S101 to S108 will be described sequentially, but this does not mean that steps S101 to S108 must be performed in a strict order. The reason this embodiment follows this order is... Figure 1 The order in which steps S101 to S108 are described is intended to facilitate those skilled in the art in focusing on the technical solutions of the embodiments of this application. In other words, in the embodiments of this application, the order of steps S101 to S108 can be appropriately adjusted according to actual needs.

[0051] pass Figure 1The method can obtain the professional skill score of each training trainee for the online training course, continue to generate the concentration score of each training trainee for the online training course, can customize the training information of each training trainee for the training course, continue to determine the training expectation level of each training trainee combination according to the concentration score and the professional skill score, can extract the inactive training trainee in time, so as to realize the training effect of each training trainee in real time, and improve the efficiency of housekeeping training.

[0052] Based on Figure 1 The method, and some specific embodiments and extended schemes of the method are also provided in the embodiments of the application, which are described below.

[0053] In some embodiments of the application, when determining the training expectation level of the training trainee combination, the training skill reserve degree of the trainees is different, and the housekeeping industry is different from other industries. The housekeeping industry has many old trainees, and memory will decrease with age, and learning and receiving ability will also decrease. Therefore, the influence of age on the training trainee's mastery is considered, and the training expectation level of the training trainee combination can be more accurately reflected.

[0054] Therefore, the professional skill score is mapped by the first mapping relationship to determine the understanding score of each training trainee for the online training course. That is, the understanding degree of the online training course under the condition of the knowledge reserve of the training trainee. The higher the professional skill score is, the higher the understanding score is.

[0055] Then, each training trainee is divided into a plurality of clustering combinations according to the pre-divided concentration score interval in each training trainee combination. It should be noted that the concentration score interval needs to be continuous. For example, 0-20, 20-40, 40-60, 60-80, and 80-100.

[0056] Then, the attention parameter of each clustering combination is determined, and the attention parameter includes age and understanding score. Then, the training expectation coefficient of each clustering combination is determined according to the attention parameter and each clustering combination.

[0057] For different concentration score intervals, the training expectation weight of each clustering combination is determined by the second mapping relationship. The higher the concentration score is, the greater the training expectation weight is. For example, the training expectation weight of the 80-100 interval is greater than that of the 60-80 interval.

[0058] Then, the training expectation coefficients of each cluster combination are weighted and summed with the corresponding training expectation weights to obtain a local training expectation coefficient of each training student combination. That is, the product of the training expectation coefficient of each cluster combination and the corresponding training expectation weight is calculated, and then each product is summed to obtain the local training expectation coefficient.

[0059] Finally, the local training expectation coefficient of each training student combination is mapped by a third mapping relationship to determine the training expectation level of each training student combination. For example, different training expectation coefficient intervals are mapped to corresponding training expectation levels.

[0060] Further, when determining the training expectation coefficient of the cluster combination according to the attention degree parameter, first, in each cluster combination, each training student is divided according to a pre-divided age interval to obtain a plurality of sub-cluster combinations. It should be noted that the age interval also needs to be continuous.

[0061] Then, for different age intervals, the weight of each sub-cluster combination is determined by a third mapping relationship. The higher the age, the lower the weight of each sub-cluster combination.

[0062] Then, the understanding scores of all sub-cluster combinations in the same cluster combination are weighted and summed with the corresponding weights to obtain a local training expectation coefficient of each cluster combination. That is, the product of the understanding score of each sub-cluster combination and the corresponding training expectation weight is calculated, and then each product is summed to obtain the local training expectation coefficient.

[0063] It should be noted that when determining the understanding score of the sub-cluster combination, the mean value between the understanding scores of the plurality of training students in the sub-cluster combination can be calculated, and the mean value is taken as the understanding score of the sub-cluster combination.

[0064] Finally, the local training expectation coefficient of each cluster combination is mapped by a fourth mapping relationship to determine the training expectation coefficient of each cluster combination. At this time, the local training expectation coefficient of each cluster combination can also be directly mapped to the training expectation coefficient of each cluster combination.

[0065] In some embodiments of the present application, the concentration of the user during the viewing process changes, especially for the parent industry group, due to age, the concentration ability is poorer, and the concentration change is more likely. Since it is always in a state of high concentration, the concentration degree is actually higher than the detection result, and the mastery of the training course knowledge is also higher, on the contrary, it is always in a state of inattention, the concentration degree is actually lower than the detection result, and the mastery of the training course knowledge is also lower.

[0066] Based on this, a concentration score sequence for the online training course is generated for each trainee, and then the concentration score of each trainee is input into the concentration score sequence.

[0067] After obtaining the concentration score each time, the concentration change level of each trainee when watching the online training course is determined according to the difference between the multiple concentration scores of each trainee in adjacent periods in the concentration score sequence.

[0068] If the concentration change level is lower than the preset change level, the compensation degree corresponding to the concentration change level is retrieved according to the concentration score in the pre-constructed compensation mapping table. The concentration change level and the compensation degree are negatively correlated. If the concentration score is higher than or equal to the preset concentration threshold, the compensation degree is positive, and if the concentration score is lower than the preset concentration threshold, the compensation degree is negative.

[0069] It should be noted that in the present application, the concentration change level includes multiple concentration change levels, which represent the concentration change degree of the trainee when watching the online training course. For example, the concentration change level is low, low, medium, high, and high.

[0070] The preset change level is an attribute for measuring the concentration change level of the trainee, which can be set according to actual needs. For example, the preset change level is medium, and when the concentration change level is low, it is lower than the preset change level.

[0071] In addition, in the present application, the concentration score includes multiple concentration scores, which represent the concentration degree of the trainee for the online training course. For example, the concentration score of trainee A is 80, the concentration score of trainee B is 90, and the concentration score of trainee C is 65.

[0072] The preset concentration threshold is an attribute for measuring the concentration score of the trainee, which can be set according to actual needs.

[0073] Finally, the concentration score is compensated according to the compensation degree to update the concentration score. For example, if the compensation degree is 10, the concentration score is increased by 10, and if the compensation degree is -10, the concentration score is decreased by 10.

[0074] If the concentration change level is higher than the preset change level, the concentration score is not compensated.

[0075] Thus, by compensating the concentration score, the concentration of the trainee can be more accurately improved, and the training expectation level of each trainee group can be more accurately reflected.

[0076] In some embodiments of the present application, the reasons for the low concentration of trainees are on the one hand the reasons of the trainees themselves, and on the other hand the problems of the teaching methods of the trainers, which lead to the trainees not understanding and reduce the learning efficiency of the trainees, so that the small actions are generated. For example, the voice of the trainer needs to be emphasized, the speed control and the volume control are required, and the language sense is adjusted as the basis, and the language sense is appropriately changed.

[0077] Based on this, if the online training course is a recording course, the video image of the trainer in the online training course is obtained, the video image is identified, and the voice data of the trainer in the online teaching is extracted.

[0078] During the online teaching of the trainer, some noise may occur, such as the sudden ringing of the trainer's mobile phone, therefore, the voice in the video image needs to be de-noised.

[0079] Specifically, when the playing voice is extracted, the audio signal of the video image is first extracted, such as the audio signal including pitch, duration, intensity, etc. Then, according to the preset frequency feature, the voice signal is extracted from the audio signal, and the voice signal is taken as the playing voice. Thus, through the preset frequency feature, the voice of the trainer can be effectively preserved and the noise can be removed.

[0080] It should be noted that the audio signal includes the frequency feature, and the preset frequency feature can be set according to actual needs for voice processing, such as the preset frequency feature being 16kHZ and 16bit sampling.

[0081] Then, according to the voice data, the speed value of the trainer in the online teaching is determined. The speed, tone, and language sense of the trainer in the playing voice can be automatically identified through the voice recognition technology based on artificial intelligence deep learning.

[0082] If the speed value exceeds the preset speed threshold, the training file bound to the online training course is called to determine the corresponding training content of the video image in the training file.

[0083] It should be noted that the preset speed threshold is an attribute for measuring the speed of the trainer, which can be set according to actual needs.

[0084] Since different course contents have different difficulty levels, if the difficulty level of the training content is lower than the preset difficulty level, it means that the trainee does not understand the trainer's teaching method more, and then the text information of the voice data is converted into a demonstration playing voice according to the preset speed threshold.

[0085] It should be noted that the difficulty level of the training content indicates the difficulty of the training content, and the preset difficulty level is an attribute for measuring the difficulty level of the training content, which can be set according to actual needs.

[0086] Confirm whether to replay the video image in the current period for each training student in the inactive student combination. It should be noted that when the video image in the current period is replayed, the playback sound of the trainer is replaced with the demonstration playback voice.

[0087] If yes, the fixed time period is updated according to the duration of the current period to increase the viewing duration of the corresponding training student.

[0088] It should be noted that if the speech speed value is lower than the preset speech speed threshold, the training file bound to the online training course is called to determine the corresponding training content of the video image in the training file, and the inactive student combination and the training content of the corresponding period are fed back to the housekeeping manager.

[0089] If the difficulty level of the training content is higher than the preset difficulty level, the inactive student combination and the training content of the corresponding period are also fed back to the housekeeping manager.

[0090] In some embodiments of the present application, the coaching workload of offline trainers can be replaced by introducing intelligent robots.

[0091] Specifically, first, determine the offline training course to be used for collective housekeeping training of multiple training students. The offline training course is a practical operation course of the online training course.

[0092] Then, obtain the training start time and the training duration of the offline training course, and determine the specified training time of the training content in the offline training course according to the training start time and the training duration. For example, according to the training time and the training duration, determine the time interval occupied by the multiple training contents during the training period, and then determine the specified training time of the training content in the offline training course according to the time interval, i.e., the start training time of the training content in the offline training course.

[0093] Then, the position parameter of the inactive training student during the practical operation in the offline training course is determined by shooting the image of the inactive training student. The wearable device can also be distributed to each training student in the offline training course, so that the position parameter of the inactive training student during the practical operation in the offline training course can be obtained through the wearable device carried by the inactive training student.

[0094] Finally, when the specified training time is reached, the training content and the location parameter are sent to the intelligent robot, so that the intelligent robot assists in training the practical operation action of each training trainee in the inactive trainee combination according to the training content.

[0095] Further, the intelligent robot generates a standard practical operation action of each training trainee in the inactive trainee combination according to the training content.

[0096] Then, a practical operation video image of each training trainee in the inactive trainee combination is obtained. Then, the practical operation video image is recognized to obtain a practical operation action of each training trainee in the inactive trainee combination.

[0097] Then, the practical operation action is compared with the standard practical operation action to determine a matching degree of the practical operation action and the standard practical operation action.

[0098] Finally, when the matching degree is lower than a preset threshold, the specified training trainee is reminded.

[0099] It should be noted that when the matching degree is higher than or equal to the preset threshold, the practical operation action of the training trainee is not processed.

[0100] Based on the same idea, some embodiments of the present application also provide a device and a non-volatile computer storage medium corresponding to the above method.

[0101] Figure 2 A structure schematic diagram of a data processing device for housekeeping training provided by an embodiment of the present application includes:

[0102] at least one processor; and

[0103] a memory in communication connection with the at least one processor; wherein

[0104] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to:

[0105] determine an online training course to be used for collective housekeeping training of multiple training trainees of the same professional category within a fixed time period, so as to push the online training course to a terminal device of each training trainee;

[0106] determine a professional skill score of each training trainee for the online training course according to the online training course and personal information pre-uploaded by each training trainee through a multi-element logistic regression professional skill prediction model;

[0107] grouping the each trainee by dividing the multiple professional skill scores into preset professional skill score intervals, to obtain multiple trainee combinations;

[0108] when playing the online training course, obtaining video images of the each trainee in a preset period, and generating a concentration score of the each trainee for the online training course by inputting the video images into a pre-trained neural network model;

[0109] determining a training expectation level of the each trainee combination according to the concentration score and the professional skill score; the training expectation level is used to represent a mastering degree of the each trainee combination on training course content in a current period;

[0110] in the multiple trainee combinations, a trainee combination lower than a preset level is taken as an inactive trainee combination;

[0111] if the online training course is a live course, generating a professional skill analysis diagram of the inactive trainee combination;

[0112] pushing the professional skill analysis diagram of the inactive trainee combination to a trainer terminal device, so that the trainer updates a preset teaching progress.

[0113] Some embodiments of the present application provide a data processing nonvolatile computer storage medium for housekeeping training, which stores computer executable instructions, and the computer executable instructions are configured to:

[0114] determining an online training course to be used for collective housekeeping training of multiple trainees of the same professional category in a fixed time period, and pushing the online training course to a terminal device of each trainee;

[0115] determining a professional skill score of each trainee for the online training course according to the online training course and personal information pre-uploaded by each trainee by using a multiple logistic regression professional skill prediction model;

[0116] grouping the each trainee by dividing the multiple professional skill scores into preset professional skill score intervals, to obtain multiple trainee combinations;

[0117] when playing the online training course, obtaining video images of the each trainee in a preset period, and generating a concentration score of the each trainee for the online training course by inputting the video images into a pre-trained neural network model;

[0118] determine a training expectation level of each of the training trainee combinations according to the concentration score and the professional skill score; the training expectation level is used to represent a mastering degree of the training trainee combinations on the training course content in a current period;

[0119] in the plurality of training trainee combinations, a training trainee combination lower than the preset level is taken as an inactive trainee combination;

[0120] if the online training course is a live course, generate a professional skill analysis diagram of the inactive trainee combination;

[0121] push the professional skill analysis diagram of the inactive trainee combination to a trainer terminal device, so that the trainer updates a preset teaching progress.

[0122] Each of the embodiments in the present application is described in a progressive manner, and the same and similar parts between the embodiments can be referred to each other. Each of the embodiments focuses on the difference from other embodiments. In particular, for the device and medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the related parts can be referred to the part of the method embodiments.

[0123] The device and medium provided by the embodiments of the present application are one-to-one corresponding to the method, and therefore, the device and medium also have the similar beneficial technical effects as the method. Since the beneficial technical effects of the method have been described in detail above, the beneficial technical effects of the device and medium will not be described here.

[0124] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0125] The present application is described with reference to flowcharts and / or block diagrams of the method, device (system), and computer program product according to the embodiments of the present application. The concentration can be achieved by computer program instructions in each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a machine that implements the flowcharts and / or block diagrams. Figure 1 one flow or multiple flows and / or blocksFigure 1 means for performing the function specified by the block or blocks.

[0126] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a Figure 1 one or more processes and / or blocks Figure 1 means for performing the function specified by the block or blocks.

[0127] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the Figure 1 one or more processes and / or blocks Figure 1 means for performing the function specified by the block or blocks.

[0128] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0129] The memory can include non-persistent memory and / or volatile memory, such as a random access memory (RAM) including a cache area for the temporary storage of data. The memory can also include non-volatile memory, such as a read only memory (ROM), EPROM, EEPROM, flash memory, or other non-volatile memory storage. The memory can be another type of computer-readable media, a magnetic-based memory, such as a magnetic disks, magnetic tapes or cassettes, or cards or other types of memory which can store data. The memory can include an interface for receiving data from or transmitting data to the bus(es).

[0130] Computer-readable media includes permanent and non-permanent, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory (ROM), electrically erasable programmable read only memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital versatile discs (DVDs) or other optical storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible to a computing device. According to the definition herein, computer-readable media does not include transitory media, such as modulated data signals and carrier waves.

[0131] It should also be noted that the terms "comprising", "comprises", "including", "includes" or any other variation thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises a... " does not, without more constraints, exclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0132] The above description is only the preferred embodiment of the present application, and is not intended to limit the present application. The present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the technical principles of the present application shall fall within the protection scope of the present application.

Claims

1. A data processing method for housekeeping training, characterized by, The method comprises: determining an online training course to be collectively provided to multiple trainees of the same professional category for home management training within a fixed time period, and pushing the online training course to a terminal device of each trainee; determining a professional skill score of each trainee for the online training course according to the online training course and personal information previously uploaded by each trainee, by using a multi-element logistic regression professional skill prediction model; grouping each trainee by dividing multiple professional skill scores into pre-set professional skill score intervals, to obtain multiple trainee combinations; acquiring a video image of each trainee within a pre-set period during playing of the online training course, and generating a concentration score of each trainee for the online training course by inputting the video image into a pre-trained neural network model; determining a training expectation level of each trainee combination according to the concentration score and the professional skill score; the training expectation level is used to represent a mastery degree of the trainee combination to the training course content in the current period; in the multiple trainee combinations, a trainee combination lower than a pre-set level is taken as an inactive trainee combination; if the online training course is a live course, generating a professional skill analysis diagram of the inactive trainee combination; pushing the professional skill analysis diagram of the inactive trainee combination to a trainer terminal device, so that the trainer updates a pre-set teaching progress; the determination of the training expectation level of each trainee combination according to the concentration score and the professional skill score specifically comprises: mapping the professional skill score by using a first mapping relationship, to determine a corresponding understanding score of each trainee for the online training course; dividing each trainee in each trainee combination according to a pre-divided concentration score interval, to obtain multiple cluster combinations; determining a corresponding attention parameter of each cluster combination, the attention parameter comprising age and understanding score; determining a training expectation coefficient of each cluster combination according to the attention parameter and each cluster combination; determining a training expectation weight of each cluster combination by using a second mapping relationship for different concentration score intervals; the higher the concentration score, the greater the training expectation weight; weighting and summing the training expectation coefficient of each cluster combination and the corresponding training expectation weight, to obtain a local training expectation coefficient of each trainee combination; mapping the local training expectation coefficient of each trainee combination by using a third mapping relationship, to determine a training expectation level of each trainee combination; the determination of the training expectation coefficient of each cluster combination according to the attention parameter specifically comprises: dividing each trainee in each cluster combination according to a pre-divided age interval, to obtain multiple sub-cluster combinations; determining a weight of each sub-cluster combination by using a fourth mapping relationship for different age intervals; the higher the age, the lower the weight of each sub-cluster combination. weighting and summing the understanding scores of all sub-cluster combinations in the same cluster combination with corresponding weights, to obtain a local training expectation coefficient of each cluster combination; mapping the local training expectation coefficient of each cluster combination through a fourth mapping relationship to determine a training expectation coefficient of each cluster combination; Before the determining the training expectation level of each training trainee combination according to the concentration score and the professional skill score, the method further comprises: determining a concentration score sequence of each training trainee for the online training course; in the concentration score sequence, determining a concentration level change of each training trainee when watching the online training course according to the difference between a plurality of concentration scores of each training trainee in adjacent periods; if the concentration level change is lower than a preset change level, retrieving a compensation degree corresponding to the concentration level change in a pre-constructed compensation mapping table according to the concentration score; the concentration level change and the compensation degree are negatively correlated in absolute value; compensating the concentration score according to the compensation degree to update the concentration score.

2. The method of claim 1, wherein, Before the generating the concentration score of each training trainee for the online training course by inputting the video image into a pre-trained neural network model, the method further comprises: obtaining a plurality of sample video images of sample training trainees of different professional categories; the professional categories include at least one of maternal and infant categories, old-age care categories, and cleaning categories; dividing the plurality of sample video images into a plurality of sample groups according to the professional categories; from each sample group, selecting a preset number of training samples to obtain a training sample set; training an initial convolutional neural network model with the training sample set as the input sample and the concentration score of the sample training trainee as the sample label until a training stop condition is reached to obtain the neural network model.

3. The method of claim 1, wherein, After the determining the inactive trainee combination from the plurality of training trainee combinations which are lower than the preset level, the method further comprises: if the online training course is a recording and broadcasting course, obtaining a video image of an online teaching of a trainer in the online training course in a current period, recognizing the video image, and extracting voice data of the trainer in the online teaching; determining a speech speed value of the trainer in the online teaching according to the voice data; if the speech speed value exceeds a preset speech speed threshold, calling a training file bound to the online training course to determine corresponding training content of the video image in the training file; if the difficulty level of the training content is lower than a preset difficulty level, converting text information of the voice data into demonstration playing voice according to the preset speech speed threshold; confirming with each training trainee in the inactive trainee combination whether to replay the video image in the current period; if yes, updating the fixed time period according to the time length of the current period to increase the viewing time length of the corresponding training trainee.

4. The method of claim 3, wherein, The method further comprises: Determine an offline training course to be performed for collective housekeeping training of a plurality of trainees, wherein the offline training course is a practical operation course of the online training course; Obtain a training start time and a training duration of the offline training course; Determine a designated training time of the training content in the offline training course according to the training start time and the training duration; Determine a position parameter of each trainee in the offline training course when performing practical operation in the inactive trainee combination; When the designated training time is reached, send the training content and the position parameter to the intelligent robot, so that the intelligent robot assists in training the practical operation action of each trainee in the inactive trainee combination according to the training content.

5. The method of claim 4, wherein, The intelligent robot assists in training the practical operation action of each trainee in the inactive trainee combination according to the training content, specifically comprising: The intelligent robot generates a standard practical operation action of each trainee in the inactive trainee combination according to the training content; Obtain a practical operation video image of each trainee in the inactive trainee combination; Identify the practical operation video image to obtain the practical operation action of each trainee in the inactive trainee combination; Compare the practical operation action with the standard practical operation action to determine the matching degree of the practical operation action and the standard practical operation action; When the matching degree is lower than a preset threshold, remind the designated trainee.

6. A data processing device for housekeeping training, characterized by, Comprise: At least one processor; And The memory is in communication connection with the at least one processor; wherein The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the data processing method for housekeeping training according to any one of claims 1-5.

7. A data processing non-transitory computer storage medium for housekeeping training, storing computer executable instructions, characterized by, The computer executable instructions are arranged to execute the data processing method for housekeeping training according to any one of claims 1-5.

Images