Learning rate-based cognitive interaction scheme pushing method and system

Abstract

The application discloses a learning rate-based cognitive interaction scheme pushing method and system. The method comprises the following steps: determining a preset task library based on a preset database, and obtaining cognitive interaction data for each cognitive interaction task in the preset task library; for each cognitive interaction task, obtaining an interaction normal mode based on the cognitive interaction data of the cognitive interaction task; for a user to be pushed, obtaining a learning rate of a current interaction frequency for each cognitive interaction task based on the interaction normal mode of the cognitive interaction task; obtaining an extraction weight and a difficulty weight of each cognitive interaction task according to the learning rate of the current interaction frequency of the user to each cognitive interaction task; and extracting a preset number of cognitive interaction tasks from the preset task library based on the extraction weight and the difficulty weight of each cognitive interaction task to combine cognitive interaction schemes and push the cognitive interaction schemes to the user for cognitive interaction.

Description

Technical Field

[0001] This invention relates to a cognitive interaction scheme recommendation method based on learning rate, and also to a corresponding cognitive interaction scheme recommendation system, belonging to the field of cognitive training technology. Background Technology

[0002] Cognitive training is an effective means of delaying cognitive decline. To achieve good training results, individuals need to train continuously, potentially for up to three months or even a lifetime. However, how to stimulate individuals' interest in cognitive training and enable them to sustain their effort and action is a key area that requires continuous innovation and breakthroughs in cognitive training design.

[0003] Traditional cognitive training typically employs classic task settings. For example, Chinese invention patent ZL201510891880.0 discloses a three-dimensional cognitive training system and its training method. This method expands the diversity of three-dimensional cognitive training by detecting the accuracy of matching results and controlling the difficulty level of shape molds accordingly.

[0004] However, traditional cognitive training primarily focuses on the dynamic learning process of a particular cognitive ability. While accuracy, achieved through a correct number of repetitions, can reflect the outcome of each training session, it is insufficient to comprehensively depict the user's actual learning curve during training. This means that although accuracy is an important metric, it may not fully reflect an individual's dynamic progress and learning depth in cognitive training. Therefore, to more comprehensively assess and improve an individual's cognitive abilities, the design of cognitive training requires further innovation and improvement. Summary of the Invention

[0005] The primary technical problem to be solved by this invention is to provide a cognitive interaction scheme push method based on learning rate.

[0006] Another technical problem to be solved by the present invention is to provide a cognitive interaction scheme push system based on learning rate.

[0007] To achieve the above-mentioned technical objectives, the present invention adopts the following technical solution:

[0008] According to a first aspect of the present invention, a method for pushing cognitive interaction schemes based on learning rates is provided, comprising the following steps:

[0009] A preset task library is determined based on a preset database, and for each cognitive interaction task in the preset task library, the cognitive interaction data of the user group for that cognitive interaction task is obtained from the preset database.

[0010] For each cognitive interaction task, an interaction norm for that cognitive interaction task is obtained based on the cognitive interaction data of the user group for that cognitive interaction task.

[0011] For users to be pushed to, based on the interaction norms of each cognitive interaction task, the learning rate of each cognitive interaction task at the current number of interactions is obtained;

[0012] For the user to be pushed to, based on the learning rate of the user for each cognitive interaction task in the current number of interactions, the extraction weight and difficulty weight of each cognitive interaction task are obtained respectively; wherein, the extraction weight is inversely proportional to the learning rate of the cognitive interaction task, and the difficulty weight is directly proportional to the learning rate of the cognitive interaction task.

[0013] Based on the extraction weight and difficulty weight of each cognitive interaction task, a preset number of cognitive interaction tasks are extracted from the preset task library to form a cognitive interaction scheme and push it to the user for cognitive interaction.

[0014] Preferably, the interaction norms of the cognitive interaction task are obtained through the following steps:

[0015] For the cognitive interaction task, cognitive interaction data of the user group on the cognitive interaction task are collected from the preset database;

[0016] Remove data points from the cognitive interaction data whose interaction interval is lower than the first threshold or higher than the second threshold;

[0017] Calculate the mean and standard deviation of the time interval between each cognitive interaction and the previous one, and establish the curve of the score of each cognitive interaction task as a function of the number of cognitive interactions, so as to jointly constitute the interaction norm of the cognitive interaction task.

[0018] Preferably, the learning rate of the cognitive interaction task in the current number of interactions is calculated through the following steps:

[0019] Based on a preset learning model, the implicit potential of the user for each cognitive interaction task k is obtained; wherein, the preset learning model is expressed as:

[0020]

[0021] in, This represents the actual score of cognitive interaction task k in the (t+1)th interaction. σ represents the implicit potential of cognitive interaction task k at the (t+1)th interaction. t+1 This represents the standard deviation of the interaction growth score for the cognitive interaction task k within the interaction norm;

[0022] Based on the implicit potential of the cognitive interaction task k at the (t+1)th interaction, the learning rate α of the user based on the interaction norm is fitted.

[0023]

[0024] in, δ represents the actual score of task k in the t-th interaction. t+1 This represents the average increase in the task score within the interaction norm.

[0025] Preferably, for each cognitive interaction task, the reciprocal of the learning rate α of the cognitive interaction task is used as the extraction weight, and the learning rate α of the cognitive interaction task is used as the difficulty weight.

[0026] Preferably, the cognitive interaction scheme push method further includes:

[0027] Set a learning limit and compare the learning rate α of each cognitive interaction task of the user with the size of the learning limit;

[0028] If the learning rate α of the current cognitive interaction task is greater than the learning limit, then the difficulty weight of the current cognitive interaction task is increased according to the magnitude of the learning rate α.

[0029] If the learning rate α of the current cognitive interaction task is less than the learning limit, the difficulty of the current cognitive interaction task remains unchanged, and the extraction weight of the current cognitive interaction task is increased according to the magnitude of the learning rate α.

[0030] Preferably, for the user to be pushed to, the current interaction stage of the cognitive interaction task is determined based on the number of interactions; wherein different numbers of interactions correspond to different interaction stages.

[0031] Based on the interaction norms of the cognitive interaction task, the learning rate α of the user to be pushed to is obtained at different interaction stages.

[0032] Based on the learning rate α of the user to be pushed to at different interaction stages, plot the learning change curve of the user to be pushed to for this cognitive interaction task.

[0033] Preferably, the cognitive interaction scheme push method further includes:

[0034] Obtain the cognitive interaction results of the user to be pushed to based on the current cognitive interaction scheme; wherein, the cognitive interaction results include the cognitive interaction data of each cognitive interaction task included in the current cognitive interaction scheme;

[0035] Based on the cognitive interaction results, the learning rate α of each cognitive interaction task corresponding to the current cognitive interaction scheme is updated respectively.

[0036] Preferably, the cognitive interaction scheme push method further includes:

[0037] Every preset number of days, the preset database and preset task library are updated based on the cognitive interaction data of each user to be pushed to.

[0038] Preferably, the preset number of days is one week, one month, or three months.

[0039] According to a second aspect of the present invention, a cognitive interaction scheme push system based on learning rate is provided, including a processor and a memory, wherein the processor reads a computer program in the memory for executing the above-described cognitive interaction scheme push method.

[0040] Compared with the prior art, the present invention has the following technical effects:

[0041] 1. Through large-scale sample testing, the scores and changes of each interaction for each task were obtained. Furthermore, through a series of analyses such as data cleaning, descriptive statistics, and modeling, the change curves and norms of the scores of each interaction task with the number of times were established.

[0042] 2. By combining traditional learning models, we have achieved modeling of the learning rate for each user. After considering the differences among user groups and the number of interactions, we can accurately characterize the learning rate of each user in different tasks and at different interaction stages.

[0043] 3. It allows for horizontal comparison of learning progress across cognitive interaction tasks, and performs normalized comparisons based on user learning within interaction norms. This enables the first-ever assessment of task learning progress using the learning rate, and allows for adjustments to task delivery. Attached Figure Description

[0044] Figure 1 A flowchart illustrating an overall process for a cognitive interaction scheme push method based on learning rate, provided in an embodiment of the present invention.

[0045] Figure 2 A detailed flowchart of a cognitive interaction scheme push method based on learning rate provided in an embodiment of the present invention;

[0046] Figure 3 This is a schematic diagram of a cognitive interaction scheme push system based on learning rate, provided in an embodiment of the present invention. Detailed Implementation

[0047] The technical content of the present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0048] The technical concept of this invention is to design each cognitive interaction task according to an initial fixed difficulty level, build a learning model based on the learning rate, calculate the user's comparison results based on norms under different conditions, and model to obtain the learning rate of each cognitive interaction task. An extraction weight is assigned to each cognitive interaction task according to the reciprocal of the learning rate; the lower the learning rate, the higher the extraction weight. Furthermore, a difficulty weight is directly assigned to each cognitive interaction task according to the learning rate; the higher the learning rate, the higher the difficulty weight. Therefore, compared to accuracy, the learning rate-based learning model depicts the process by which autonomous and proactive users adjust their actions in a timely manner based on feedback when interacting with the external environment to achieve a certain goal. It is a more accurate model reflecting the learning level relative to the user group. Moreover, the learning rate based on population norms can more accurately reflect the user's actual learning situation during the interaction process.

[0049] Specifically, such as Figure 1 and Figure 2 As shown, the present invention provides a cognitive interaction scheme push method based on learning rate, which specifically includes steps S1 to S5:

[0050] S1: Obtain cognitive interaction data of the user group.

[0051] Specifically, based on a preset database (historical big data in this embodiment), all types of cognitive interaction tasks included in the historical big data are identified, thereby determining a preset task library. Then, for each cognitive interaction task in the preset task library, cognitive interaction data of a user group (consisting of tens of thousands of users) on that cognitive interaction task are obtained from the preset database.

[0052] S2: For each cognitive interaction task, obtain the interaction norms of the cognitive interaction task based on the cognitive interaction data of the user group.

[0053] Specifically, this includes steps S21 to S24:

[0054] S21: For a specific cognitive interaction task, collect cognitive interaction data of the user group on the cognitive interaction task from a preset database.

[0055] S22: Remove data points from cognitive interaction data whose interaction interval is lower than the first threshold (i.e., the interaction interval is too short) or higher than the second threshold (i.e., the interaction interval is too long).

[0056] S23: Calculate the mean and standard deviation of the time interval between each cognitive interaction and the previous one, and establish the curve of the score of each cognitive interaction task as a function of the number of cognitive interactions, so as to jointly constitute the interaction norm of the cognitive interaction task.

[0057] Where, δ t+1 σ represents the average increase in the score of this task among the user group in the (t+1)th interaction. t+1 This represents the standard deviation of the score increase in the user group at the t+1th interaction, i.e., the individual differences in how the interaction affects the improvement of cognitive level.

[0058] S24: Repeat the above process S21 to S23 until the interaction norms for each cognitive interaction task are obtained.

[0059] S3: For users to be pushed to, based on the interaction norms of each cognitive interaction task, obtain the learning rate of each cognitive interaction task at the current number of interactions.

[0060] Specifically, this includes steps S31 to S33:

[0061] S31: Based on a pre-defined learning model, obtain the user's implicit potential for each cognitive interaction task k; wherein, the pre-defined learning model is represented as:

[0062]

[0063] in, This represents the actual score of cognitive interaction task k in the (t+1)th interaction. σ represents the implicit potential of cognitive interaction task k at the (t+1)th interaction. t+1 This represents the standard deviation of the interaction growth score of cognitive interaction task k in the interaction norm.

[0064] S32: Based on the implicit potential of cognitive interaction task k at the (t+1)th interaction, fit the user's learning rate α based on the interaction norm.

[0065]

[0066] in, δ represents the actual score of task k in the t-th interaction. t+1 This represents the average increase in the task score within the interaction norm.

[0067] S33: Repeat the above process S31 to S32 until the learning rate α of each cognitive interaction task in the current number of interactions is obtained.

[0068] S4: For users to be pushed to, based on the user's learning rate for each cognitive interaction task in the current number of interactions, obtain the extraction weight and difficulty weight for each cognitive interaction task.

[0069] In this embodiment, the extraction weight is inversely proportional to the learning rate of the cognitive interaction task, while the difficulty weight is directly proportional to the learning rate of the cognitive interaction task. Specifically, for each cognitive interaction task, the reciprocal of the learning rate α of the cognitive interaction task is used as the extraction weight, and the learning rate α of the cognitive interaction task is used as the difficulty weight.

[0070] Furthermore, preferably, a learning threshold can be set (e.g., a learning threshold of 1), and the learning rate α of each user's cognitive interaction task can be compared with the learning threshold. If the learning rate α of the current cognitive interaction task is greater than the learning threshold, the difficulty weight of the current cognitive interaction task is increased according to the magnitude of the learning rate α. If the learning rate α of the current cognitive interaction task is less than the learning threshold, the difficulty of the current cognitive interaction task remains unchanged, and the extraction weight of the current cognitive interaction task is increased according to the magnitude of the learning rate α.

[0071] S5: Extract a preset number of cognitive interaction tasks from the preset task library, combine them to form a cognitive interaction scheme, and push it to the user for cognitive interaction.

[0072] Understandably, since the probability of extracting each cognitive interaction task is weighted according to the reciprocal of the learning rate, conditions with lower learning rates have higher extraction weights, thereby increasing users' further practice on tasks where they are currently not performing well. Furthermore, the difficulty coefficient of each cognitive interaction task is directly weighted according to the learning rate value; therefore, conditions with higher learning rates have higher difficulty coefficients, thereby increasing the challenge of practice for tasks where patients are currently performing well.

[0073] S6: Cognitive interaction solution updated.

[0074] In this embodiment, the cognitive interaction results of the user based on the current cognitive interaction scheme are obtained; wherein, the cognitive interaction results include the cognitive interaction data of each cognitive interaction task included in the current cognitive interaction scheme. Then, based on the cognitive interaction results, the learning rate α of each cognitive interaction task corresponding to the current cognitive interaction scheme is updated respectively.

[0075] Understandably, as the number of cognitive interactions increases, the learning rate α of the corresponding cognitive interaction task will also increase. By updating the learning rate α of the cognitive interaction task, the extraction weight and difficulty weight of the cognitive interaction task can be adjusted, thereby updating the cognitive interaction scheme for the next time.

[0076] Furthermore, as users engage in continuous cognitive interaction, the number of interactions for each cognitive interaction task increases, with different interaction counts corresponding to different interaction stages. Therefore, based on the number of interactions a user has with a particular cognitive interaction task, the current interaction stage of that task can be determined. Then, based on the interaction norms of the cognitive interaction task, the user's learning rate α is obtained at each interaction stage. Finally, based on the user's learning rate α at different interaction stages, a learning change curve for the cognitive interaction task is plotted, thus providing a visual representation of the user's interaction effect on that cognitive interaction task and helping the user build learning confidence.

[0077] S7: Preset database update.

[0078] Understandably, every preset number of days (e.g., a week, a month, or three months), a new batch of cognitive interaction data is generated based on each user's cognitive interaction behavior. This new data is then used to update the original database and reconstruct the interaction norms for each cognitive interaction task. This allows the interaction norms for each cognitive interaction task to adapt to the user's stage of development, improving the cognitive enhancement effect of the cognitive interaction solution on the user's cognitive development.

[0079] Based on the aforementioned cognitive interaction scheme recommendation method based on learning rate, this invention further provides a cognitive interaction scheme recommendation system based on learning rate. For example... Figure 3 As shown, the cognitive interaction scheme push system includes one or more processors 21 and a memory 22. The memory 22 is coupled to the processors 21 and is used to store one or more programs. When the one or more programs are executed by the one or more processors 21, the one or more processors 21 implement the learning rate-based cognitive interaction scheme push method as described in the above embodiment.

[0080] The processor 21 controls the overall operation of the cognitive interaction scheme delivery system to complete all or part of the steps of the learning rate-based cognitive interaction scheme delivery method. The processor 21 can be a central processing unit (CPU), graphics processing unit (GPU), field-programmable gate array (FPGA), application-specific integrated circuit (ASIC), digital signal processing (DSP), etc. The memory 22 stores various types of data to support the operation of the cognitive interaction scheme delivery system. This data may include, for example, instructions for any application or method operating on the cognitive interaction scheme delivery system, as well as application-related data. The memory 22 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, etc.

[0081] In one exemplary embodiment, the cognitive interaction scheme push system can be implemented by a computer chip or physical entity, or by a product with certain functions, to execute the above-described learning rate-based cognitive interaction scheme push method and achieve the same technical effect as the method described above. A typical embodiment is a computer. Specifically, the computer can be, for example, a personal computer, a laptop computer, an in-vehicle human-machine interaction device, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0082] In another exemplary embodiment, the present invention also provides a computer-readable storage medium including program instructions, which, when executed by a processor, implement the steps of the learning rate-based cognitive interaction scheme push method in any of the above embodiments. For example, the computer-readable storage medium may be the aforementioned memory including program instructions, which may be executed by the processor of the cognitive interaction scheme push system to complete the aforementioned learning rate-based cognitive interaction scheme push method and achieve the same technical effects as the method described above.

[0083] In summary, the cognitive interaction scheme push method and system based on learning rate provided by the embodiments of the present invention have the following beneficial effects:

[0084] 1. Through large-scale sample testing, the scores and changes of each interaction for each task were obtained. Furthermore, through a series of analyses such as data cleaning, descriptive statistics, and modeling, the change curves and norms of the scores of each interaction task with the number of times were established.

[0085] 2. By combining traditional learning models, we have achieved modeling of the learning rate for each user. After considering the differences among user groups and the number of interactions, we can accurately characterize the learning rate of each user in different tasks and at different interaction stages.

[0086] 3. A horizontal comparison of learning progress across cognitive interaction tasks is performed, and a normalized comparison is conducted based on the user's learning progress within the interaction norms. This allows for the first-ever assessment of task learning progress using the learning rate, and subsequent adjustments to task delivery.

[0087] It should be noted that the above embodiments are merely illustrative examples, and the technical solutions of each embodiment can be combined, all of which are within the protection scope of this invention.

[0088] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0089] The above provides a detailed description of the learning rate-based cognitive interaction scheme push method and system provided by this invention. Any obvious modifications made by those skilled in the art without departing from the essence of this invention will constitute an infringement of the patent rights of this invention and will incur corresponding legal liability.

Claims

1. A cognitive interaction scheme push method based on learning rate, characterized in that... The steps include the following: A preset task library is determined based on a preset database, and for each cognitive interaction task in the preset task library, the cognitive interaction data of the user group for that cognitive interaction task is obtained from the preset database. For each cognitive interaction task, an interaction norm for that cognitive interaction task is obtained based on the cognitive interaction data of the user group for that cognitive interaction task. The interaction norms of the cognitive interaction task are obtained through the following steps: For the cognitive interaction task, collect cognitive interaction data of the user group on the cognitive interaction task from the preset database; remove data points from the cognitive interaction data whose interaction interval is lower than a first threshold or higher than a second threshold; calculate the mean and standard deviation of the time interval between each cognitive interaction and the previous cognitive interaction, and establish the curve of the score of each cognitive interaction task as a function of the number of cognitive interactions, so as to jointly constitute the interaction norms of the cognitive interaction task. For the user to be pushed to, based on the interaction norm of each cognitive interaction task, the learning rate of each cognitive interaction task at the current number of interactions is obtained; wherein, the learning rate of the cognitive interaction task at the current number of interactions is calculated through the following steps: Based on a preset learning model, the implicit potential of the user for each cognitive interaction task k is obtained; wherein, the preset learning model is expressed as: in, This represents the actual score of cognitive interaction task k in the (t+1)th interaction. σ represents the implicit potential of cognitive interaction task k at the (t+1)th interaction. t+1 This represents the standard deviation of the interaction growth score for the cognitive interaction task k within the interaction norm; Based on the implicit potential of the cognitive interaction task k at the (t+1)th interaction, the learning rate α of the user based on the interaction norm is fitted. in, δ represents the actual score of task k in the t-th interaction. t+1 This represents the average increase in the task's score within the interaction norm; For the user to be pushed to, based on the learning rate of the user for each cognitive interaction task in the current number of interactions, the extraction weight and difficulty weight of each cognitive interaction task are obtained respectively; wherein, the extraction weight is inversely proportional to the learning rate of the cognitive interaction task, and the difficulty weight is directly proportional to the learning rate of the cognitive interaction task. Set a learning limit and compare the learning rate α of each cognitive interaction task of the user with the size of the learning limit; If the learning rate α of the current cognitive interaction task is greater than the learning limit, then the difficulty weight of the current cognitive interaction task is increased according to the magnitude of the learning rate α. If the learning rate α of the current cognitive interaction task is less than the learning limit, then the difficulty of the current cognitive interaction task remains unchanged, and the extraction weight of the current cognitive interaction task is increased according to the magnitude of the learning rate α. Based on the extraction weight and difficulty weight of each cognitive interaction task, a preset number of cognitive interaction tasks are extracted from the preset task library to form a cognitive interaction scheme and push it to the user for cognitive interaction.

2. The cognitive interaction scheme push method as described in claim 1, characterized in that: For each cognitive interaction task, the reciprocal of the learning rate α of the cognitive interaction task is used as the extraction weight, and the learning rate α of the cognitive interaction task is used as the difficulty weight.

3. The cognitive interaction scheme push method as described in claim 1, characterized in that: For the user to be pushed to, the current interaction stage of the cognitive interaction task is determined based on the number of interactions; wherein, different numbers of interactions correspond to different interaction stages. Based on the interaction norms of the cognitive interaction task, the learning rate α of the user to be pushed to is obtained at different interaction stages. Based on the learning rate α of the user to be pushed to at different interaction stages, plot the learning change curve of the user to be pushed to for this cognitive interaction task.

4. The cognitive interaction scheme push method as described in claim 1, characterized in that... Also includes: Obtain the cognitive interaction results of the user to be pushed to based on the current cognitive interaction scheme; wherein, the cognitive interaction results include the cognitive interaction data of each cognitive interaction task included in the current cognitive interaction scheme; Based on the cognitive interaction results, the learning rate α of each cognitive interaction task corresponding to the current cognitive interaction scheme is updated respectively.

5. The cognitive interaction scheme push method as described in claim 1, characterized in that... Also includes: Every preset number of days, the preset database and preset task library are updated based on the cognitive interaction data of each user to be pushed to.

6. The cognitive interaction scheme push method as described in claim 5, characterized in that: The preset number of days is one week, one month, or three months.

7. A cognitive interaction scheme push system based on learning rate, characterized in that... It includes a processor and a memory, wherein the processor reads a computer program from the memory for executing the cognitive interaction scheme push method as described in any one of claims 1 to 6.

Images