Barrier-free residence permit endorsement service system and method based on artificial intelligence

Abstract

The invention belongs to the technical field of endorsement service, and discloses a barrier-free residence permit endorsement service method based on artificial intelligence, and the method comprises the steps: recognizing the certificate information of a user; sending the certificate information to a query chain; nodes of the query chain are composed of an endorsement server, a social security query server, a medical query server, an educational background query server and a mode setting server; receiving a setting serial number sent by the mode setting server, and setting an adjustment interface and an interaction mode corresponding to the endorsement terminal according to the setting serial number; user intervention setting is not needed, barrier-free residence permit endorsement service is provided for special group users, and the endorsement success rate of each type of users is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of endorsement services, and more specifically, to an artificial intelligence-based barrier-free residence permit endorsement service system and method. Background Art

[0002] With the continuous improvement of the level of social informatization and intelligence, various government services are gradually developing in the direction of online, automated and barrier-free. Taking the endorsement of residence permits as an example, the traditional handling method usually requires users to conduct manual review and endorsement at offline windows, which is not only time-consuming and laborious, but also often cannot provide sufficient assistance and guidance when facing users with special needs (such as illiterate, color blind, visually impaired, hearing impaired, anxious, depressed, etc.), resulting in such users encountering greater difficulties in the process of handling, and even unable to successfully complete the endorsement.

[0003] Although in the prior art, in order to improve the success rate of visa application, interface and interactive mode option settings are set, users still need to set them themselves. Especially for special groups, they may not know how to set and cannot complete these settings, which may cause cognitive pressure, irritability or directly give up the visa application service, thereby reducing the success rate of visa application.

[0004] In view of this, the present invention proposes an artificial intelligence-based barrier-free residence permit endorsement service system and method to solve the above problems. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art and achieve the above-mentioned purpose, the present invention provides the following technical solution: an artificial intelligence-based barrier-free residence permit endorsement service method, comprising:

[0006] Identify the user's certificate information;

[0007] Send the certificate information to the query chain; the nodes of the query chain are composed of an endorsement server, a social security query server, a medical query server, an education query server and a mode setting server; the certificate information is sent by the endorsement server;

[0008] Receive the setting sequence number sent by the mode setting server, and set the corresponding adjustment interface and interaction mode of the endorsement terminal according to the setting sequence number;

[0009] The mode setting server receives the social security vector fed back by the social security query server based on the certificate information, the health vector fed back by the medical query server based on the certificate information, and the academic qualification vector fed back by the social security query server based on the certificate information, and inputs them into the pre-trained endorsement service setting model in a preset manner, and outputs the setting serial number.

[0010] Furthermore, the setting sequence number generation method includes:

[0011] Step 1: The mode setting server receives the continuous social security payment duration in the social security vector fed back by the social security query server according to the certificate information, and determines whether the certificate information is eligible for residence permit endorsement;

[0012] Step 2. If the person is eligible for residence permit endorsement, the mode setting server receives the health vector fed back by the medical query server based on the certificate information, the education vector fed back by the social security query server based on the certificate information, and the age vector in the social security vector fed back by the social security query server based on the certificate information, and inputs them into the pre-trained endorsement service setting model in a preset manner, and outputs the setting serial number.

[0013] Furthermore, the method for determining whether the certificate information is eligible for residence permit endorsement includes:

[0014] If the continuous payment period of social security is greater than or equal to the preset value, then you are deemed eligible for residence permit endorsement, otherwise you are not eligible.

[0015] Furthermore, the health vector includes the relevant disease vector representation and position serial number corresponding to the certificate information; the position serial number and the relevant disease vector representation are fixed positions in the health vector; the relevant diseases include visual impairment, hearing impairment, cognitive impairment or psychological impairment, and the psychological impairment includes anxiety, depression or obsessive-compulsive disorder;

[0016] The educational background vector includes the educational background corresponding to the certificate information; the social security vector includes the age corresponding to the certificate information and the length of time for continuous payment of social security at the location of the endorsement terminal.

[0017] Furthermore, the training method of the endorsement service setting model includes:

[0018] Collecting group B of endorsement service setting data in advance, where B is an integer greater than 1, and the endorsement service setting data includes feature data and setting sequence numbers corresponding to the feature data; the feature data includes a health vector, an education vector, and an age vector arranged in a preset manner;

[0019] Each group of endorsement service setting data is used as the input of the endorsement service setting model, the endorsement service setting model takes the setting serial number corresponding to each group of endorsement service setting data as the output, takes the actual setting serial number corresponding to each group of endorsement service setting data as the prediction target, and takes minimizing the sum of the loss function values ​​of all the endorsement service setting data as the training target; the endorsement service setting model is trained until the sum of the prediction errors reaches convergence and the training is stopped; the endorsement service setting model is specifically a deep neural network model.

[0020] Furthermore, the loss function value is:

[0021]

[0022] In the formula, is the loss function value; B is the number of data sets set for the endorsement service; Set the prediction function of the model for the endorsement service, based on the parameters Set data x for the bth endorsement service (b) Make prediction output; y (b) Set the actual output corresponding to the data for the b-th endorsement service, x (b) Set data for the bth endorsement service; Set the parameters of the model for the endorsement service, namely the weights of the health vector, education vector, and age vector.

[0023] Furthermore, the parameters of the endorsement service setting model are optimized by optimizing algorithm training. The methods include:

[0024] Step 1: Randomly generate H groups of initial solutions, namely nests, each nest represents a set of parameters for the annotation service setting model

[0025]

[0026] In the formula, Set the model output for the endorsement service; J is the health vector; X is the education vector; N is the age vector; is the health vector weight; is the education vector weight; is the age vector weight;

[0027] Step 2: Iterate and update to generate a new nest;

[0028] Step 3: Use Calculate the loss function value of the new nest;

[0029] Step 4: If the loss function value of the new nest is less than the loss function value of any nest in group H, then the nest corresponding to the loss function value greater than the new nest in group H is replaced with the new nest with a preset probability;

[0030] When the preset number of iterations is reached, the nest with the smallest loss function value is selected as the parameter of the endorsement service setting model

[0031] Further, the method of generating a new nest comprises:

[0032]

[0033] In the formula, is the parameter of the tth iteration; is the parameter of the t+1th iteration; α is the step size factor; Levy(λ) is the random step size that obeys the Levy distribution. Levy flight helps to jump out of the local optimal solution; λ is the scale parameter.

[0034] The barrier-free residence permit endorsement service system based on artificial intelligence is applied in the endorsement server to implement the barrier-free residence permit endorsement service method based on artificial intelligence, including:

[0035] A collection module is used to identify the user's certificate information;

[0036] The first processing module sends the certificate information to the query chain; the nodes of the query chain are composed of an endorsement server, a social security query server, a medical query server, an education query server and a mode setting server; the certificate information is sent by the endorsement server;

[0037] The second processing module is used to receive the setting serial number sent by the mode setting server, and set the corresponding adjustment interface and interaction mode of the endorsement terminal according to the setting serial number;

[0038] The mode setting server receives the social security vector fed back by the social security query server based on the certificate information, the health vector fed back by the medical query server based on the certificate information, and the academic qualification vector fed back by the social security query server based on the certificate information, and inputs them into the pre-trained endorsement service setting model in a preset manner, and outputs the setting serial number.

[0039] The technical effects and advantages of the artificial intelligence-based barrier-free residence permit endorsement service system and method of the present invention are as follows:

[0040] The certificate information is sent out through the endorsement server, and various servers send the health vector, education vector, and social security vector corresponding to the certificate information to the mode setting server. The mode setting server inputs these vectors into the endorsement service setting model according to the preset method, outputs the setting serial number, and sends the setting serial number to the endorsement server. The endorsement server sets the corresponding adjustment interface and interaction method of the endorsement terminal. The setting serial number represents the preset interface and interaction method corresponding to each type of user (including normal users and special users) in the endorsement terminal. The whole process adaptively sets the interface and interaction method of the endorsement terminal, taking into account that special groups of users are not familiar with specific setting operations, and adaptively completes the setting of relevant interfaces and interaction methods without user intervention. This provides barrier-free residence permit endorsement services for special groups of users, and effectively improves the endorsement success rate of each type of user.

[0041] At the same time, each server only sends the corresponding feature vector based on the user's identity information, and does not directly transmit the user's real information; the mode setting server uses its own preset vector arrangement method to train the endorsement service setting model. Even if the usage record data of the later model is accidentally leaked, the specific user privacy information cannot be restored from it. Since the mode setting server does not know the actual meaning of the feature vector when receiving the data, the user's privacy is always protected during the entire data transmission and processing process and will not be leaked due to the transmission of feature data. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic diagram of the structure of the query chain of the present invention;

[0043] Figure 2 This is a schematic diagram of the modules of the endorsement server of the present invention;

[0044] Figure 3 This is a flow chart of the method for providing barrier-free residence permit endorsement services based on artificial intelligence of the present invention. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] Example 1

[0047] See also Figure 2 As shown, the barrier-free residence permit endorsement service system based on artificial intelligence in this embodiment is applied to the endorsement server, and the endorsement server includes:

[0048] The acquisition module is used to identify the user's certificate information. The user's certificate is such as an ID card, and the acquisition module is such as an ID card identifier.

[0049] The first processing module sends the certificate information to the query chain; Figure 1 As shown, the nodes of the query chain are composed of an endorsement server, a social security query server, a medical query server, an academic degree query server and a mode setting server. The certificate information is issued by the endorsement server. The certificate information is encrypted information and is decrypted by corresponding decryption programs set in the social security query server, the medical query server and the academic degree query server.

[0050] The second processing module is used to receive the setting serial number sent by the mode setting server, and set the corresponding adjustment interface and interaction mode of the endorsement terminal according to the setting serial number;

[0051] The method for generating a setting sequence number comprises:

[0052] Step 1: The mode setting server receives the continuous social security payment duration in the social security vector fed back by the social security query server according to the certificate information, and determines whether the certificate information is eligible for residence permit endorsement;

[0053] Step 2. If the person is eligible for residence permit endorsement, the mode setting server receives the health vector fed back by the medical query server based on the certificate information, the education vector fed back by the social security query server based on the certificate information, and the age vector in the social security vector fed back by the social security query server based on the certificate information, and inputs them into the pre-trained endorsement service setting model in a preset manner, and outputs the setting serial number.

[0054] The health vector includes the relevant disease vector representation and position number corresponding to the certificate information; the position number and the relevant disease vector representation are fixed in the health vector; it is convenient for the subsequent mode setting server to set the endorsement service setting instructions corresponding to the user; the relevant disease vector indicates whether the corresponding user suffers from the corresponding relevant disease, represented by "1" or "0", "1" means suffering from it, and "0" means not.

[0055] Related diseases include visual impairment, hearing impairment, cognitive impairment and psychological disorder, etc. For example, when visually impaired users use the signing terminal, they need the interactive methods provided by the signing terminal, such as voice prompts, tactile feedback, and screen reading function assistance, in order to complete the entire signing process; hearing impairments require the interactive methods provided by the signing terminal, such as sign language support and vibration prompts; cognitive impairments include mental retardation, autism, learning disabilities, etc., and the signing terminal interface needs to provide an inclusive design of graphical interface and error prompts.

[0056] Psychological disorders include anxiety, depression or obsessive-compulsive disorder, which require the terminal interface to display a progress bar, provide an "archive" function (allowing users to exit and continue the operation to avoid the pressure of completing the operation all at once), use soft blue or green interface colors, white noise or light music as background music (to help relax the mood), and add friendly prompts to the interaction method (for example, "Don't worry, you can modify the information at any time" or "If you have any questions, please contact us"); so as to reduce the cognitive burden of users of various categories, shorten the service process and reduce the difficulty of decision-making, provide barrier-free residence permit endorsement services for special groups of users, and improve the endorsement success rate of each type of user.

[0057] The education vector includes the education level corresponding to the certificate information, for example, a doctorate is represented by "1", a master's degree is represented by "2", and no education level is represented by "0". Education level reflects the user's ability to learn about unknown objects, which is related to cognitive impairment to a certain extent.

[0058] The social security vector includes the age corresponding to the certificate information and the recent continuous payment period of social security in the location of the endorsement terminal; the continuous payment period of social security is one of the important measurement conditions for whether the certificate information is eligible for residence permit endorsement; for older users, the endorsement terminal needs to provide large fonts, handwriting input and other functional support.

[0059] The training method of the endorsement service setting model includes:

[0060] Collect the B group of endorsement service setting data in advance, where B is an integer greater than 1. The endorsement service setting data includes characteristic data and setting serial numbers corresponding to the characteristic data; the characteristic data includes health vectors, education vectors, and age vectors arranged in a preset manner. The setting serial number represents the preset interface and interaction method corresponding to the characteristic data in the endorsement terminal.

[0061] It should be noted that arranging the health vector, education vector and age vector in a preset manner has the following two effects:

[0062] Ensure the consistency of model input data: The fixed vector arrangement method makes the endorsement service setting model have a clear structure and uniform format when receiving data, thereby avoiding confusion and errors in the model recognition process.

[0063] Protect user privacy: Each server only sends the corresponding feature vector based on the user's identity information, and does not directly transmit the user's real information; the mode setting server uses its own preset vector arrangement method to train the endorsement service setting model. Even if the usage record data of the later model is accidentally leaked, the specific user privacy information cannot be restored from it. Since the mode setting server does not know the actual meaning of the feature vector when receiving the data, the user's privacy is always protected during the entire data transmission and processing process and will not be leaked due to the transmission of feature data.

[0064] Each group of endorsement service setting data is used as the input of the endorsement service setting model, the endorsement service setting model takes the setting serial number corresponding to each group of endorsement service setting data as the output, takes the actual setting serial number corresponding to each group of endorsement service setting data as the prediction target, and takes minimizing the sum of the loss function values ​​of all the endorsement service setting data as the training target; the endorsement service setting model is trained until the sum of the prediction errors reaches convergence and the training is stopped; the endorsement service setting model is specifically a deep neural network model.

[0065] The loss function value:

[0066]

[0067] In the formula, is the loss function value; B is the number of data sets set for the endorsement service; Set the prediction function of the model for the endorsement service, based on the parameters Set data x for the bth endorsement service (b) Make prediction output; y (b) Set the actual output corresponding to the data for the b-th endorsement service, x (b) Set data for the bth endorsement service; Set the parameters of the model for the endorsement service, namely the weights of the health vector, education vector, and age vector.

[0068] Optimize the parameters of the endorsement service setting model by optimizing algorithm training The methods include:

[0069] Step 1: Randomly generate H groups of initial solutions, namely nests, each nest represents a set of parameters for the annotation service setting model as follows:

[0070]

[0071] In the formula, Set the model output for the endorsement service; J is the health vector; X is the education vector; N is the age vector; is the health vector weight; is the education vector weight; is the age vector weight.

[0072] Step 2: Iterate and update to generate a new nest. The method of generating a new nest includes:

[0073]

[0074] In the formula, is the parameter of the tth iteration; is the parameter of the t+1th iteration; α is the step size factor; Levy(λ) is the random step size that obeys the Levy distribution. Levy flight helps to jump out of the local optimal solution; λ is the scale parameter.

[0075] Step 3: Use Calculate the loss function value of the new nest.

[0076] Step 4: If the loss function value of the new nest is smaller than the loss function value of any nest in group H, the nest corresponding to the loss function value greater than the new nest in group H is replaced with the new nest with a preset probability.

[0077] When the preset number of iterations is reached, the nest with the smallest loss function value is selected as the parameter of the endorsement service setting model

[0078] In the complex parameter space of the endorsement service setting model, traditional algorithms are often prone to falling into local optimal solutions, limiting the potential for model optimization. By introducing the optimization algorithm, the parameter space can be explored more comprehensively and the overall performance can be improved. Specifically, the optimization algorithm uses a combination of parameters with different weights, and the complex interactions between these parameters constitute a huge parameter space. In this process, the optimization algorithm simulates the nest parasitic behavior of birds, abandons the current solution with a certain probability and generates a new solution, so that there is a greater chance of finding the global optimal parameter combination. This strategy can not only break through the local optimal dilemma, but also significantly improve the recommendation accuracy of the endorsement service setting model when facing different setting numbers, which helps to provide users with more accurate and personalized services.

[0079] In this embodiment, the certificate information is issued by the signing server, and various servers send the health vector, education vector, and social security vector corresponding to the certificate information to the mode setting server. The mode setting server inputs these vectors into the signing service setting model according to the preset method, outputs the setting serial number, and sends the setting serial number to the signing server. The signing server sets the corresponding adjustment interface and interaction method of the signing terminal. The setting serial number represents the preset interface and interaction method corresponding to each type of user (including normal users and special users) in the signing terminal. The whole process adaptively sets the interface and interaction method of the signing terminal, taking into account that special groups of users are not familiar with specific setting operations, and adaptively completes the setting of relevant interfaces and interaction methods without user intervention. Accessible residence permit signing services are provided for special groups of users, effectively improving the signing success rate of each type of user.

[0080] This embodiment is based on an artificial intelligence barrier-free residence permit endorsement service system and can also be applied to other similar endorsement services. It only requires targeted replacement of feature data, such as driver's license replacement or other types of services that require users to go to a specific endorsement terminal to complete.

[0081] Example 2

[0082] See also Figure 3 As shown, this embodiment provides an artificial intelligence-based barrier-free residence permit endorsement service method, which is applied in an endorsement server and includes:

[0083] Identify the user's certificate information;

[0084] Send the certificate information to the query chain; the nodes of the query chain are composed of an endorsement server, a social security query server, a medical query server, an education query server and a mode setting server; the certificate information is sent by the endorsement server;

[0085] Receive the setting sequence number sent by the mode setting server, and set the corresponding adjustment interface and interaction mode of the endorsement terminal according to the setting sequence number;

[0086] The mode setting server receives the social security vector fed back by the social security query server based on the certificate information, the health vector fed back by the medical query server based on the certificate information, and the academic qualification vector fed back by the social security query server based on the certificate information, and inputs them into the pre-trained endorsement service setting model in a preset manner, and outputs the setting serial number.

[0087] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0088] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. The barrier-free residence permit endorsement service method based on artificial intelligence is applied in the endorsement server, which is characterized by: include: Identify the user's certificate information; Send the certificate information to the query chain; The nodes of the query chain are composed of a signature server, a social security query server, a medical query server, an education query server and a mode setting server; the certificate information is issued by the signature server; Receive the setting sequence number sent by the mode setting server, and set the corresponding adjustment interface and interaction mode of the endorsement terminal according to the setting sequence number; The mode setting server receives the social security vector fed back by the social security query server based on the certificate information, the health vector fed back by the medical query server based on the certificate information, and the academic qualification vector fed back by the social security query server based on the certificate information, and inputs them into the pre-trained endorsement service setting model in a preset manner, and outputs the setting serial number.

2. The method for providing barrier-free residence permit endorsement service based on artificial intelligence according to claim 1 is characterized in that: The method for generating a setting sequence number comprises: Step 1: The mode setting server receives the continuous social security payment duration in the social security vector fed back by the social security query server according to the certificate information, and determines whether the certificate information is eligible for residence permit endorsement; Step 2. If the person is eligible for residence permit endorsement, the mode setting server receives the health vector fed back by the medical query server based on the certificate information, the education vector fed back by the social security query server based on the certificate information, and the age vector in the social security vector fed back by the social security query server based on the certificate information, and inputs them into the pre-trained endorsement service setting model in a preset manner, and outputs the setting serial number.

3. The method for providing barrier-free residence permit endorsement service based on artificial intelligence according to claim 1 is characterized in that: Methods for determining whether the certificate information is eligible for residence permit endorsement include: If the continuous payment period of social security is greater than or equal to the preset value, then you are deemed eligible for residence permit endorsement, otherwise you are not eligible.

4. The method for providing barrier-free residence permit endorsement service based on artificial intelligence according to claim 1 is characterized in that: The health vector includes the relevant disease vector representation and position serial number corresponding to the certificate information; the position serial number and the relevant disease vector representation are fixed positions in the health vector; the relevant diseases include visual impairment, hearing impairment, cognitive impairment or psychological impairment, and the psychological impairment includes anxiety, depression or obsessive-compulsive disorder; The educational background vector includes the educational background corresponding to the certificate information; the social security vector includes the age corresponding to the certificate information and the length of time for continuous payment of social security at the location of the endorsement terminal.

5. The method for providing barrier-free residence permit endorsement service based on artificial intelligence according to claim 4 is characterized in that: The training method of the endorsement service setting model includes: Collecting group B of endorsement service setting data in advance, where B is an integer greater than 1, and the endorsement service setting data includes feature data and setting sequence numbers corresponding to the feature data; the feature data includes a health vector, an education vector, and an age vector arranged in a preset manner; Each group of endorsement service setting data is used as the input of the endorsement service setting model, the endorsement service setting model takes the setting serial number corresponding to each group of endorsement service setting data as the output, takes the actual setting serial number corresponding to each group of endorsement service setting data as the prediction target, and takes minimizing the sum of the loss function values ​​of all the endorsement service setting data as the training target; the endorsement service setting model is trained until the sum of the prediction errors reaches convergence and the training is stopped; the endorsement service setting model is specifically a deep neural network model.

6. The method for providing barrier-free residence permit endorsement service based on artificial intelligence according to claim 5 is characterized in that: The loss function value: In the formula, is the loss function value; B is the number of data sets set for the endorsement service; Set the prediction function of the model for the endorsement service, based on the parameters Set data x for the bth endorsement service (b) Make prediction output; y (b) Set the actual output corresponding to the data for the b-th endorsement service, x (b) Set data for the bth endorsement service; Set the parameters of the model for the endorsement service, namely the weights of the health vector, education vector, and age vector.

7. The method for providing barrier-free residence permit endorsement service based on artificial intelligence according to claim 6 is characterized in that: Optimize the parameters of the endorsement service setting model by optimizing algorithm training The methods include: Step 1: Randomly generate H groups of initial solutions, namely nests, each nest represents a set of parameters for the annotation service setting model In the formula, Set the model output for the endorsement service; J is the health vector; X is the education vector; N is the age vector; is the health vector weight; is the education vector weight; is the age vector weight; Step 2: Iterate and update to generate a new nest; Step 3: Use Calculate the loss function value of the new nest; Step 4: If the loss function value of the new nest is less than the loss function value of any nest in group H, then the nest corresponding to the loss function value greater than the new nest in group H is replaced with the new nest with a preset probability; When the preset number of iterations is reached, the nest with the smallest loss function value is selected as the parameter of the endorsement service setting model 8. The method for providing barrier-free residence permit endorsement service based on artificial intelligence according to claim 7 is characterized in that: Methods for producing new nests include: In the formula, is the parameter of the tth iteration; is the parameter of the t+1th iteration; α is the step size factor; Levy(λ) is the random step size that obeys the Levy distribution. Levy flight helps to jump out of the local optimal solution; λ is the scale parameter.

9. The barrier-free residence permit endorsement service system based on artificial intelligence is applied in the endorsement server to implement the barrier-free residence permit endorsement service method based on artificial intelligence according to any one of claims 1 to 8, characterized in that: include: A collection module is used to identify the user's certificate information; The first processing module sends the certificate information to the query chain; The nodes of the query chain are composed of a signature server, a social security query server, a medical query server, an education query server and a mode setting server; the certificate information is issued by the signature server; The second processing module is used to receive the setting serial number sent by the mode setting server, and set the corresponding adjustment interface and interaction mode of the endorsement terminal according to the setting serial number; The mode setting server receives the social security vector fed back by the social security query server based on the certificate information, the health vector fed back by the medical query server based on the certificate information, and the academic qualification vector fed back by the social security query server based on the certificate information, and inputs them into the pre-trained endorsement service setting model in a preset manner, and outputs the setting serial number.

Images