A government service intelligent customer service system and service method thereof

Abstract

The present invention discloses an intelligent customer service system for government services and a service method thereof. The present invention relates to the technical field of government service customer service. The present invention establishes a communication connection while popping up a conversation box for a pre-consultation user, obtains each consultation information typed by the pre-consultation user in the conversation box, converts the consultation information into ciphertext using a conversation timestamp for each consultation information identified as non-business intention, and transmits the ciphertext-converted data as a secret sequence of the corresponding consultation information to a government service customer service platform. The government service customer service platform decrypts the consultation information using a reply timestamp. In this way, the encryption process for consultation information and reply information is freed from the involvement of keys, and regular replacement of keys will not cause system compatibility problems, thereby ensuring the stable encryption and decryption processes of consultation information and reply information.

Description

Technical Field

[0001] The present invention relates to the technical field of government service customer service, and specifically to a government service intelligent customer service system and a service method thereof. Background Art

[0002] In the field of government information services, the government service intelligent customer service system uses natural language processing technology to analyze user intentions in user consultation content and provide services based on the government knowledge base. However, user consultation content often contains sensitive information. Existing technologies generally use key encryption technology to protect text data containing intentions. That is, after the user inputs, it is encrypted with a preset key and then transmitted to the background for analysis;

[0003] However, the above approach has significant problems in high-frequency scenarios with large volumes of government service inquiries and frequent interactions: fixed keys are stored in system memory or databases for a long time, and high-frequency calls lead to an exponentially increased risk of being intercepted by malicious programs. Furthermore, regular key replacement to reduce the risk of leakage can cause system compatibility issues. Dynamic key replacement is often accompanied by encryption algorithm version upgrades (such as from AES-128 to AES-256), but some old components in the system (such as legacy API interfaces and embedded hardware modules) do not support the new algorithm, causing compatibility failures.

[0004] In order to solve the above problems, the present invention proposes a solution. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent customer service system for government affairs services and a service method thereof, in order to solve the problems raised in the above-mentioned background technology.

[0006] The present invention provides a service method of an intelligent customer service system for government services, comprising the following steps:

[0007] Step 1: After the current pre-consultation user clicks the conversation button, the client module pops up a conversation box for the current pre-consultation user and monitors the conversation box, and simultaneously establishes a communication connection with the government service customer service platform;

[0008] Step 2: When it is detected that the current pre-consultation user types a consultation message in the conversation box, the client module obtains the consultation message;

[0009] Step 3: Selecting the first ciphertext generation rule or the second ciphertext generation rule to generate the secret sequence of the consultation information and the session prefix stamp of the current pre-consultation user's current consultation based on whether the session prefix stamp of the current pre-consultation user's current consultation is temporarily stored in the client module;

[0010] Step 4: The client module transmits the encrypted sequence of the consultation information to the government service customer service platform, which receives and decrypts it to restore the consultation information;

[0011] Step 5: The government service customer service platform inputs the restored consultation information into the government service big model to obtain reply information of the consultation information.

[0012] Furthermore, in step one, after the communication connection is successfully established, the client module and the government service customer service platform synchronously record the time of successful establishment, and convert the time into a timestamp as the session timestamp and reply timestamp of the current pre-consultation user for this consultation for temporary storage.

[0013] Furthermore, in step three, if the client module temporarily stores the session prefix stamp of the current pre-consulting user's current consultation, the secret sequence of the consultation information and the session prefix stamp of the current pre-consulting user's current consultation are generated according to the second ciphertext generation rule.

[0014] Furthermore, in step 4, during the decryption process, if the government service customer service platform stores the session prefix stamp of the current pre-consultation user's current consultation, the session prefix stamp and session timestamp of the current pre-consultation user's current consultation are obtained from the government service customer service platform, and the secret sequence is decrypted and restored based on the stamp to obtain the corresponding consultation information, and the consultation information is transmitted to the government big model for input, and the government big model outputs data based on the consultation information as the reply information of the consultation information;

[0015] At this time, if the reply prefix stamp of the current pre-consultation user is not temporarily stored in the government service customer service platform, then according to steps S11 to S17, a secret sequence of the reply information and a reply prefix stamp of the current pre-consultation user are generated, the reply prefix stamp is temporarily stored, and the secret sequence is transmitted to the client module;

[0016] After receiving the transmitted secret sequence, the client module uses the stored reply timestamp of the current pre-consulting user's current consultation to decrypt and restore the secret sequence to obtain reply information, and displays the reply information in the conversation box.

[0017] An intelligent customer service system for government services, comprising:

[0018] A client module, configured to pop up a conversation box for the current pre-consulting user and monitor the conversation box after the current pre-consulting user clicks a conversation button;

[0019] The client module is further configured to acquire the consultation information upon detecting that the current pre-consultation user types a consultation information in the conversation box;

[0020] The client module is also used to identify the intention of the consultation information. If it is identified as non-business intention, the consultation information is directly transmitted to the government service customer service platform. Otherwise, the first ciphertext generation rule or the second ciphertext generation rule is selected according to whether the session prefix stamp of the current pre-consultation user is temporarily stored in the client module to generate a secret sequence of the consultation information and a session prefix stamp of the current pre-consultation user, and the secret sequence is transmitted to the government service customer service platform.

[0021] The government service customer service platform is used to generate reply information of the consultation information based on the consultation information of the pre-consultation user.

[0022] Compared with the existing technology, it has the following beneficial effects:

[0023] The present invention establishes a communication connection at the same time as a dialog box pops up for the pre-consultation user, and temporarily stores the time when the communication connection is successfully established as the conversation timestamp and reply timestamp of this consultation in the client module and the government service customer service platform respectively. Each consultation information typed by the pre-consultation user in the dialog box is obtained, and for each consultation information identified as non-business intention, the session timestamp is used to convert it into ciphertext, and the ciphertext-converted data is transmitted as the secret sequence of the corresponding consultation information to the government service customer service platform. The government service customer service platform decrypts it through the reply timestamp and generates its reply information using the government big model. At the same time, the reply information is ciphertext-converted according to the session timestamp to generate the corresponding secret sequence, and the client module decrypts it using the reply timestamp. In this way, the encryption process for the consultation information and the reply information is freed from the participation of the key, and there will be no system compatibility problems caused by regular key replacement, thereby ensuring the stable encryption and decryption process of the consultation information and the reply information.

[0024] In the process of encrypting the consultation information and reply information, the present invention uses the session timestamp and reply timestamp to perform initial encryption and decryption for the first consultation information and the first reply information sent by the pre-consultation user, respectively. In the initial encryption and decryption, the binary conversion results of the consultation information and the reply information are grouped to determine the assignment frequency, and the assignment frequency is used as the basis for the number of execution times of the assignment encryption operation. Each number of execution arrays is sequentially assigned with digital numbers, and then a number of missing characters are determined according to the composition of the characters in the session timestamp and the reply timestamp. Based on the digital number combined with the characters in the session time or the reply timestamp, the missing characters are determined. The key mapping array of the different execution arrays in each key mapping operation is determined by the symbol composition. In this way, the same execution array based on different key mapping operations and different assigned digital numbers has the same key mapping array, which effectively increases the difficulty for attackers to infer the key from historical ciphertexts and further enhances the randomness of the encryption result. In addition, for each subsequent consultation message and reply message, the encryption and decryption process is based on the missing characters determined during the encryption process of the previous consultation message and reply message, as well as the updated new session timestamp and reply timestamp. In this way, the encryption and decryption method for the consultation information is dynamic.

[0025] The present invention also assigns the corresponding execution array to the missing characters determined during the encryption and decryption process of each consultation information and reply information, so that the secret mapping array that has no direct connection with the original information is mixed into the secret mapping sequence, resulting in semantic confusion of the encryption and decryption results. In addition, the missing characters recorded in the encryption are mapped as part of the prefix stamp in the subsequent encryption, making the encryption logic present a forward dependency feature. Attackers cannot infer the complete encryption and decryption rules based on a single ciphertext, which greatly increases the cost of reverse engineering. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is a system block diagram of the present invention;

[0027] Figure 2 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0028] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0029] See also Figure 1 、 Figure 2 , this application provides a government service intelligent customer service system and service method thereof, including a client module and a government service customer service platform;

[0030] A client module is used for pre-consulting users to conduct government affairs consultation. The client module provides a conversation button for the pre-consulting user. The pre-consulting user can click the conversation button to enter the consultation information in the conversation box that pops up for the pre-consulting user in the client module;

[0031] In this application, when the pre-consulting user clicks the conversation button, the client module pops up a conversation box for the pre-consulting user and establishes a communication connection with the government service customer service platform;

[0032] It should be noted here that one client module serves one pre-consultation user;

[0033] When the current pre-consulting user clicks the conversation button, the client module pops up a conversation box for the current pre-consulting user and monitors the conversation box. Simultaneously, when the conversation box pops up, a communication connection is established with the government service customer service platform. After the communication connection is successfully established, the client module and the government service customer service platform synchronously record the time of successful establishment, convert the time into a timestamp, and temporarily store it as the conversation timestamp and reply timestamp of the current pre-consulting user's current consultation. That is, the client module and the government service customer service platform temporarily store two copies of the time, respectively.

[0034] When the current pre-consultation user is detected typing a consultation message in the conversation box, the intention recognition algorithm is first used to identify the intention of the consultation message. If the intention of the consultation message is identified as non-business intention, the consultation message is transmitted to the government service customer service platform, where non-business intention means that it does not involve specific service needs (such as consultation, complaint, business processing), and is only used to confirm the online status of the customer service or to start a conversation;

[0035] The government service customer service platform is used to provide government consultation services to pre-consultation users. The government service customer service platform is equipped with a government big model. After receiving the transmitted consultation information, the government service customer service platform inputs it into the government big model. The government big model outputs data based on the consultation information as reply information to the consultation information and transmits it to the client module. After receiving the reply information of the transmitted consultation message, the client module displays it in the conversation box.

[0036] Otherwise, the client module is first searched to see whether the session prefix stamp of the current pre-consulting user is temporarily stored in the client module. If not, the secret sequence of the consultation information and the session prefix stamp of the current pre-consulting user are generated according to the preset first ciphertext generation rule. The first ciphertext generation rule is as follows:

[0037] S11: performing a base conversion operation on the consulting information to obtain execution data of the consulting information, wherein the base conversion operation has the following contents: using a text base converter to perform binary conversion on the consulting information;

[0038] S12: Specifying a grouping step of 4 characters, grouping the execution data in order from left to right to obtain a plurality of execution arrays, and labeling all the execution arrays obtained by grouping the execution data as A1, A2, ..., Aa, from left to right according to their positions in the execution data, where a ≥ 1, and the number of characters in the execution array Aa is less than or equal to 4;

[0039] S13: Calculate the assignment frequency b of the execution data using the formula b=ceil(a / 9), where ceil() is a round-up function;

[0040] S14: Perform a keying operation on the execution arrays A1, A2, ..., A9 to obtain a keying sequence of the execution arrays, as follows:

[0041] S141: Assign numerical numbers 1, 2, ..., 9 to the execution arrays A1, A2, ..., A9 in ascending order of the marking subscripts. At this time, the execution array A1 is numbered 1, and the execution arrays A2, A3, ..., A9 are numbered in the same way.

[0042] S142: Obtain the temporarily stored session timestamp of the current pre-consultation user's current consultation, and determine whether the session timestamp is empty. The determination content is as follows:

[0043] Comparing the numbers 1, 2, ..., 9 with each character constituting the session timestamp in the order of 1, 2, ..., 9 for consistency; during the comparison process, if any number is inconsistent with any character constituting the session timestamp, it is first determined that the session timestamp is missing, and then the character is determined to be a missing character in the session timestamp;

[0044] After the consistency comparison between the number 9 and each character constituting the session timestamp is completed, if the session timestamp is determined to be blank, all blank characters determined to be in the session timestamp are obtained; otherwise, no processing is performed;

[0045] S143: Mark each non-zero character in the session timestamp as C1, C2, ..., Cc in order from left to right, where c≥1. That is, when a character that is zero is encountered, the character is skipped during marking and is not marked.

[0046] If it is determined that there are missing characters in the session timestamp, all missing characters in the session timestamp are marked as D1, D2, ..., Dd in sequence according to the determined order, where d≥1;

[0047] S144: Determine the key mapping array for character C1 from execution arrays A1, A2, ..., A9. The process is as follows: Obtain the execution array whose number matches that of character C1 and determine the execution array as the key mapping array for character C1. Similarly, determine the key mapping arrays for characters C2, C3, ..., Cc from execution arrays A1, A2, ..., A9 in sequence.

[0048] S145: Determine the key mapping array of the missing characters D1, D2, ..., Dd from the execution arrays A1, A2, ..., A9 according to S144. It should be noted that if the session timestamp is determined to be not missing, step S145 is not executed;

[0049] S146: Concatenate the key mapping arrays of the characters C1, C2, ..., Cc in the order of characters C1, C2, ..., Cc to obtain a key mapping sequence of the session timestamp. It should be noted that if the session timestamp is determined to be missing, concatenate the key mapping arrays of the characters C1, C2, ..., Cc and the missing characters D1, D2, ..., Dd in the order of characters C1, C2, ..., Cc, D1, D2, ..., Dd to obtain a key mapping sequence of the session timestamp.

[0050] S15: Perform a keying operation on execution arrays A10, A11, ..., A18, execution arrays A19, A20, ..., A27, ..., Aa-8, Aa-7, ..., Aa in sequence according to S14 to obtain b-1 keying sequences of the session timestamp. It should be noted that if the number of characters in execution array Aa is less than 4, perform a keying operation on Aa-8, Aa-7, ..., Aa-1 and concatenate the execution array Aa to the rightmost end of the obtained keying sequences.

[0051] S16: Concatenate the b key mapping sequences of the session timestamps in the order in which each key mapping sequence is obtained to obtain the key mapping sequence of the consultation information;

[0052] S17: If the session timestamp is not determined to be missing, all characters with an appearance frequency greater than 1 are extracted from the session timestamp, and all the extracted characters are spliced ​​from left to right according to the position of the characters in the session timestamp to obtain the session prefix stamp of the current pre-consultation user for this consultation. It should be noted here that if the session timestamp is determined to be missing, the missing characters D1, D2, ..., Dd are first spliced ​​in the order of missing characters D1, D2, ..., Dd to obtain the missing prefix stamp of the current consultation, and then all characters with an appearance frequency greater than 1 are extracted from the session timestamp, and all the extracted characters are spliced ​​from left to right according to the position of the characters in the session timestamp, and the spliced ​​data are spliced ​​with the missing prefix stamp to obtain the session prefix stamp of the current pre-consultation user for this consultation. In the process of splicing to obtain the session prefix stamp, the missing prefix stamp is spliced ​​at the rightmost end;

[0053] After the session prefix stamp of the current pre-consultation user is generated, the temporarily stored session timestamp of the current pre-consultation user is deleted;

[0054] The client module temporarily stores the session prefix stamp generated for the current pre-consultation user and sets the coverage time to 30 minutes. If the session prefix stamp is not covered within 30 minutes, the session prefix stamp is deleted;

[0055] It should be noted here that if the current pre-consulting user closes the conversation box, the client module will immediately delete the conversation prefix stamp of the current pre-consulting user stored in it;

[0056] If it exists, the second ciphertext generation rule is used to generate the secret sequence of the consultation information and the session prefix stamp of this consultation. The second ciphertext generation rule is as follows:

[0057] S21: performing a base conversion operation on the consulting information to obtain execution data of the consulting information, wherein the base conversion operation comprises the following steps: performing binary conversion on the consulting information using a text base converter;

[0058] S22: Specifying a grouping step of 4 characters, grouping the execution data in order from left to right to obtain a plurality of execution arrays, and labeling all the execution arrays obtained by grouping the execution data as E1, E2, ..., Ee, from left to right according to their positions in the execution data, where e≥1, and the number of characters in the execution array Ee is less than or equal to 4;

[0059] S23: Calculate the assignment frequency f of the execution data using the formula f=ceil(e / 9);

[0060] S24: Perform a keying operation on the execution arrays E1, E2, ..., E9 to obtain a keying sequence of the execution arrays, as follows:

[0061] S241: Assign numerical numbers 1, 2, ..., 9 to the execution arrays E1, E2, ..., E9 in ascending order of the marking subscripts. At this time, the execution array E1 is numbered 1, and the numbers of the execution arrays E2, E3, ..., E9 are numbered in the same way;

[0062] S242: Obtain the temporarily stored session timestamp and session prefix stamp of the current pre-consulting user's current consultation, concatenate the session timestamp and the session prefix stamp to obtain a new session timestamp for the current pre-consulting user's current consultation, and determine whether the session timestamp is missing. It should be noted that in the process of concatenating the new session timestamp, the session prefix stamp is concatenated to the rightmost end of the session timestamp. The determination is as follows:

[0063] Comparing the numbers 1, 2, ..., 9 with each character constituting the session timestamp in the order of 1, 2, ..., 9 for consistency; during the comparison process, if any number is inconsistent with any character constituting the session timestamp, it is first determined that the session timestamp is missing, and then the character is determined to be a missing character in the session timestamp;

[0064] After the consistency comparison between the number 9 and each character constituting the session timestamp is completed, if the session timestamp is determined to be blank, all blank characters determined to be in the session timestamp are obtained; otherwise, no processing is performed;

[0065] S243: Mark each non-zero character in the session timestamp as G1, G2, ..., Gg in order from left to right, where g≥1. That is, when a character that is zero is encountered, the character is skipped during marking and is not marked.

[0066] If it is determined that there are missing characters in the session timestamp, all the missing characters in the session timestamp are marked as H1, H2, ..., Hh in sequence according to the determined order, where h≥1;

[0067] S244: Determine the key mapping array for character G1 from execution arrays E1, E2, ..., E9. The process is as follows: Obtain the execution array whose number matches that of character G1 and determine the execution array as the key mapping array for character G1. Similarly, determine the key mapping arrays for characters G2, G3, ..., Gg from execution arrays E1, E2, ..., E9 in sequence.

[0068] S245: Determine the key mapping array of the missing characters H1, H2, ..., Hh from the execution arrays E1, E2, ..., E9 according to S244. It should be noted that if the session timestamp is determined to be not missing, step S245 is not executed;

[0069] S246: Concatenate the key mapping arrays of the characters G1, G2, ..., Gg in the order of characters G1, G2, ..., Gg to obtain a key mapping array of the session timestamp. It should be noted that if the session timestamp is determined to be missing, concatenate the key mapping arrays of the characters G1, G2, ..., Gg and the missing characters H1, H2, ..., Hh in the order of characters G1, G2, ..., Gg to obtain a key mapping array of the session timestamp.

[0070] S25: Perform a keying operation on execution arrays E10, E11, ..., E18, execution arrays E19, E20, ..., E27, ..., Ee-8, Ee-7, ..., Ee in sequence according to S24 to obtain f-1 keying sequences of the session timestamp. It should be noted that if the number of characters in execution array Ee is less than 4, perform a keying operation on Ee-8, Ee-7, ..., Ee-1 and concatenate the execution array Ee to the rightmost end of the obtained keying sequence.

[0071] S26: Concatenate the obtained f secret mapping sequences of the session timestamp in the order in which each secret mapping sequence is obtained to obtain the secret mapping sequence of the consultation information;

[0072] S27: If the session timestamp is not determined to be missing, all characters with an appearance frequency greater than 1 are extracted from the session timestamp, and all the extracted characters are spliced ​​from left to right according to the position of the characters in the session timestamp to obtain the session prefix stamp of this consultation. It should be noted here that if the session timestamp is determined to be missing, the missing characters H1, H2, ..., Hh are first spliced ​​in the order of missing characters H1, H2, ..., Hh to obtain the missing prefix stamp of this consultation, and then all characters with an appearance frequency greater than 1 are extracted from the session timestamp, and all the extracted characters are spliced ​​from left to right according to the position of the characters in the session timestamp, and the spliced ​​data are spliced ​​with the missing prefix stamp to obtain the session prefix stamp of this consultation. In the process of splicing to obtain the session prefix stamp, the missing prefix stamp is spliced ​​at the rightmost end;

[0073] The client module overwrites and stores the stored session prefix stamp according to the generated session prefix stamp of the current pre-consultation user;

[0074] The client module transmits the generated secret sequence of the consultation information to the government service customer service platform;

[0075] After receiving the encrypted sequence of the transmitted consultation information, the government service customer service platform decrypts the encrypted sequence according to a preset decryption step;

[0076] During the decryption process, if the government service customer service platform stores the session prefix stamp of the current pre-consultation user's current consultation, the session prefix stamp and session timestamp of the current pre-consultation user's current consultation are obtained from the government service customer service platform, and the secret sequence is decrypted and restored based on the stamp to obtain the corresponding consultation information, and the consultation information is transmitted to the government big model for input, and the government big model outputs data based on the consultation information as the reply information of the consultation information;

[0077] At this time, if the reply prefix stamp of the current pre-consultation user is not temporarily stored in the government service customer service platform, then according to steps S11 to S17, a secret sequence of the reply information and a reply prefix stamp of the current pre-consultation user are generated, the reply prefix stamp is temporarily stored, and the secret sequence is transmitted to the client module;

[0078] After receiving the transmitted secret sequence, the client module uses the stored timestamp of the current pre-consulting user's current consultation to decrypt and restore the secret sequence to obtain a reply message, and displays the reply message in the conversation box;

[0079] At this time, if the reply prefix stamp of the current pre-consultation user is temporarily stored in the government service customer service platform, a secret sequence of the reply information and the reply prefix stamp of the current pre-consultation user are generated according to steps S21 to S27, the reply prefix stamp is overwritten and stored, and the secret sequence is transmitted to the client module;

[0080] After receiving the transmitted secret sequence, the client module uses the stored timestamp of the current pre-consulting user's current consultation to decrypt and restore the secret sequence to obtain a reply message, and displays the reply message in the conversation box;

[0081] During the decryption process, if the government service customer service platform does not store the session prefix stamp of the current pre-consulting user's current consultation, the session timestamp of the current pre-consulting user's current consultation is obtained from the government service customer service platform, and the secret sequence is decrypted and restored based on the timestamp to obtain the corresponding consultation information, and the consultation information is transmitted to the government big model for input, and the government big model outputs data based on the consultation information as the reply information of the consultation information;

[0082] At this time, if the reply prefix stamp of the current pre-consultation user is not temporarily stored in the government service customer service platform, then according to steps S11 to S17, a secret sequence of the reply information and a reply prefix stamp of the current pre-consultation user are generated, the reply prefix stamp is temporarily stored, and the secret sequence is transmitted to the client module;

[0083] After receiving the transmitted secret sequence, the client module uses the stored timestamp of the current pre-consulting user's current consultation to decrypt and restore the secret sequence to obtain a reply message, and displays the reply message in the conversation box;

[0084] At this time, if the reply prefix stamp of the current pre-consultation user is temporarily stored in the government service customer service platform, a secret sequence of the reply information and the reply prefix stamp of the current pre-consultation user are generated according to steps S21 to S27, the reply prefix stamp is overwritten and stored, and the secret sequence is transmitted to the client module;

[0085] After receiving the transmitted secret sequence, the client module uses the stored timestamp of the current pre-consulting user's current consultation to decrypt and restore the secret sequence to obtain a reply message, and displays the reply message in the conversation box;

[0086] Some of the data in the above formulas are dimensionless and numerically calculated. Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0087] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A government service intelligent customer service method, characterized in that: The following steps are involved: Step 1: After the current pre-consulting user clicks the conversation button, the client module pops up a conversation box for the current pre-consulting user and monitors the conversation box, and simultaneously establishes a communication connection with the government service customer service platform. After the communication connection is successfully established, the client module and the government service customer service platform simultaneously record the time of successful establishment, and convert the time into a timestamp as the conversation timestamp and reply timestamp of the current pre-consulting user for temporary storage; Step 2: When it is detected that the current pre-consultation user types a consultation message in the conversation box, the client module obtains the consultation message; Step 3: Selecting a first ciphertext generation rule or a second ciphertext generation rule based on whether the client module temporarily stores the session prefix stamp of the current pre-consulting user's current consultation, generates a secret sequence of the consultation information and a session prefix stamp of the current pre-consulting user's current consultation, where the session prefix stamp is a sequence of digits from 0 to 9; Step 4: The client module transmits the encrypted sequence of the consultation information to the government service customer service platform, which receives and decrypts it to restore the consultation information; Step 5: The government service customer service platform inputs the restored consultation information into the government service big model to obtain reply information of the consultation information; If the session prefix stamp of the current pre-consulting user is not temporarily stored, the secret sequence of the consultation information and the session prefix stamp of the current pre-consulting user are generated according to the first ciphertext generation rule, as follows: S11: performing a base-based conversion operation on the consultation information to obtain execution data of the consultation information; S12: Specifying a grouping step of 4 characters, grouping the execution data in order from left to right to obtain a plurality of execution arrays, and labeling all the execution arrays obtained by grouping the execution data as A1, A2, ..., Aa, from left to right according to their positions in the execution data, where a ≥ 1, and the number of characters in the execution array Aa is less than or equal to 4; S13: Calculate the assignment frequency b of the execution data using the formula b=ceil(a / 9), where ceil() is a round-up function; S14: Perform a keying operation on the execution arrays A1, A2, ..., A9 to obtain a keying sequence of the execution arrays, as follows: S141: Assign numerical numbers 1, 2, ..., 9 to the execution arrays A1, A2, ..., A9 in ascending order of the marking subscripts. At this time, the execution array A1 is numbered 1, and the execution arrays A2, A3, ..., A9 are numbered in the same way. S142: Obtain the temporarily stored session timestamp of the current pre-consultation user's current consultation, determine whether the session timestamp is missing, and determine a number of missing characters in the session timestamp based on the determination result. The determination content is as follows: Comparing the numbers 1, 2, ..., 9 with each character constituting the session timestamp for consistency in sequence. During the comparison process, if any number is inconsistent with any character constituting the session timestamp, it is first determined that the session timestamp is missing, and then the character is determined to be a missing character in the session timestamp. After the consistency comparison between the number 9 and each character constituting the session timestamp is completed, if the session timestamp is determined to be blank, all blank characters determined to be in the session timestamp are obtained; otherwise, no processing is performed; S143: Mark each non-zero character in the session timestamp as C1, C2, ..., Cc, from left to right, where c≥1. If it is determined that there are missing characters in the session timestamp, mark all the missing characters in the session timestamp as D1, D2, ..., Dd, where d≥1, in the determined order. S144: Determine the key mapping array for character C1 from execution arrays A1, A2, ..., A9. The process is as follows: Obtain the execution array whose number matches that of character C1 and determine the execution array as the key mapping array for character C1. Similarly, determine the key mapping arrays for characters C2, C3, ..., Cc from execution arrays A1, A2, ..., A9 in sequence. S145: Determine the key mapping array of the missing characters D1, D2, ..., Dd from the execution arrays A1, A2, ..., A9 according to S144, wherein if the session timestamp is determined to be not missing, step S145 is not executed; S146: Concatenate the key mapping arrays of the characters C1, C2, ..., Cc in the order of characters C1, C2, ..., Cc to obtain a key mapping array of the session timestamp. If the session timestamp is determined to be missing, concatenate the key mapping arrays of the characters C1, C2, ..., Cc and the missing characters D1, D2, ..., Dd in the order of characters C1, C2, ..., Cc, D1, D2, ..., Dd to obtain a key mapping array of the session timestamp. S15: Perform a keying operation on the execution arrays A10, A11, ..., A18, the execution arrays A19, A20, ..., A27, ..., Aa-8, Aa-7, ..., Aa in sequence according to S14 to obtain b-1 keying sequences of the session timestamps; S16: Concatenate the b key mapping sequences of the session timestamps in the order in which each key mapping sequence is obtained to obtain the key mapping sequence of the consultation information; S17: When the session timestamp is determined to be not empty, extract all characters with a frequency greater than 1 from the session timestamp, and concatenate all the extracted characters from left to right according to their positions in the session timestamp to obtain a session prefix stamp for the current consultation of the pre-consulting user; If the session prefix stamp of the current pre-consulting user is temporarily stored, the secret sequence of the consultation information and the session prefix stamp of the current pre-consulting user are generated according to the second ciphertext generation rule; wherein the first ciphertext generation rule and the second ciphertext generation rule are identical in content, and differ only in that: The second ciphertext generation rule is based on the first ciphertext generation rule. Through step S142, the temporarily stored session timestamp and session prefix stamp of the current pre-consultation user for this consultation are obtained, and the session timestamp and the session prefix stamp are spliced ​​to obtain a new session timestamp of the current pre-consultation user for this consultation.

2. A government service intelligent customer service method according to claim 1, characterized in that: S17, if the session timestamp is determined to be missing, the missing characters D1, D2, ..., Dd are first concatenated in the order of the missing characters D1, D2, ..., Dd to obtain a missing prefix stamp for this consultation, and then all characters with an appearance frequency greater than 1 are extracted from the session timestamp, and all the extracted characters are concatenated from left to right according to the position of the characters in the session timestamp, and the concatenated data is concatenated with the missing prefix stamp to obtain the session prefix stamp for this consultation of the current pre-consulting user.

3. A government service intelligent customer service method according to claim 1, characterized in that: In S17 , after the session prefix stamp of the current pre-consultation user's current consultation is generated, the temporarily stored session timestamp of the current pre-consultation user's current consultation is deleted.

4. A government service intelligent customer service method according to claim 1, characterized in that: In step 3, if the client module temporarily stores the session prefix stamp of the current pre-consulting user's current consultation, the secret sequence of the consultation information and the session prefix stamp of the current pre-consulting user's current consultation are generated according to the second ciphertext generation rule.

5. The intelligent customer service method for government services according to claim 1, characterized in that: In step 4, during the decryption process, if the government service customer service platform stores the session prefix stamp of the current pre-consulting user's current consultation, the session prefix stamp and session timestamp of the current pre-consulting user's current consultation are obtained from the government service customer service platform, and the secret sequence is decrypted and restored based on the stamp to obtain the corresponding consultation information, and the consultation information is transmitted to the government big model for input, and the government big model outputs data based on the consultation information as the reply information of the consultation information; At this time, if the reply prefix stamp of the current pre-consultation user is not temporarily stored in the government service customer service platform, then according to steps S11 to S17, a secret sequence of the reply information and a reply prefix stamp of the current pre-consultation user are generated, the reply prefix stamp is temporarily stored, and the secret sequence is transmitted to the client module; After receiving the transmitted secret sequence, the client module uses the stored reply timestamp of the current pre-consulting user's current consultation to decrypt and restore the secret sequence to obtain reply information, and displays the reply information in the conversation box.

6. An intelligent customer service system for government services, characterized in that: include: A client module, configured to pop up a conversation box for the current pre-consulting user and monitor the conversation box after the current pre-consulting user clicks a conversation button; The client module is further configured to acquire the consultation information upon detecting that the current pre-consultation user types a consultation information in the conversation box; The client module is also used to identify the intention of the consultation information. If it is identified as non-business intention, the consultation information is directly transmitted to the government service customer service platform. Otherwise, the first ciphertext generation rule or the second ciphertext generation rule is selected according to whether the session prefix stamp of the current pre-consultation user is temporarily stored in the client module to generate a secret sequence of the consultation information and a session prefix stamp of the current pre-consultation user, and the secret sequence is transmitted to the government service customer service platform. The government service customer service platform is used to generate reply information of the consultation information according to the consultation information of the pre-consultation user; If the session prefix stamp of the current pre-consulting user is not temporarily stored, the secret sequence of the consultation information and the session prefix stamp of the current pre-consulting user are generated according to the first ciphertext generation rule, as follows: S11: performing a base-based conversion operation on the consultation information to obtain execution data of the consultation information; S12: Specifying a grouping step of 4 characters, grouping the execution data in order from left to right to obtain a plurality of execution arrays, and labeling all the execution arrays obtained by grouping the execution data as A1, A2, ..., Aa, from left to right according to their positions in the execution data, where a ≥ 1, and the number of characters in the execution array Aa is less than or equal to 4; S13: Calculate the assignment frequency b of the execution data using the formula b=ceil(a / 9), where ceil() is a round-up function; S14: Perform a keying operation on the execution arrays A1, A2, ..., A9 to obtain a keying sequence of the execution arrays, as follows: S141: Assign numerical numbers 1, 2, ..., 9 to the execution arrays A1, A2, ..., A9 in ascending order of the marking subscripts. At this time, the execution array A1 is numbered 1, and the execution arrays A2, A3, ..., A9 are numbered in the same way. S142: Obtain the temporarily stored session timestamp of the current pre-consultation user's current consultation, determine whether the session timestamp is missing, and determine a number of missing characters in the session timestamp based on the determination result. The determination content is as follows: Comparing the numbers 1, 2, ..., 9 with each character constituting the session timestamp for consistency in sequence. During the comparison process, if any number is inconsistent with any character constituting the session timestamp, it is first determined that the session timestamp is missing, and then the character is determined to be a missing character in the session timestamp. After the consistency comparison between the number 9 and each character constituting the session timestamp is completed, if the session timestamp is determined to be blank, all blank characters determined to be in the session timestamp are obtained; otherwise, no processing is performed; S143: Mark each non-zero character in the session timestamp as C1, C2, ..., Cc, from left to right, where c≥1. If it is determined that there are missing characters in the session timestamp, mark all the missing characters in the session timestamp as D1, D2, ..., Dd, where d≥1, in the determined order. S144: Determine the key mapping array for character C1 from execution arrays A1, A2, ..., A9. The process is as follows: Obtain the execution array whose number matches that of character C1 and determine the execution array as the key mapping array for character C1. Similarly, determine the key mapping arrays for characters C2, C3, ..., Cc from execution arrays A1, A2, ..., A9 in sequence. S145: Determine the key mapping array of the missing characters D1, D2, ..., Dd from the execution arrays A1, A2, ..., A9 according to S144, wherein if the session timestamp is determined to be not missing, step S145 is not executed; S146: Concatenate the key mapping arrays of the characters C1, C2, ..., Cc in the order of characters C1, C2, ..., Cc to obtain a key mapping array of the session timestamp. If the session timestamp is determined to be missing, concatenate the key mapping arrays of the characters C1, C2, ..., Cc and the missing characters D1, D2, ..., Dd in the order of characters C1, C2, ..., Cc, D1, D2, ..., Dd to obtain a key mapping array of the session timestamp. S15: Perform a keying operation on the execution arrays A10, A11, ..., A18, the execution arrays A19, A20, ..., A27, ..., Aa-8, Aa-7, ..., Aa in sequence according to S14 to obtain b-1 keying sequences of the session timestamps; S16: Concatenate the b key mapping sequences of the session timestamps in the order in which each key mapping sequence is obtained to obtain the key mapping sequence of the consultation information; S17: When the session timestamp is determined to be not empty, extract all characters with a frequency greater than 1 from the session timestamp, and concatenate all the extracted characters from left to right according to their positions in the session timestamp to obtain a session prefix stamp for the current consultation of the pre-consulting user; If the session prefix stamp of the current pre-consulting user is temporarily stored, the secret sequence of the consultation information and the session prefix stamp of the current pre-consulting user are generated according to the second ciphertext generation rule; wherein the first ciphertext generation rule and the second ciphertext generation rule are identical in content, and differ only in that: The second ciphertext generation rule is based on the first ciphertext generation rule. Through step S142, the temporarily stored session timestamp and session prefix stamp of the current pre-consultation user for this consultation are obtained, and the session timestamp and the session prefix stamp are spliced ​​to obtain a new session timestamp of the current pre-consultation user for this consultation.

Images