Account processing method, medium, device and computing equipment

Abstract

Embodiments of the present disclosure provide an account processing method, medium, apparatus, and computing device. The method comprises: obtaining an account to be registered; identifying character information for each character in the account to be registered using an element recognition model; and determining target elements in the account to be registered based on the character information corresponding to the multiple characters. The target elements include at least one of a pinyin element, an English word element, or an English name element. This method can identify a richer range of element types, thereby fully mining useful information in the account to be registered, providing more elements for the account recommendation process, ensuring the richness of recommended accounts, and improving the user experience.

Description

Technical Field

[0001] The embodiments of the present disclosure relate to the field of Internet technology, and more specifically, to account processing methods, media, apparatuses, and computing devices. Background Art

[0002] This section is intended to provide a background or context to the embodiments of the present disclosure that are recited in the claims. No statement herein is admitted to be prior art by virtue of its inclusion in this section.

[0003] As the internet continues to expand, more and more internet platforms are emerging. Each internet platform usually requires users to register a personal account before logging in, such as an email account, WeChat account, Weibo account, etc.

[0004] In recent years, the number of Internet users has gradually increased, and a large number of user accounts have been registered on various platforms. In order to prevent the occurrence of duplicate accounts on the same platform and bring security risks, when users register new accounts, account recommendations can be made based on the elements in the new account. However, currently only symbolic elements and numerical elements in the account can be extracted, which makes the recommended accounts limited and difficult to meet user needs. Summary of the Invention

[0005] The embodiments of the present disclosure provide an account processing method, medium, apparatus, and computing device for solving the technical problem that the extracted elements have limitations in the current element extraction process.

[0006] In a first aspect of the embodiments of the present disclosure, an account processing method is provided, including: obtaining an account to be registered, the account to be registered including multiple characters; identifying the character information of each character through an element recognition model, and determining the target element in the account to be registered based on the character information corresponding to the multiple characters, the target element including at least one of a pinyin element, an English word element, or an English name element, the element recognition model is trained based on the character information of each character in the registered account, and the character information is used to indicate the position of each character in the element and / or the element type of the element.

[0007] In one embodiment of the present disclosure, the feature recognition model includes: an encoding layer, a forward long short-term memory network LSTM, a reverse LSTM and a conditional random field CRF; the encoding layer is used to obtain the vector representation of each character in the account to be registered and generate a string vector of the account to be registered; the forward LSTM is used to obtain a first vector matrix corresponding to the string vector; the reverse LSTM is used to obtain a second vector matrix corresponding to the string vector; the CRF is used to perform feature recognition on the first vector matrix and the second vector matrix, obtain character information of each character, and obtain the target feature in the account to be registered based on the character information.

[0008] In one embodiment of the present disclosure, an element recognition model is obtained by: obtaining multiple registered accounts; identifying elements contained in the registered accounts, and determining the element type of each element, which includes pinyin, English words, and English names; determining the position information of each character in the element based on the position of each character in the element; obtaining the character information of each character in the element based on the element type of the element and the position information of each character in the element; training the initial element recognition model through the character information of each character in the registered account to obtain the element recognition model.

[0009] In one embodiment of the present disclosure, the account processing method further includes: generating a target account set according to target elements and preset generation rules, wherein the target account set includes at least one target account.

[0010] In one embodiment of the present disclosure, the preset generation rules include: a first generation rule and / or a second generation rule; generating a target account set based on the target elements and the preset generation rules, including: adjusting at least part of the characters in the target elements in the account to be registered according to the first generation rule to generate the target account set; and / or, adjusting the positions between multiple target elements in the account to be registered according to the second generation rule to generate the target account set.

[0011] In one embodiment of the present disclosure, the first generation rule includes at least one of a deletion rule, a character position adjustment rule and a replacement rule; according to the first generation rule, at least part of the characters in the target element in the account to be registered is adjusted to generate a target account set, including at least one of the following methods: according to the deletion rule, at least part of the characters of the pinyin element in the account to be registered is deleted to generate a target account set; according to the character position adjustment rule, the position between at least part of the characters of the pinyin element in the account to be registered is adjusted to generate a target account set; according to the replacement rule, the English word element in the account to be registered is replaced to generate a target account set.

[0012] In one embodiment of the present disclosure, after generating a target account set based on target elements and preset generation rules, the process further includes: determining a priority score for each target account in the target account set based on a string adjustment distance of each target account relative to the account to be registered, a first string length of the target account, and a second string length of the account to be registered; and outputting a recommendation list corresponding to the target account set based on the priority score of each target account in the target account set.

[0013] In one embodiment of the present disclosure, the priority score of each target account in the target account set is determined based on the string adjustment distance of each target account relative to the account to be registered, the second string length of the target account, and the first string length of the account to be registered, including: determining the adjustment distance influence score of the target account based on the string adjustment distance corresponding to each target account, the first string length of the target account, and the second string length of the account to be registered; determining the length influence score of the target account based on the first string length of the target account, the string length of the shortest target account in the target account set, and the string length of the longest target account in the target account set; determining the priority score of each target account in the target account set based on the distance influence score, the length influence score, and the preset adjustment coefficient of each target account.

[0014] In one embodiment of the present disclosure, after generating the target account set according to the target elements and preset generation rules, the method further includes: determining a first target account in the target account set that is duplicated with a registered account; and deleting the first target account in the target account set.

[0015] In one embodiment of the present disclosure, after obtaining the account to be registered, it also includes: determining whether the account to be registered is a duplicate of a registered account on the target platform; if the account to be registered is a duplicate of the registered account, determining that the account to be registered is an available account on the target platform; if the account to be registered is a duplicate of the registered account, identifying the character information of each character through the element recognition model, and determining the target element in the account to be registered based on the character information corresponding to multiple characters.

[0016] In one embodiment of the present disclosure, after identifying the character information of each character through the element recognition model and determining the target element in the account to be registered based on the character information corresponding to multiple characters, it also includes: determining the randomness score corresponding to the account to be registered based on the total length of the target element in the account to be registered and the length of the account to be registered; if the randomness score is less than a preset value, determining that the account to be registered is a risk account; if the randomness score is greater than or equal to the preset value, determining that the account to be registered is an available account of the target platform.

[0017] In one embodiment of the present disclosure, before identifying the character information of each character through an element recognition model and determining the target element in the account to be registered based on the character information corresponding to multiple characters, it also includes: extracting symbol characters and numeric characters in the account to be registered through regular expressions.

[0018] In a second aspect of the embodiments of the present disclosure, a computer-readable storage medium is provided, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, the account processing method of the first aspect is implemented.

[0019] In a third aspect of the embodiments of the present disclosure, an account processing device is provided, including: an acquisition module for acquiring an account to be registered, wherein the account to be registered includes multiple characters; an identification module for identifying the character information of each character through an element recognition model, and determining the target element in the account to be registered based on the character information corresponding to the multiple characters, wherein the target element includes at least one of a pinyin element, an English word element, or an English name element, and the element recognition model is trained based on the character information of each character in the registered account, and the character information is used to indicate the position of each character in the element and / or the element type of the element.

[0020] In one embodiment of the present disclosure, the feature recognition model includes: an encoding layer, a forward long short-term memory network LSTM, a reverse LSTM and a conditional random field CRF; the encoding layer is used to obtain the vector representation of each character in the account to be registered and generate a string vector of the account to be registered; the forward LSTM is used to obtain a first vector matrix corresponding to the string vector; the reverse LSTM is used to obtain a second vector matrix corresponding to the string vector; the CRF is used to perform feature recognition on the first vector matrix and the second vector matrix, obtain character information of each character, and obtain the target feature in the account to be registered based on the character information.

[0021] In one embodiment of the present disclosure, an element recognition model is obtained by: obtaining multiple registered accounts; identifying elements contained in the registered accounts, and determining the element type of each element, which includes pinyin, English words, and English names; determining the position information of each character in the element based on the position of each character in the element; obtaining the character information of each character in the element based on the element type of the element and the position information of each character in the element; training the initial element recognition model through the character information of each character in the registered account to obtain the element recognition model.

[0022] In one embodiment of the present disclosure, the account processing device further includes: a generation module, configured to generate a target account set according to target elements and preset generation rules, wherein the target account set includes at least one target account.

[0023] In one embodiment of the present disclosure, the preset generation rules include: a first generation rule and / or a second generation rule; the generation module is specifically used to: adjust at least part of the characters in the target elements in the account to be registered according to the first generation rule to generate a target account set; and / or, adjust the positions between multiple target elements in the account to be registered according to the second generation rule to generate a target account set.

[0024] In one embodiment of the present disclosure, the first generation rule includes at least one of a deletion rule, a character position adjustment rule and a replacement rule; the generation module specifically generates the target account set in at least one of the following ways: according to the deletion rule, deleting at least part of the characters of the pinyin elements in the account to be registered to generate the target account set; according to the character position adjustment rule, adjusting the positions between at least part of the characters of the pinyin elements in the account to be registered to generate the target account set; according to the replacement rule, replacing the English word elements in the account to be registered to generate the target account set.

[0025] In one embodiment of the present disclosure, the account processing device also includes: a first determination module, used to determine the priority score of each target account in the target account set based on the string adjustment distance of each target account relative to the account to be registered, the first string length of the target account, and the second string length of the account to be registered; a recommendation module, used to output a recommendation list corresponding to the target account set based on the priority score of each target account in the target account set.

[0026] In one embodiment of the present disclosure, the first determination module is specifically used to: determine the adjustment distance influence score of the target account based on the string adjustment distance corresponding to each target account, the first string length of the target account, and the second string length of the account to be registered; determine the length influence score of the target account based on the first string length of the target account, the string length of the shortest target account in the target account set, and the string length of the longest target account in the target account set; determine the priority score of each target account in the target account set based on the distance influence score, length influence score and preset adjustment coefficient of each target account.

[0027] In one embodiment of the present disclosure, the account processing device further includes: a second determination module for determining a first target account in the target account set that is a duplicate of a registered account; and a deletion module for deleting the first target account in the target account set.

[0028] In one embodiment of the present disclosure, the account processing device also includes: a third determination module, used to determine whether the account to be registered is a duplicate of a registered account on the target platform; if the account to be registered is a duplicate of the registered account, then determining that the account to be registered is an available account on the target platform; if the account to be registered is a duplicate of the registered account, then identifying the character information of each character through the element recognition model, and determining the target element in the account to be registered based on the character information corresponding to multiple characters.

[0029] In one embodiment of the present disclosure, the account processing device also includes: a fourth determination module, which is used to determine the randomness score corresponding to the account to be registered based on the total length of the target elements in the account to be registered and the length of the account to be registered; if the randomness score is less than a preset value, the account to be registered is determined to be a risk account; if the randomness score is greater than or equal to the preset value, the account to be registered is determined to be an available account of the target platform.

[0030] In one embodiment of the present disclosure, the account processing device further includes: an extraction module, configured to extract symbolic characters and numeric characters in the account to be registered by using a regular expression.

[0031] In a fourth aspect of the embodiments of the present disclosure, a computing device is provided, comprising: at least one processor and a memory; the memory stores computer-executable instructions; at least one processor executes the computer-executable instructions stored in at least one memory, so that the at least one processor executes the account processing method of the first aspect.

[0032] In a fifth aspect of the embodiments of the present disclosure, a computer program product is provided, the computer program product including a computer program; when the computer program is executed, the account processing method of the first aspect is implemented.

[0033] The account processing method, medium, apparatus, and computing device provided by the disclosed embodiments include: obtaining an account to be registered; identifying character information for each character in the account to be registered using an element recognition model; and determining target elements in the account to be registered based on the character information corresponding to the multiple characters. The target elements include at least one of a pinyin element, an English word element, or an English name element. This method can identify a richer range of element types, fully mining useful information in the account to be registered, thereby providing more elements for the account recommendation process, ensuring the richness of recommended accounts and improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The above and other objects, features and advantages of the exemplary embodiments of the present disclosure will become readily understood by reading the detailed description below with reference to the accompanying drawings, in which several embodiments of the present disclosure are shown by way of example and not limitation, wherein:

[0035] Figure 1 A schematic diagram of an application scenario provided by an embodiment of the present disclosure;

[0036] Figure 2 Schematic diagram of the account processing method provided in the embodiment of the present disclosure Figure 1 ;

[0037] Figure 3 A schematic diagram of the structure of the element recognition model provided in the embodiment of the present disclosure;

[0038] Figure 4 A flowchart of a method for training an element recognition model according to an embodiment of the present disclosure;

[0039] Figure 5 Schematic diagram of the account processing method provided in the embodiment of the present disclosure Figure 2 ;

[0040] Figure 6 A schematic diagram of a process for recommending a target account to a user according to an embodiment of the present disclosure;

[0041] Figure 7 Schematic diagram of the account processing method provided in the embodiment of the present disclosure Figure 3 ;

[0042] Figure 8 A schematic diagram of a program product provided by an embodiment of the present disclosure;

[0043] Figure 9 A schematic diagram of the structure of the account processing device provided in the embodiment of the present disclosure Figure 1 ;

[0044] Figure 10 A schematic diagram of the structure of the account processing device provided in the embodiment of the present disclosure Figure 2 ;

[0045] Figure 11 A schematic diagram of the structure of a computing device provided in an embodiment of the present disclosure.

[0046] In the drawings, the same or corresponding reference numerals denote the same or corresponding parts. DETAILED DESCRIPTION

[0047] The principles and spirit of the present disclosure will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and implement the present disclosure, and are not intended to limit the scope of the present disclosure in any way. On the contrary, these embodiments are provided to make the present disclosure more thorough and complete, and to fully convey the scope of the present disclosure to those skilled in the art. In the process of a user registering a platform account, it is easy for the new account set by the user to overlap with an account already registered on the platform, making it impossible to complete the registration.

[0048] Those skilled in the art will appreciate that the embodiments of the present disclosure may be implemented as a system, apparatus, device, method, or computer program product. Therefore, the present disclosure may be implemented in the following forms: entirely in hardware, entirely in software (including firmware, resident software, microcode, etc.), or in a combination of hardware and software.

[0049] According to an embodiment of the present disclosure, an account processing method, medium, apparatus, and computing device are proposed.

[0050] In addition, the number of any elements in the drawings is for illustration and not for limitation, and any names are for distinction only and do not have any limiting meaning. The principle and spirit of the present disclosure will be explained in detail with reference to several representative embodiments of the present disclosure. SUMMARY OF THE INVENTION

[0052] The inventors have discovered that current platforms typically recommend accounts based on elements of the user-entered account number, such as numeric elements and symbolic elements. For example, if the user enters the account number "zhangsan0123," the elements of this account number are the numeric elements "0123." When recommending accounts, the numeric elements can be replaced with other numbers, or the positions of the numbers can be changed to obtain recommended accounts. For example, the recommended accounts could be "zhangsan3210," "zhangsan5678," and so on.

[0053] However, since the modified characters are usually randomly generated, the random numbers contained in the account number will make the recommended account number difficult for users to remember, resulting in a poor user experience. In addition, since only the numeric elements and symbol elements in the account number can be recognized and only the numeric elements and symbol elements can be modified, the account numbers obtained by this method have certain limitations. Moreover, when the numbers in the user's account number input have special meanings, the recommended account numbers obtained by this solution can easily deviate from the meaning and thus fail to meet user needs. For example, 0123 may be used to represent the user's birthday as January 23. If it is modified to other numbers or the position of the numbers in the element is replaced, the recommended account number obtained will obviously not meet user needs.

[0054] In view of this, embodiments of the present disclosure provide an account processing method, medium, apparatus, and computing device. These methods utilize an element recognition model to identify other elements in a user's input account, such as pinyin elements, English word elements, or English name elements, thereby providing more elements for the account recommendation process, ensuring a richer selection of recommended accounts and enhancing the user experience.

[0055] After introducing the basic principles of the present disclosure, various non-limiting embodiments of the present disclosure are described in detail below.

[0056] Application Scenario Overview

[0057] First reference Figure 1 , Figure 1 This is a schematic diagram of an application scenario provided by an embodiment of the present disclosure. Figure 1As shown, in the application scenario provided by the embodiment of the present disclosure, the devices involved include a terminal device 101.

[0058] In the feature recognition scenario, the user can enter the account registration page of the platform to be registered through the terminal device, and enter the account to be registered on the account registration page to obtain the target feature in the account to be registered.

[0059] On the one hand, the platform may be an offline platform, for example, a stand-alone application that can complete registration and login without an Internet connection. In this scenario, the terminal device 101 may determine the target elements in the account to be registered.

[0060] On the other hand, if the platform is an Internet platform, the scenario may further include: a server 102. The server 102 communicates with the terminal device 101 via a network. When the user enters the account to be registered on the account registration page, the terminal device 102 sends the account to be registered to the server 102 via the network. The server 102 determines the target element in the account to be registered based on the received account to be registered.

[0061] Among them, the terminal device can be a personal digital assistant (PDA) device, a handheld device with wireless communication function (such as a smart phone, a tablet computer), a computing device (such as a personal computer (PC)), a vehicle-mounted device, a wearable device (such as a smart watch, a smart bracelet), a smart home device (such as a smart display device), etc.

[0062] The server may be a single server, a server cluster, a distributed server, a centralized server, or a cloud server, etc., which is not limited in the embodiments of the present disclosure.

[0063] Exemplary Methods

[0064] The following combination Figure 1 For application scenarios, refer to Figures 2 to 7 The following describes an account processing method provided according to an exemplary embodiment of the present disclosure. It should be noted that the above scenarios are merely provided to facilitate understanding of the spirit and principles of the present disclosure, and the embodiments of the present disclosure are not limited in this respect. Rather, the embodiments of the present disclosure can be applied to any applicable scenario.

[0065] The execution device applicable to the embodiments of the present disclosure is a computing device, such as the terminal or server mentioned above. Figure 2 Schematic diagram of the account processing method provided in the embodiment of the present disclosure Figure 1 .like Figure 2As shown, the account processing method provided by the embodiment of the present disclosure may include the following steps:

[0066] S201. Obtain the account to be registered.

[0067] The account to be registered includes multiple characters.

[0068] S202: Identify the character information of each character through the element recognition model, and determine the target element in the account to be registered based on the character information corresponding to the multiple characters.

[0069] Among them, the target element includes at least one of a pinyin element, an English word element or an English name element, and the element recognition model is trained based on the character information of each character in the registered account. The character information is used to indicate the position of each character in the element and / or the element type of the element.

[0070] For example, taking the account to be registered as "hellozhangsanlucy", which contains the characters "h, e, l, l, o, z, h, a, n, g, s, a, n, l, u, c, y", the element recognition model can identify that the character "h" is the first letter of the English word element; the characters "e, l, l" are the middle characters of the English word element; and the character "o" is the last character of the English word element. Based on this, it can be concluded that the account to be registered includes the English word element "hello". It should be understood that the principles for determining the elements "zhangsan" and "lucy" are similar to those described above and will not be listed here one by one.

[0071] In the account processing method provided by the embodiment of the present disclosure, a richer variety of element types can be identified through the element recognition model, and useful information in the account to be registered can be fully mined, thereby providing more elements for the account recommendation process to ensure the richness of the recommended accounts and enhance the user experience.

[0072] Next, combine Figure 3 and Figure 4 The feature recognition principle and training process of the feature recognition model are described in detail:

[0073] Figure 3 This is a schematic diagram of the structure of the element recognition model provided by the embodiment of the present disclosure. Figure 3 As shown, the feature recognition model provided by the embodiment of the present disclosure includes: an encoding layer, a forward long short-term memory network LSTM, a reverse LSTM and a conditional random field CRF;

[0074] Among them, the encoding layer is used to obtain the vector representation of each character in the account to be registered and generate the string vector of the account to be registered; the forward LSTM is used to obtain the first vector matrix corresponding to the string vector; the backward LSTM is used to obtain the second vector matrix corresponding to the string vector; the CRF is used to perform element recognition on the first vector matrix and the second vector matrix to obtain the character information of each character, and then obtain the target element in the account to be registered based on the character information.

[0075] It should be noted that the embodiment of the present disclosure does not limit the type of the coding layer, for example, it can be an embedding layer (Embedding) and the like.

[0076] During the feature recognition process, the account to be registered is input into the feature recognition model. First, the encoding layer encodes each character in the account to be registered into a vector representation. For example, taking the dimension of the encoding layer as 100, if the input account to be registered is "cutelucy-163", which contains 12 characters, then after processing by the encoding layer, a vector matrix with a dimension of [12,100] is obtained. Each vector in the vector matrix corresponds to the vector representation of a character in the account to be registered.

[0077] Furthermore, the forward LSTM scans the character string vector of the account to be registered (i.e., the [12, 100] matrix output by the encoding layer) from left to right to obtain a first vector matrix. For example, assuming that the dimension parameter of the forward LSTM is 200, in this step, the dimension of the first vector matrix obtained by the forward LSTM is [12, 200].

[0078] At the same time, the backward LSTM scans the string vector of the account to be registered from right to left (i.e., the [12, 100] matrix output by the encoding layer) to obtain a second vector matrix. For example, if the dimension parameter of the backward LSTM is 200, in this step, the dimension of the second vector matrix obtained by the backward LSTM is also [12, 200].

[0079] Furthermore, the first vector matrix obtained by the forward LSTM and the second vector matrix obtained by the backward LSTM are concatenated to obtain a vector matrix with a dimension of [12, 400].

[0080] Finally, the CRF layer recognizes the vector matrix of dimension [12, 400] to obtain the entity recognition result of the final string sequence. The CRF layer determines the character information of each character based on the vector matrix. The character information is used to indicate the position of each character in the element and / or the element type of the element.

[0081] Exemplarily, after passing through the CRF layer, its output is a vector of [12, 7], where 7 represents 7 types of character information: BW, IW, BP, IP, BN, IN, and O. BW represents the character at the first letter position in an English word element, and IW represents the character at the middle letter position or the end position in an English word element; BP represents the character at the first letter position in a pinyin element, and IP represents the character at the middle letter position or the end position in a pinyin element; BN represents the character at the first letter position in an English name element, and IN represents the character at the middle letter position or the end position in an English name element; O represents a character category other than the above characters, such as letters, numbers, symbols, etc.

[0082] The CRF layer will give the probability distribution of these 12 characters corresponding to these 7 categories of character information respectively, and the character information category with the highest probability is the character information of this character.

[0083] Please continue to refer to Figure 3 ,like Figure 3 As shown, taking the English word element "cute" as an example, the character information category of the character "c" in the element is BW, which means that the character "c" is the character at the first letter position in the English word; the character information category of the characters "u, t" is IW, that is, the characters "u, t" are the characters at the middle letter position in the English word; the character information category of the character "e" is IW, that is, the character "e" is the character at the last letter position in the English word.

[0084] Furthermore, after obtaining the character information of each character, it can be determined that the elements corresponding to these characters are the English word elements "cute".

[0085] It should be understood that the character information determination method and element determination method of other characters are similar and will not be described in detail here.

[0086] In the disclosed embodiments, the CRF layer ensures that string classification results conform to a priori rules. For example, if the results of two LSTM layers were directly output, two consecutive strings might be recognized as BW and IP, respectively. This model clearly does not meet the requirements for element composition: BW should be followed by IP, not IP. The CRF layer, however, calculates the correct transition results for each classification, avoiding this situation and improving the accuracy of the element recognition model.

[0087] Next, combine Figure 4 Explain the training process of the feature recognition model:

[0088] Figure 4 This is a flow chart of the element recognition model training method provided by the embodiment of the present disclosure. Figure 4As shown, the element recognition model training method of the embodiment of the present disclosure includes the following steps:

[0089] S401. Obtain multiple registered accounts.

[0090] The registered account may be a registered account of the same platform or a registered account of another platform, which is not limited in the embodiment of the present disclosure.

[0091] S402: Identify the elements included in the registered account and determine the element type of each element.

[0092] Among them, the element types include pinyin, English words and English names.

[0093] S403: Determine position information of each character according to the position of each character in the element.

[0094] S404: Acquire character information of each character in the element according to the element type and position information of each character in the element.

[0095] For example, taking the registered accounts "gzhangsan123", "stanleylsx", "cutelucy-163", and "gdfgpyfsz" as examples, each character in each registered account can be annotated according to the position of each character in the element, thereby obtaining the character information of each character. In this example, the annotated results of the above registered accounts are shown in the following table:

[0096] 1) Registered account "gzhangsan123":

[0097] character g z z h a n g s a n 1 2 3 Location information O O BP IP IP IP IP BP IP IP O O O

[0098] 2) Registered account "stanleylsx":

[0099] character s t a n l e y l s y Location information BN IN IN IN IN IN IN O O O

[0100] 3) Registered account "cutelucy-163":

[0101] character c u t e l u c y - 1 6 3 Location information BW IW IW IW BN IN IN IN O O O O

[0102] 4) Registered account "gdfgpyfsz"

[0103] character g d f g p f s z Location information O O O O O O O O

[0104] S405 . Train the initial element recognition model using the character information of each character in the registered account to obtain an element recognition model.

[0105] Furthermore, the character information of each character in the registered account is input into the initial element recognition model, and the initial element recognition model is trained with the corresponding character information of these accounts to obtain the element recognition model. The model training process is not described here in detail.

[0106] In some scenarios, you can also recommend accounts to users based on the elements in the account to be registered. Figure 5 and Figure 6 The illustrated embodiment provides a detailed description of the account recommendation process provided by the embodiment of the present disclosure:

[0107] Figure 5 Schematic diagram of the account processing method provided in the embodiment of the present disclosure Figure 2 It should be understood that the execution device applicable to the embodiments of the present disclosure is a computing device, such as the terminal or server mentioned above. Figure 5 As shown, the account recommendation method provided by the embodiment of the present disclosure includes the following steps:

[0108] S501. Obtain the account to be registered.

[0109] The account to be registered includes multiple characters.

[0110] S502: Identify the character information of each character through the element recognition model, and determine the target element in the account to be registered based on the character information corresponding to the multiple characters.

[0111] The principles and beneficial effects of steps S501 to S502 are similar to those of Figure 2 For details of steps S201 to S202 in the illustrated embodiment, reference may be made to the above embodiment and will not be repeated here.

[0112] S503: Generate a target account set based on target elements and preset generation rules.

[0113] Among them, the preset generation rules include: a first generation rule and / or a second generation rule; the first generation rule is used to adjust at least part of the characters in the target element in the account to be registered; the second generation rule is used to adjust the position between multiple target elements in the account to be registered to generate a target account set.

[0114] (1) With respect to the first generation rule, the first generation rule includes at least one of a deletion rule, a character position adjustment rule, and a replacement rule. For different element types, the corresponding adjustment methods are different.

[0115] Taking the account to be registered "hellozhangsanlucy" as an example, it contains the English word element "hello", the pinyin element "zhangsan" and the English name element "lucy".

[0116] In one example, for the pinyin elements, at least some characters of the pinyin elements in the account to be registered may be deleted according to a deletion rule to generate a target account, thereby obtaining a target account set.

[0117] It should be noted that the embodiments of the present disclosure do not specifically limit the deletion of at least some of the characters. For example, one of the two pinyin characters "zhangsan" can be deleted, and the resulting target account can be "hellosanlucy" or "hellozhanglucy"; alternatively, the other characters except the first letter of the two pinyin characters "zhangsan" can be deleted, and the resulting target account can be "hellozslucy", "hellozsanlucy", "hellozhangslucy", etc., and these target accounts can then constitute the target account set.

[0118] In another example, for the pinyin elements, the positions of at least some characters of the pinyin elements in the account to be registered may be adjusted according to character position adjustment rules to generate target accounts, thereby obtaining a target account set.

[0119] It should be noted that the embodiment of the present disclosure does not specifically limit the adjustment of at least part of the characters. For example, the positions of the two pinyins of "zhangsan" can be swapped, and the resulting target account can be: "hellosanzhanglucy".

[0120] In other examples, for English word elements, the English word elements in the account to be registered can be replaced according to the replacement rules to generate a target account, thereby obtaining a target account set.

[0121] Specifically, the English word element may be replaced by a synonym of the English word element. For example, "hello" may be replaced by "hi", and the target account obtained may be "hisanzhanglucy".

[0122] (2) With respect to the second generation rule, the target accounts can be generated by adjusting the positions of multiple target elements in the account to be registered, thereby obtaining a target account set.

[0123] Specifically, the position of any element in the pinyin element, English word element or English name element in the account to be registered can be adjusted to obtain the target account. Taking the above example, the target account obtained can be "zhangsanhellolucy", "lucyzhangsanhello", "zhangsanhellolucy", etc., which will not be repeated here.

[0124] It should be noted that for the same account to be registered, the target element can be processed by any one of the first generation rule and the second generation rule to obtain the target account, or the target element can be processed by both rules at the same time to obtain more diverse target accounts.

[0125] For example, still taking the account to be registered "hellozhangsanlucy" as an example, first delete the other characters except the first letters of the two pinyins of "zhangsan" through the first generation rule, and then swap the positions of the pinyin elements and the English name elements through the second generation rule. The target account that can be obtained is "hellolucyzs". As for other generation method combinations, they will not be repeated here.

[0126] S504: Recommend target accounts to the user based on the target account set.

[0127] Specifically, the target account set can be displayed on the account registration interface to achieve target account recommendation. Among them, all target accounts in the target account set can be displayed, or part of the target accounts in the target account set can be displayed, which is not limited in the embodiment of the present disclosure.

[0128] In the disclosed embodiment, by obtaining the target elements in the account to be registered, the available information in the account to be registered can be fully mined, and then account recommendations can be made based on the target elements. Compared with the existing account recommendation method, the target accounts obtained through this solution are richer and more in line with user needs.

[0129] It should be noted that in the embodiments of the present disclosure, after a user enters the account to be registered, target accounts can be directly recommended to the user based on the account to be registered, thereby providing a reference for the user, increasing account diversity, and improving the user experience. That is, after the user enters the account to be registered, steps S502 to S504 can be directly executed, and the target accounts can be displayed on the terminal device for the user to select.

[0130] In some embodiments, before step S502, it is also possible to first determine whether the account to be registered is a duplicate of a registered account on the platform. If the account to be registered is a duplicate of a registered account, steps S502 to S504 are executed to recommend a target account to the user; if the account to be registered is not a duplicate of a registered account, it is determined that the account to be registered is an available account on the target platform.

[0131] Optionally, when it is determined that the account to be registered is the same as the registered account, a prompt message may be displayed on the terminal device to prompt the user to change the currently entered account to be registered; alternatively, a recommended target account may be directly displayed on the terminal device.

[0132] In some embodiments, the account to be registered may contain other types of elements, such as symbolic characters and / or numeric characters, etc. During the account recommendation process, the target account may also be obtained based on these elements.

[0133] For example, let's take the account number "hellozhangsanlucy-123" to be registered, which contains the numeric elements "1, 2, 3" and the symbol element "-". During the target account recommendation process, the positions of the numeric elements, symbol elements, English name elements, English word elements, and pinyin elements can be swapped according to the second pre-generated rule to generate more target accounts and further improve the diversity of target accounts.

[0134] In practical applications, on the one hand, the above-mentioned element recognition model can be used to extract the symbol characters and / or numeric characters in the account to be registered, wherein the element recognition model is obtained by training the symbol characters and numeric characters in the registered account. As for the training method and principle of the element recognition model, Figure 4 In the disclosed embodiment, the symbol characters and / or numeric characters in the account to be registered are extracted by the element recognition model, which can ensure the accuracy of the extraction result and thus ensure the rationality of the target account for account recommendation.

[0135] Alternatively, you can use regular expressions to extract symbols and numbers from the account to be registered. Compared to extracting symbols and numbers using feature recognition models, using regular expressions simplifies the process by eliminating the need to train feature recognition models based solely on symbols and numbers, which helps improve feature extraction efficiency.

[0136] Optionally, for an account to be registered, the symbol characters and / or numeric characters in the account to be registered can be first extracted through a regular expression, and the characters corresponding to the symbol characters and / or numeric characters in the account to be registered can be excluded, and then the pinyin elements, English word elements or English name elements in the remaining characters in the account to be registered can be identified through the element recognition model. For example, for the account to be registered "hellozhangsanlucy-123", the numeric elements "1, 2, 3" and the symbol element "-" can be first extracted through a regular expression, and then the pinyin elements, English word elements or English name elements in "hellozhangsanlucy" can be extracted through the element recognition model. This solution can reduce the length of the character string input into the element recognition model, and there is no need to recognize the characters corresponding to the symbol characters and / or numeric characters. At the same time, it can also exclude the influence of the characters corresponding to the symbol characters and / or numeric characters on the target elements, which can further improve the extraction efficiency and accuracy of the target elements.

[0137] Figure 6 This is a flow chart of recommending target accounts to users according to an embodiment of the present disclosure. Figure 5 Based on the illustrated embodiment, the target account recommendation process in step S504 is described in more detail. Specifically, the account processing method of the embodiment of the present disclosure includes the following steps:

[0138] S601: Determine the first target account in the target account set that is the same as the registered account.

[0139] S602: Delete the first target account in the target account set.

[0140] In actual applications, after obtaining the target account set, the target account in the target account set may be compared with the registered accounts of the platform to determine the registered account in the target account set (ie, the first target account).

[0141] As an optional solution, in the process of generating target accounts using preset generation rules, after each target account is generated, the account is compared to see if it is a duplicate of the registered account on the platform. If it is a duplicate, the target account is directly deleted and the next target account is generated.

[0142] This solution can ensure the availability of recommended accounts and avoid recommending registered accounts to users, thereby improving user experience.

[0143] S603 : Determine a priority score for each target account in the target account set based on the string adjustment distance of each target account relative to the account to be registered, the first string length of the target account, and the second string length of the account to be registered.

[0144] The string adjustment distance indicates the similarity between the target account and the account to be registered. For the same account to be registered, if fewer strings are adjusted (i.e., the adjustments are shorter), the similarity with the target account is relatively high, and it is closer to the user's expectations. Furthermore, the string length of the target account can be used as a measure of account value: the shorter the string length, the higher the value.

[0145] Therefore, in this step, the above two metrics can be combined to determine the priority of the recommended accounts, so as to recommend high-quality target accounts to users based on the priority.

[0146] Different adjustment methods for registered accounts correspond to different string adjustment distances. The string adjustment distance for each target account can be determined based on the correspondence between the adjustment method and the string adjustment distance.

[0147] Exemplarily, in a deletion operation, the value of the distance dist of the string corresponding to each character is adjusted to n1; in an insertion operation, the value of the distance dist of the string corresponding to each character is adjusted to n2; in a replacement operation, the value of the distance dist of the string corresponding to each character is adjusted to n3.

[0148] For example, taking the string "abc" as an example, the string adjustment distance dist of the adjusted string "abc" is 0, the string adjustment distance dist of the adjusted string "ab" is n1, the string adjustment distance dist of the string "abcd" is n2, and the string adjustment distance dist of the adjusted string "abd" is n3.

[0149] It should be understood that when the same target account adopts multiple of the above rules, the string adjustment distance is the sum of the string adjustment distances of each operation adopted.

[0150] Next, we will explain this solution in detail with specific steps:

[0151] (1) Determine the distance influence score of the target account based on the character string adjustment distance corresponding to each target account, the length of the first character string of the target account, and the length of the second character string of the account to be registered.

[0152] Specifically, the distance influence score L1 can be obtained according to the following formula:

[0153]

[0154] Where len1 is the sum of the length of the first string of the target account and the length of the second string of the account to be registered, and dist is the string adjustment distance corresponding to the target account.

[0155] Taking the values ​​of n1 and n2 as 1 and the value of n3 as 2, in the above example, the distance influence score L1 of the string "abc" is 1, the distance influence score L1 of the string "abcd" is 0.8, and the distance influence score L1 of the string "abd" is 0.666.

[0156] (2) Determine the length influence score of the target account based on the length of the first character string of the target account, the length of the shortest character string of the target account in the target account set, and the length of the longest character string of the target account in the target account set.

[0157] The target account's length impact score L2 is determined by the min-max normalized value of the string lengths of all target accounts in the target account set. Assuming that the shortest length of a target account in this set is minlen and the longest length is maxlen, for a target account with a length of len2, the target account's length impact score L2 can be calculated using the following formula:

[0158]

[0159] (3) Determine the priority score of each target account in the target account set based on the distance influence score, length influence score and preset adjustment coefficient of each target account.

[0160] Specifically, the priority score of each target can be obtained according to the following formula:

[0161] score=α*L1+(1-α)*L2

[0162] Wherein, α is a preset adjustment coefficient, and α is a positive number less than or equal to 1.

[0163] It is worth noting that α is used to indicate the weight of the distance influence score L1 when calculating the priority score. The closer the α value is to 1, the more the recommended account pays attention to the distance influence score, and the more similar the recommended account is to the account entered by the user. Correspondingly, (1-α) is used to indicate the weight of the length influence score L2 when calculating the priority score. The closer the α value is to 0, the greater the weight of the length influence score L2, which means that the recommended account pays more attention to the length influence score, and the length of the recommended account is shorter.

[0164] In view of this, in the embodiments of the present disclosure, the value of α can be adjusted to achieve personalized account recommendations to users. For example, if it is necessary to recommend accounts that are highly similar to the user-entered account, the value of α can be appropriately increased. If it is necessary to recommend accounts with shorter lengths, the value of α can be appropriately decreased. The specific value of α can be set according to needs and is not limited in the embodiments of the present disclosure.

[0165] S604: Output a recommendation list corresponding to the target account set according to the priority score of each target account in the target account set.

[0166] In this step, after obtaining the priority score of each target account, the target accounts in the target account set can be sorted according to the priority score, so as to recommend target accounts with higher priority to the user. The embodiment of this disclosure does not limit the number of target accounts in the recommendation list.

[0167] In the disclosed embodiment, by using the character string adjustment distance and the character string length of the target account as the measurement criteria for recommending an account, the similarity between the recommended account and the account to be registered can be fully considered while ensuring the account value of the recommended account, thereby recommending high-quality target accounts to users and improving user experience.

[0168] In actual applications, malicious registrations occur on many platforms, which brings certain risks to the platforms and causes losses to the platforms.

[0169] The inventors found that these maliciously registered accounts are highly random. For example, the account to be registered "gdfgpyfsz" is composed of random letters. It is very likely that it is not an account registered by a natural person. The probability of this account being repeated with a registered account is also relatively small. It is very likely that it will be considered as an unregistered account and registered. In view of this, in the embodiment of the present disclosure, after identifying the target element, it is also possible to determine whether the current account to be registered is a risky account based on the target element to prevent malicious registration, thereby reducing the risk of the platform. Next, combined with Figure 7 Describe this plan in detail.

[0170] Figure 7 Schematic diagram of the account processing method provided in the embodiment of the present disclosure Figure 3 The embodiments of the present disclosure will be Figure 2 A more detailed description will be given based on the illustrated embodiment. Specifically, the account processing method of the embodiment of the present disclosure includes the following steps:

[0171] S701. Obtain the account to be registered.

[0172] The account to be registered includes multiple characters.

[0173] S702: Identify the character information of each character through the element recognition model, and determine the target element in the account to be registered based on the character information corresponding to the multiple characters.

[0174] Among them, steps S701 to S702 are Figure 2 The principles and beneficial effects of steps S201 to S202 in the illustrated embodiment are similar, and reference may be made to the above embodiment for details, which will not be described in detail here.

[0175] S703: Determine a randomness score corresponding to the account to be registered based on the total length of the target elements in the account to be registered and the length of the account to be registered.

[0176] Specifically, the randomness score corresponding to the account to be registered is the ratio of the total length of the target element to the length of the account to be registered. For example, if the account to be registered is "hellozhangsanlucy", which contains the English word element "hello", the pinyin element "zhangsan" and the English name element "lucy", the total length of the target element is 17, and the total length of the account to be registered is also 17, then the randomness score of the account to be registered is 1.

[0177] For the account to be registered "gdfgpyfsz", the total length of its target element is 0 (if the target element is not extracted, the total length of the target element is 0), and the total length of the account to be registered is 9, so the randomness score of the account to be registered is 0.

[0178] It should be understood that the method for determining the randomness scores of other accounts to be registered is similar and will not be described in detail here.

[0179] S704: Determine whether the randomness score is less than a preset value.

[0180] S705: If the randomness score is less than a preset value, the account to be registered is determined to be a risky account.

[0181] S706: If the randomness score is greater than or equal to the preset value, determine that the account to be registered is an available account on the target platform.

[0182] It should be noted that the embodiment of the present disclosure does not specifically limit the size of the preset value. For example, it can be 0.5. That is, when the randomness score of the account to be registered is greater than 0.5, it means that the account is an available account on the platform. Otherwise, a prompt message is output to prompt the user to re-enter.

[0183] In some scenarios, if the account to be registered is determined to be available, it may be further determined whether the account to be registered duplicates an account already registered on the platform. If so, a target account may be recommended to the user based on the target factors, as described in the aforementioned embodiment. It should be understood that the target account recommendation process is similar to that described in the aforementioned embodiment and will not be further described here.

[0184] In the disclosed embodiment, by determining whether the current account to be registered is a risky account through target elements, malicious registration can be prevented, thereby reducing platform risks.

[0185] Exemplary media

[0186] After introducing the method of the exemplary embodiment of the present disclosure, next, reference is made to Figure 9 A storage medium according to an exemplary embodiment of the present disclosure is described.

[0187] Figure 8 Schematic diagram of the program product provided by the embodiment of the present disclosure. Figure 8 As shown, a program product 800 for implementing the above method according to an embodiment of the present disclosure is described, which can be a portable compact disk read-only memory (CD-ROM) and includes program code and can be run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited thereto.

[0188] The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or component, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0189] A readable signal medium may include a data signal transmitted in baseband or as part of a carrier wave, which carries readable program code. Such a transmitted data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable signal medium may also be any readable medium other than a readable storage medium.

[0190] Program code for performing the operations disclosed herein can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device via any type of network, including a local area network (LAN) or a wide area network (WAN).

[0191] Exemplary devices

[0192] After introducing the medium of the exemplary embodiment of the present disclosure, next, reference is made to Figure 9 and Figure 10 A device according to an exemplary embodiment of the present disclosure will be described.

[0193] Figure 9 A schematic diagram of the structure of the account processing device provided in the embodiment of the present disclosure Figure 1 .like Figure 9As shown, the account processing device 900 of the embodiment of the present disclosure includes: an acquisition module 901, used to obtain the account to be registered, which includes multiple characters; an identification module 902, used to identify the character information of each character through an element recognition model, and determine the target element in the account to be registered based on the character information corresponding to the multiple characters, the target element includes at least one of a pinyin element, an English word element or an English name element, the element recognition model is trained based on the character information of each character in the registered account, and the character information is used to indicate the position of each character in the element and / or the element type of the element.

[0194] In one embodiment of the present disclosure, the feature recognition model includes: an encoding layer, a forward long short-term memory network LSTM, a reverse LSTM and a conditional random field CRF; the encoding layer is used to obtain the vector representation of each character in the account to be registered and generate a string vector of the account to be registered; the forward LSTM is used to obtain a first vector matrix corresponding to the string vector; the reverse LSTM is used to obtain a second vector matrix corresponding to the string vector; the CRF is used to perform feature recognition on the first vector matrix and the second vector matrix, obtain character information of each character, and obtain the target feature in the account to be registered based on the character information.

[0195] In one embodiment of the present disclosure, an element recognition model is obtained by: obtaining multiple registered accounts; identifying elements contained in the registered accounts, and determining the element type of each element, which includes pinyin, English words, and English names; determining the position information of each character in the element based on the position of each character in the element; obtaining the character information of each character in the element based on the element type of the element and the position information of each character in the element; training the initial element recognition model through the character information of each character in the registered account to obtain the element recognition model.

[0196] Figure 10 A schematic diagram of the structure of the account processing device provided in the embodiment of the present disclosure Figure 2 .like Figure 10 As shown, the account processing device 900 of the embodiment of the present disclosure further includes: a generation module 903, which is used to generate a target account set according to target elements and preset generation rules, and the target account set includes at least one target account.

[0197] In one embodiment of the present disclosure, the preset generation rules include: a first generation rule and / or a second generation rule; the generation module is specifically used to: adjust at least part of the characters in the target elements in the account to be registered according to the first generation rule to generate a target account set; and / or, adjust the positions between multiple target elements in the account to be registered according to the second generation rule to generate a target account set.

[0198] In one embodiment of the present disclosure, the first generation rule includes at least one of a deletion rule, a character position adjustment rule and a replacement rule; the generation module specifically generates the target account set in at least one of the following ways: according to the deletion rule, deleting at least part of the characters of the pinyin elements in the account to be registered to generate the target account set; according to the character position adjustment rule, adjusting the positions between at least part of the characters of the pinyin elements in the account to be registered to generate the target account set; according to the replacement rule, replacing the English word elements in the account to be registered to generate the target account set.

[0199] In one embodiment of the present disclosure, the account processing device also includes: a first determination module 904, which is used to determine the priority score of each target account in the target account set based on the string adjustment distance of each target account relative to the account to be registered, the first string length of the target account, and the second string length of the account to be registered; a recommendation module 905, which is used to output a recommendation list corresponding to the target account set based on the priority score of each target account in the target account set.

[0200] In one embodiment of the present disclosure, the first determination module 906 is specifically used to: determine the adjustment distance influence score of the target account based on the string adjustment distance corresponding to each target account, the first string length of the target account, and the second string length of the account to be registered; determine the length influence score of the target account based on the first string length of the target account, the string length of the shortest target account in the target account set, and the string length of the longest target account in the target account set; determine the priority score of each target account in the target account set based on the distance influence score, length influence score and preset adjustment coefficient of each target account.

[0201] In one embodiment of the present disclosure, the account processing device 900 further includes: a second determining module 906 for determining a first target account in the target account set that is duplicated with a registered account; and a deleting module 907 for deleting the first target account in the target account set.

[0202] In one embodiment of the present disclosure, the account processing device 900 also includes: a third determination module 908, which is used to determine whether the account to be registered is a duplicate of a registered account on the target platform; if the account to be registered is not a duplicate of the registered account, then the account to be registered is determined to be an available account on the target platform; if the account to be registered is a duplicate of the registered account, then the character information of each character is identified through the element recognition model, and the target element in the account to be registered is determined based on the character information corresponding to multiple characters.

[0203] In one embodiment of the present disclosure, the account processing device 900 also includes: a fourth determination module 909, which is used to determine the randomness score corresponding to the account to be registered based on the total length of the target elements in the account to be registered and the length of the account to be registered; if the randomness score is less than a preset value, the account to be registered is determined to be a risk account; if the randomness score is greater than or equal to the preset value, the account to be registered is determined to be an available account of the target platform.

[0204] In one embodiment of the present disclosure, the account processing device 900 further includes: an extraction module 910, configured to extract symbolic characters and numeric characters in the account to be registered by using regular expressions.

[0205] The above-mentioned account processing device provided by the embodiment of the present disclosure can implement each process in the above-mentioned deadline time acquisition method embodiment and achieve the same functions and effects, which will not be described in detail here.

[0206] Exemplary Computing Devices

[0207] After introducing the method, medium and apparatus of the exemplary embodiments of the present disclosure, Figure 11 The computing device of the exemplary embodiment of the present disclosure is described. It should be understood that Figure 11 The computing device 1100 shown is only an example and should not limit the functionality and scope of use of the embodiments of the present disclosure.

[0208] Figure 11 This is a schematic diagram of the structure of the computing device provided in the embodiment of the present disclosure. Figure 11 As shown, computing device 1100 is implemented as a general-purpose computing device. Components of computing device 1100 may include, but are not limited to, the aforementioned at least one processing unit 1101, the aforementioned at least one storage unit 1102, and a bus 1103 connecting various system components (including processing unit 1101 and storage unit 1102).

[0209] The bus 1103 includes a data bus, a control bus, and an address bus. The storage unit 1102 may include a readable medium in the form of a volatile memory, such as a random access memory (RAM) 1111 and / or a cache memory 1122, and may further include a readable medium in the form of a non-volatile memory, such as a read-only memory (ROM) 1132.

[0210] The storage unit 1102 may also include a program / utility 1152 having a set (at least one) of program modules 1142, such program modules 1142 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0211] The computing device 1100 may also communicate with one or more external devices 1104 (e.g., a keyboard, a pointing device, etc.). Such communication may be performed via an input / output (I / O) interface 1105. Furthermore, the computing device 1100 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 1106. Figure 11 As shown, network adapter 1106 communicates with other modules of computing device 1100 via bus 1103. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with computing device 1100, including but not limited to microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0212] It should be noted that although several units / modules or sub-units / modules of the timing update device are mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more units / modules described above can be embodied in a single unit / module. Conversely, the features and functions of a single unit / module described above can be further divided and embodied by multiple units / modules.

[0213] Furthermore, although the operations of the disclosed method are described in a particular order in the accompanying drawings, this does not require or imply that the operations must be performed in this particular order, or that all illustrated operations must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps.

[0214] Although the spirit and principles of the present disclosure have been described with reference to several specific embodiments, it should be understood that the present disclosure is not limited to the specific embodiments disclosed, and the division into various aspects does not mean that the features in these aspects cannot be combined to benefit. Such division is only for the convenience of expression. The present disclosure is intended to cover various modifications and equivalent arrangements included in the spirit and scope of the appended claims.

Claims

1. An account processing method, comprising: Obtaining an account to be registered, where the account to be registered includes multiple characters; Identifying character information of each character using an element recognition model, and determining a target element in the account to be registered based on the character information corresponding to the multiple characters, the target element including at least one of a pinyin element, an English word element, or an English name element, the element recognition model being trained based on the character information of each character in the registered account, the character information being used to indicate the position of each character in the element and / or the element type of the element; generating a target account set according to the target elements and preset generation rules, wherein the target account set includes at least one target account; determining an adjustment distance influence score of each target account based on a character string adjustment distance of each target account relative to the account to be registered, a first character string length of the target account, and a second character string length of the account to be registered; determining a length influence score of the target account based on the first character string length of the target account, the shortest character string length of the target account in the target account set, and the longest character string length of the target account in the target account set; Determining a priority score for each target account in the target account set based on the distance influence score, the length influence score, and a preset adjustment coefficient of each target account; Output a recommendation list corresponding to the target account set according to the priority score of each target account in the target account set.

2. The account processing method according to claim 1, wherein the element recognition model comprises: Encoding layer, forward long short-term memory network LSTM, reverse LSTM and conditional random field CRF; The encoding layer is used to obtain a vector representation of each character in the account to be registered and generate a character string vector of the account to be registered; The forward LSTM is used to obtain a first vector matrix corresponding to the string vector; The reverse LSTM is used to obtain a second vector matrix corresponding to the string vector; The CRF is used to perform element recognition on the first vector matrix and the second vector matrix, obtain character information of each character, and obtain a target element in the account to be registered based on the character information.

3. The account processing method according to claim 1, wherein the element recognition model is obtained by: Get multiple registered accounts; Identify elements included in the registered account and determine the element type of each element, wherein the element type includes pinyin, English words, and English names; Determining position information of each character in the element according to the position of each character in the element; acquiring character information of each character in the element according to the element type of the element and the position information of each character in the element; The initial element recognition model is trained using the character information of each character in the registered account to obtain the element recognition model.

4. The account processing method according to claim 1, wherein the preset generation rule comprises: first generation rule and / or second generation rule; Generating a target account set according to the target elements and preset generation rules includes: adjusting at least some characters in the target elements in the account to be registered according to the first generation rule to generate the target account set; And / or, according to the second generation rule, positions between multiple target elements in the account to be registered are adjusted to generate the target account set.

5. The account processing method according to claim 4, wherein the first generation rule comprises at least one of a deletion rule, a character position adjustment rule, and a replacement rule; The step of adjusting at least some characters in the target elements in the to-be-registered account according to the first generation rule to generate the target account set includes at least one of the following methods: Deleting at least some of the pinyin elements in the account to be registered according to the deletion rule to generate the target account set; Adjusting the positions of at least some characters of the pinyin elements in the to-be-registered account according to the character position adjustment rule to generate the target account set; According to the replacement rule, the English word elements in the account to be registered are replaced to generate the target account set.

6. The account processing method according to claim 1, after generating the target account set according to the target elements and the preset generation rules, further comprising: Determine a first target account in the target account set that is a duplicate of a registered account; Delete the first target account in the target account set.

7. The account processing method according to any one of claims 1 to 5, further comprising: Determine whether the account to be registered is the same as an existing account on the target platform; If the account to be registered does not overlap with the registered account, determining that the account to be registered is an available account on the target platform; If the account to be registered overlaps with the registered account, the character information of each character is identified through the element recognition model, and the target element in the account to be registered is determined based on the character information corresponding to the multiple characters.

8. The account processing method according to any one of claims 1 to 5, further comprising: after identifying the character information of each character using the element recognition model and determining the target element in the account to be registered based on the character information corresponding to the multiple characters; Determining a randomness score corresponding to the account to be registered based on the total length of the target elements in the account to be registered and the length of the account to be registered; If the randomness score is less than a preset value, the account to be registered is determined to be a risky account; If the randomness score is greater than or equal to a preset value, the account to be registered is determined to be an available account on the target platform.

9. The account processing method according to any one of claims 1 to 5, further comprising: before identifying the character information of each character using the element recognition model and determining the target element in the account to be registered based on the character information corresponding to the multiple characters: Use regular expressions to extract symbol characters and numeric characters in the account to be registered.

10. A computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when a processor executes the computer-executable instructions, the account processing method according to any one of claims 1 to 9 is implemented.

11. An account processing device, comprising: An acquisition module is used to acquire an account to be registered, where the account to be registered includes multiple characters; an identification module, configured to identify character information of each character using an element recognition model, and determine a target element in the account to be registered based on the character information corresponding to the multiple characters, wherein the target element includes at least one of a pinyin element, an English word element, or an English name element, wherein the element recognition model is trained based on the character information of each character in the registered account, and the character information indicates the position of each character in the element and / or the element type of the element; a generating module, configured to generate a target account set according to the target elements and preset generating rules, wherein the target account set includes at least one target account; a first determining module, configured to determine an adjusted distance influence score of each target account based on a character string adjustment distance of each target account relative to the account to be registered, a first character string length of the target account, and a second character string length of the account to be registered; determining a length influence score of the target account based on the first character string length of the target account, the shortest character string length of the target account in the target account set, and the longest character string length of the target account in the target account set; Determining a priority score for each target account in the target account set based on the distance influence score, the length influence score, and a preset adjustment coefficient of each target account; The recommendation module is configured to output a recommendation list corresponding to the target account set according to the priority score of each target account in the target account set.

12. The account processing device according to claim 11, wherein the element recognition model comprises: Encoding layer, forward long short-term memory network LSTM, reverse LSTM and conditional random field CRF; The encoding layer is used to obtain a vector representation of each character in the account to be registered and generate a character string vector of the account to be registered; The forward LSTM is used to obtain a first vector matrix corresponding to the string vector; The reverse LSTM is used to obtain a second vector matrix corresponding to the string vector; The CRF is used to perform element recognition on the first vector matrix and the second vector matrix, obtain character information of each character, and obtain a target element in the account to be registered based on the character information.

13. The account processing device according to claim 11, wherein the element recognition model is obtained by: Get multiple registered accounts; Identify elements included in the registered account and determine the element type of each element, wherein the element type includes pinyin, English words, and English names; Determining position information of each character in the element according to the position of each character in the element; acquiring character information of each character in the element according to the element type of the element and the position information of each character in the element; The initial element recognition model is trained using the character information of each character in the registered account to obtain the element recognition model.

14. The account processing device according to claim 11, wherein the preset generation rule comprises: The first generation rule and / or the second generation rule; the generation module is specifically configured to: adjust at least some characters in the target element of the account to be registered according to the first generation rule to generate the target account set; And / or, according to the second generation rule, positions between multiple target elements in the account to be registered are adjusted to generate the target account set.

15. The account processing device according to claim 14, wherein the first generation rule comprises at least one of a deletion rule, a character position adjustment rule, and a replacement rule; The generating module generates the target account set in at least one of the following ways: Deleting at least some of the pinyin elements in the account to be registered according to the deletion rule to generate the target account set; Adjusting the positions of at least some characters of the pinyin elements in the to-be-registered account according to the character position adjustment rule to generate the target account set; According to the replacement rule, the English word elements in the account to be registered are replaced to generate the target account set.

16. The account processing device according to any one of claims 11, further comprising: A second determining module is configured to determine a first target account in the target account set that is a duplicate of a registered account; A deletion module is configured to delete the first target account from the target account set.

17. The account processing device according to any one of claims 11 to 15, further comprising: A third determination module is used to determine whether the account to be registered is the same as a registered account on the target platform; If the account to be registered does not overlap with the registered account, determining that the account to be registered is an available account on the target platform; If the account to be registered overlaps with the registered account, the character information of each character is identified through the element recognition model, and the target element in the account to be registered is determined based on the character information corresponding to the multiple characters.

18. The account processing device according to any one of claims 11 to 15, further comprising: a fourth determining module, configured to determine a randomness score corresponding to the account to be registered based on the total length of the target elements in the account to be registered and the length of the account to be registered; If the randomness score is less than a preset value, the account to be registered is determined to be a risky account; If the randomness score is greater than or equal to a preset value, the account to be registered is determined to be an available account on the target platform.

19. The account processing device according to any one of claims 11 to 15, further comprising: The extraction module is used to extract the symbol characters and numeric characters in the account to be registered through regular expressions.

20. A computing device comprising: at least one processor and memory; The memory stores computer-executable instructions; The at least one processor and at least the computer execution instructions stored in the memory enable the at least one processor to execute the account processing method according to any one of claims 1 to 9.

21. A computer program product, comprising a computer program; when the computer program is executed, the account processing method according to any one of claims 1 to 9 is implemented.

Images