Method for generating message signature

By constructing candidate keyword graphs and using linear layer functions to determine node weight values, and generating message signatures, the problems of low message signature efficiency and risk of identity abuse in the existing technology are solved, and efficient and accurate message signature generation is achieved.

CN119026181APending Publication Date: 2024-11-26CHINA MOBILE INTERNET CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410999981.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-24
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The prior art is difficult to efficiently generate message signatures that can effectively characterize the identity of the message sending end, which poses the risk of identity impersonation and fraud, and manual editing and review are time-consuming and inefficient.

Method used

By obtaining the identity information of the message sending end, a candidate keyword graph is constructed, a preset linear layer function and keyword extraction algorithm is used to determine the node weight value, and a target node with a large weight value is selected to generate a message signature to ensure the authenticity and accuracy of the signature.

Benefits of technology

It realizes efficient generation of message signatures that can accurately characterize the identity of the message sending end, reducing the risk of identity abuse and improving generation efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119026181A_ABST
    Figure CN119026181A_ABST
Patent Text Reader

Abstract

The embodiment of the invention discloses a method for generating a message signature, which is used for efficiently generating a message signature capable of representing the identity of a message sending end. The scheme comprises the following steps: acquiring identity information of a message sending end, wherein the identity information comprises a plurality of candidate keywords; constructing a candidate keyword graph based on the plurality of candidate keywords, wherein the candidate keyword graph comprises nodes and edges connecting the nodes; according to a preset linear layer function, the weight value of each node in the candidate keyword graph is determined through a keyword extraction algorithm, the preset linear layer function is used for determining the matching degree of the candidate keywords and positive sample keywords in a preset word bank, and the positive sample keywords are noun keywords with identity property characteristics; target nodes are selected from the candidate keyword graph according to the weight values of all the nodes, the message signature of the message sending end is generated through candidate keywords corresponding to the target nodes, and the weight value of the selected target nodes is larger than that of the unselected nodes.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of communications, and in particular to a method for generating a message signature. Background Art

[0002] In the field of communication technology, a message signature can indicate the source of a message and can be sent to the receiving end together with the message content. Some criminals may imitate the user's relatives or friends to send messages and defraud the user.

[0003] For a large number of message senders, if manual editing and review are used to generate message signatures, it will not only consume a lot of manpower and time, but also fail to ensure that the message signature can effectively represent the identity of the message sender, and there may be risks of fraud such as identity impersonation.

[0004] How to efficiently generate a message signature that can represent the identity of the message sender is a technical problem to be solved by this application. Summary of the invention

[0005] The purpose of the embodiment of the present application is to provide a method for generating a message signature, so as to efficiently generate a message signature that can represent the identity of the message sender.

[0006] In a first aspect, a method for generating a message signature is provided, comprising:

[0007] Acquire identity information of a message sender, wherein the identity information includes a plurality of candidate keywords;

[0008] constructing a candidate keyword graph based on the multiple candidate keywords, the candidate keyword graph comprising nodes and edges connecting the nodes, the nodes corresponding to the candidate keywords in the identity information, and the edges representing the co-occurrence relationship between the connected nodes;

[0009] Determine the weight value of each node in the candidate keyword graph through a keyword extraction algorithm according to a preset linear layer function, wherein the preset linear layer function is used to determine the matching degree between the candidate keyword and the positive sample keyword in the preset vocabulary, wherein the positive sample keyword is a noun keyword with identity characteristics;

[0010] At least one target node is selected from the candidate keyword graph according to the weight value of each node, and the message signature of the message sender is generated with the candidate keyword corresponding to the at least one target node, wherein the weight value of any selected target node in the candidate keyword graph is greater than the weight value of any unselected node.

[0011] In a second aspect, a device for generating a message signature is provided, comprising:

[0012] An acquisition module, for acquiring identity information of a message sender, wherein the identity information includes a plurality of candidate keywords;

[0013] A construction module, constructing a candidate keyword graph based on the multiple candidate keywords, the candidate keyword graph including nodes and edges connecting the nodes, the nodes corresponding to the candidate keywords in the identity information, and the edges representing the co-occurrence relationship between the connected nodes;

[0014] A determination module, which determines the weight value of each node in the candidate keyword graph through a keyword extraction algorithm according to a preset linear layer function, wherein the preset linear layer function is used to determine the matching degree between the candidate keyword and the positive sample keyword in the preset vocabulary, wherein the positive sample keyword is a noun keyword with identity characteristics;

[0015] A generation module selects at least one target node from the candidate keyword graph according to the weight value of each node, and generates a message signature of the message sender with the candidate keyword corresponding to the at least one target node, wherein the weight value of any selected target node in the candidate keyword graph is greater than the weight value of any unselected node.

[0016] According to a third aspect, an electronic device is provided. The electronic device comprises a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps of the method according to the first aspect are implemented.

[0017] According to a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method according to the first aspect are implemented.

[0018] In a fifth aspect, a computer program product is provided, the computer program product comprising a non-transitory computer-readable storage medium storing a computer program, the computer program being operable to cause a computer to execute part or all of the steps of the method of the first aspect.

[0019] In an embodiment of the present application, first, the identity information of the message sender is obtained, and the identity information includes multiple candidate keywords. Then, a candidate keyword graph is constructed based on the multiple candidate keywords, wherein the candidate keyword graph includes nodes and edges connecting the nodes, the nodes correspond to the candidate keywords in the identity information, and the edges represent the co-occurrence relationship between the connected nodes. Next, the weight values ​​of each node in the candidate keyword graph are determined by a keyword extraction algorithm according to a preset linear layer function, wherein the preset linear layer function is used to determine the degree of matching between the candidate keywords and the positive sample keywords in the preset vocabulary, and the positive sample keywords are nominal keywords with identity characteristics. Subsequently, at least one target node is selected from the candidate keyword graph according to the weight value of each node, and a message signature of the message sender is generated with the candidate keywords corresponding to at least one target node, wherein the weight value of any selected target node in the candidate keyword graph is greater than the weight value of any unselected node. Through the scheme provided in the embodiment of the present application, the weight values ​​of multiple candidate keywords in the identity information of the message sender can be determined by a keyword extraction algorithm according to a preset linear layer function, and the weight values ​​can represent the degree to which the candidate keywords have identity characteristics. Further, the candidate keywords corresponding to the target nodes with large weight values ​​are selected to generate the message signature of the sender. The message signature generated by this solution is generated by the keywords in the identity information of the message sender, which is authentic. The keywords of the target node selected by the weight value are keywords with relatively identity characteristics in the identity information, which can accurately express the identity of the message sender, thereby efficiently generating a message signature that can represent the identity of the message sender. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0021] Figure 1 This is one of the flowcharts of a method for generating a message signature in an embodiment of the present application;

[0022] Figure 2a This is a second flow chart of a method for generating a message signature according to an embodiment of the present application;

[0023] Figure 2b This is one of the interface schematic diagrams of a method for generating a message signature in an embodiment of the present application;

[0024] Figure 2c This is the second interface diagram of a method for generating a message signature in an embodiment of the present application;

[0025] Figure 2dThis is a third interface diagram of a method for generating a message signature in an embodiment of the present application;

[0026] Figure 3 This is a flowchart of a method for generating a message signature according to an embodiment of the present application;

[0027] Figure 4 This is a fourth flowchart of a method for generating a message signature according to an embodiment of the present application;

[0028] Figure 5 This is a flowchart of a method for generating a message signature according to an embodiment of the present application;

[0029] Figure 6a This is a flowchart of a method for generating a message signature according to an embodiment of the present application;

[0030] Figure 6b This is a schematic diagram of a system structure of a method for generating a message signature according to an embodiment of the present application;

[0031] Figure 6c is based on Figure 6b A flowchart of a method for generating a message signature according to an embodiment of the present application;

[0032] Figure 7 It is a schematic diagram of the structure of a device for generating a message signature according to an embodiment of the present application. DETAILED DESCRIPTION

[0033] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application. The numbering of the drawings in this application is only used to distinguish the various steps in the scheme, and is not used to limit the execution order of the various steps. The specific execution order is subject to the description in the specification.

[0034] In the field of communication technology, rich media messages can provide users with content information in various forms such as text, pictures, audio, video, location, contacts, etc. For example, 5G (5th Generation Mobile Communication Technology) messages are an upgrade of traditional SMS services, which can provide users with enhanced messaging services based on native SMS, without downloading applications. The main interactive forms of 5G messages currently include P2P (Person to Person), A2P (Application to Person) and P2A (Person to Application). Among them, A2P and P2A are mainly developed by B-side enterprises, organizations or individual developers based on CSP (Chatbot Server Platform) to interact with C-side users. The interactive forms include but are not limited to rich media or card message sending, card button / floating menu / fixed menu suggested replies, card button / floating menu / fixed menu suggested operations for link jumps, etc.

[0035] As the proliferation of spam text messages brings risks such as harassment, fraud and information security to users, in order to control spam text messages and prevent information fraud, the source of the message can be indicated by the message signature, thereby effectively controlling the problem of port-type spam text messages disturbing the public.

[0036] In order to solve the problems existing in the related art, the embodiment of the present application provides a method for generating a message signature. The generated message signature can be used to identify the identity of the message sender. The message signature can be widely used for message sending and receiving. Figure 1 As shown, the solution provided by the embodiment of the present application includes:

[0037] S11: Acquire identity information of a message sender, where the identity information includes a plurality of candidate keywords.

[0038] The embodiments of the present application can be applied to a variety of message platforms, and the message platform obtains the identity information of the message sender, and the message information can be the information provided by the message sender when registering on the message platform. For example, on the Chatbot management platform, merchants can register as merchants and create Chatbots required for their business. During the registration process, the identity information provided includes, for example: company name, company business scope information, business license image, province information, Chatbot business name, Chatbot service description, etc. In this step, words and phrases can be identified and extracted from the information provided by the merchant as candidate keywords and recorded as the identity information of the message sender.

[0039] S12: constructing a candidate keyword graph based on the multiple candidate keywords, the candidate keyword graph including nodes and edges connecting the nodes, the nodes corresponding to the candidate keywords in the identity information, and the edges representing the co-occurrence relationship between the connected nodes.

[0040] In this step, a candidate keyword graph corresponding to the identity information of the message sender is constructed. The candidate keyword graph includes multiple nodes and connections between nodes. The nodes correspond to the candidate keywords in the identity information, and the connections are used to represent the co-occurrence relationship between the candidate keywords.

[0041] For example, the candidate keyword graph constructed in this step is represented as G=(VN). A node set can be first constructed based on multiple candidate keywords, represented as V. Then, the co-occurrence relationship is used to construct an edge between any two points in the node set.

[0042] Optionally, in the step of constructing the edge connecting the nodes, the edge can be screened and optimized based on preset conditions. For example, for the words corresponding to each node, the two nodes are considered to have a common edge if the words corresponding to the two nodes co-occur in a window of length k. By setting the length k, the correlation between the nodes can be constrained so that the generated edge can represent a co-occurrence relationship that satisfies the constraint conditions.

[0043] S13: Determine the weight value of each node in the candidate keyword graph through a keyword extraction algorithm according to a preset linear layer function, wherein the preset linear layer function is used to determine the degree of matching between the candidate keywords and the positive sample keywords in the preset vocabulary, wherein the positive sample keywords are noun keywords with identity characteristics.

[0044] The above keyword extraction algorithm can be flexibly selected according to actual needs. In this example, the keyword extraction algorithm can be, for example, the TextRank algorithm, which can be used to determine the weight value of the importance of the nodes in the candidate keyword graph.

[0045] The above-mentioned preset linear layer function is used to evaluate whether the candidate keyword has identity characteristics based on the positive sample keywords in the preset vocabulary, that is, to evaluate whether the candidate keyword can express the identity of the message sender.

[0046] The preset word library includes positive sample keywords, which are nouns that can express identity. For example, positive sample keywords can be province names, company full names, company abbreviations, business abbreviations, etc.

[0047] Optionally, the preset linear layer function and TextRank algorithm in the present application example can be expressed as:

[0048]

[0049] Among them, W(v i )、W(v j ) is the node v i and v j The weight value; d is the damping coefficient or learning rate; k is the window size, which is used to limit the maximum number of co-occurring words to k; In(v i ) indicates all points to node v i Node, Out(v j ) represents the slave node v j Point to other nodes; W ij Represents node v i and v j The weight value of connecting common edges.

[0050] In this step, the weights of each node in the candidate keyword graph are iteratively propagated through the preset linear layer function and the TextRank algorithm, and the TextRank convergence value is calculated.

[0051] The preset linear layer function in this example can also be called f linear layer function, which is used to judge the importance of candidate keywords and score the candidate words so that the keywords in the data set D are hit. The candidate words with higher weights are used as the preferred candidate words to construct signatures. The dataset D is the preset word library, and the positive sample keywords are represented by p + , positive sample keywords are keywords that can represent identity and have identity characteristics, and negative sample keywords are represented by p - ,Negative sample keywords are keywords that cannot represent identity and do not have identity characteristics.

[0052] Through this step, each node in the candidate keyword graph can be equally divided. For meaningful keywords in the D data set, the f linear layer function score is positively superimposed and the weight convergence value is higher. For meaningless keywords, the f linear layer function score is negatively superimposed and the weight convergence value is lower. Therefore, based on the preset linear layer function, the keyword extraction algorithm is used to extract keywords with identity characteristics such as province, enterprise name, business name, etc. from the messy information through scoring.

[0053] S14: Select at least one target node from the candidate keyword graph according to the weight value of each node, and generate a message signature of the message sender with the candidate keyword corresponding to the at least one target node, wherein the weight value of any selected target node in the candidate keyword graph is greater than the weight value of any unselected node.

[0054] The weight values ​​determined for each node through the above steps can express the importance of the node in the candidate keyword graph and whether it has identity characteristics. In this step, at least one target node with a larger weight value is selected from the candidate keyword graph. The selected target node is relatively important in the candidate keyword graph and the corresponding keyword can better express the identity of the message sender. Generating a message signature with the keyword corresponding to at least one selected target node can enable the message signature to accurately express the identity of the message sender, thereby efficiently generating a message signature that can characterize the identity of the message sender.

[0055] For example, in this step, the nodes in the candidate keyword graph can be sorted in descending order based on the weight values, the first N nodes are selected as target nodes, and then the keywords corresponding to the N target nodes are used to generate message signatures. Wherein, N is a positive integer and can be flexibly set according to actual needs.

[0056] Through the scheme provided by the embodiment of the present application, the weight value can be determined for multiple candidate keywords in the identity information of the message sender through a keyword extraction algorithm according to a preset linear layer function, and the weight value can indicate the degree to which the candidate keywords have identity characteristics. Then, the candidate keywords corresponding to the target node with a large weight value are selected to generate the message signature of the sender. The message signature generated by this scheme is generated by the keywords in the identity information of the message sender and is authentic. And the keywords of the target node selected by the weight value are keywords with relatively identity characteristics in the identity information, which can accurately express the identity of the message sender, thereby efficiently generating a message signature that can characterize the identity of the message sender.

[0057] For the generated message signature, in actual applications, a message carrying the message signature can be flexibly generated by static templates and dynamic templates, so as to send a message that can effectively represent the identity of the message sender to the terminal. The following describes the specific schemes for generating messages using static templates and dynamic templates.

[0058] Based on the solution provided in the above embodiment, optionally, Figure 2a As shown, after the above step S14, that is, after selecting at least one target node from the candidate keyword graph according to the weight value of each node, and generating the message signature of the message sending end with the candidate keyword corresponding to the at least one target node, it also includes:

[0059] S21: Acquire information of a message to be sent from the message sending end, where the information of the message to be sent includes first message content and a first message type.

[0060] In this step, the information of the message to be sent can be obtained flexibly in a variety of ways. For example, the merchant can log in to the web page, applet, application, etc., and provide the message content and message type by uploading or online editing.

[0061] The message content may be, for example, text content typed online by the merchant, or may include other forms of content such as images and videos. The above message type may be related to the message content. For example, the merchant may select the desired first message type from a plurality of candidate message type options, and the platform may display a template corresponding to the first message type to the merchant, so that the merchant can edit the first message content based on the template of the first message type.

[0062] Figure 2b The left side shows an example interface where a merchant edits a message to be sent through the platform. Figure 2b The right side shows the target message generated based on the content edited on the left side.

[0063] The message type may include, for example, a text message fallback style type. Figure 2b The upper left side shows a variety of SMS fallback style types that merchants can choose from, including "SMS Mini Program 2.0", "Text SMS", "Video SMS", and "No Fallback". In this example, the merchant selected "Text SMS", and the message editing template corresponding to the "Text SMS" message type is displayed in the interface accordingly.

[0064] Based on the first message type selected by the merchant, an editing interface for the message content is displayed accordingly, in which the title of the message and the content under the title can be edited. In this example, the title is "5G message and automatic extraction and tail-end signature test of fallback", and the content under the title includes "5G message and automatic extraction and tail-end signature test of fallback" and a webpage link.

[0065] S22: Generate a first target message of the first message type including the first message content and a message signature of the message sender.

[0066] In this step, a first target message of the first message type is generated, and the first target message includes the first message content and the message signature of the message sender. The message signature can be located after the first message content, for example Figure 2b In the first target message shown on the right, the message signature [Polymenzation] is located after the link address of the first message content. In actual applications, the message signature can also be added to other positions of the first target message, such as before the title of the first message content, and the specific position can be flexibly set.

[0067] S23: Send the first target message to the user receiving terminal associated with the message sending end.

[0068] After the first target message is generated, the message is sent, wherein the information of the user receiving terminal associated with the message sending end can be provided in advance by the message sending end. For example, if a 5G rich media message is sent, the information of the user receiving terminal can include a mobile phone number. In this step, the message sending end can determine which mobile phone numbers the first target message is actually sent to. Alternatively, the first target message can be sent directly to the user receiving terminal associated with the message sending end according to a pre-set association relationship. The first target message displayed by the user receiving terminal is, for example, Figure 2c shown.

[0069] Through the solution provided in the embodiment of the present application, a first target message including a message signature is generated, wherein the message is generated based on a static template, and the message signature is already included in the message before the sending interface is called. The message signature that can represent the identity of the message sender can be efficiently generated, and then a message that can indicate the identity of the sender is sent to the user side.

[0070] Optionally, this step may call a message sending interface of the Chatbot message processing platform to send the first target message to a user receiving terminal associated with the message sending end.

[0071] Next, an example is used to illustrate the process of sending a tail-end signature message based on a static template.

[0072] First, the merchant creates a static template under the Chatbot specified by the CSP. The static template can correspond to the message type, such as 5G message plain text, short 2.0, video SMS, text SMS, etc. Subsequently, the CSP template submission review interface adds the AddSign field to automatically fill in the latest signature of the Chatbot, and assembles the complete template and submits it to the CMP review template. Then, the CMP returns the review result. If the review is passed, the merchant sends the template message through the CSP related message sending page function or the developer interface calling the CMPP message sending interface, and the message is sent to the terminal to display the tail signature in the message.

[0073] Among them, the user receiving terminal can receive messages sent by different message sending terminals. Different messages can carry various forms of content such as videos and links, for example Figure 2d shown.

[0074] Based on the solution provided in the above embodiment, optionally, Figure 3 As shown, after the above step S14, that is, after selecting at least one target node from the candidate keyword graph according to the weight value of each node, and generating the message signature of the message sending end with the candidate keyword corresponding to the at least one target node, it also includes:

[0075] S31: Acquire information of a message to be sent from the message sending end, where the information of the message to be sent includes a second message content and a second message type.

[0076] In this step, the information of the message to be sent can be obtained flexibly in a variety of ways. For example, the merchant can log in to the web page, applet, application, etc., and provide the message content and message type by uploading or online editing.

[0077] In practical applications, the interaction method may be similar to the above step S21, and the actual execution method may be flexibly set according to requirements.

[0078] S32: Generate a message body message of the second message type including the second message content and a preset placeholder, wherein the length of the preset placeholder is the length of the message signature of the message sender.

[0079] In this step, the merchant creates a dynamic template under the Chatbot specified by the CSP, and the dynamic template can correspond to the message type. The parameters in the dynamic template can change according to the mobile phone number of the target user on the message receiving side, and the template can be temporarily assembled by splicing parameters when the message is sent. Based on this, the message content, message type and other information edited by the message sending side can be submitted to the CMP for review. If the review is passed, the CMPP message sending interface is called to send the message. When the message body Bodytext content is added, a blank placeholder AddSign is added to fill in the latest signature, and the terminal displays the tail signature.

[0080] S33: Sending a second target message including the message signature to a user receiving terminal associated with the message sending end, wherein the message signature is located at the position of the preset placeholder.

[0081] Through the solution provided by the embodiment of the present application, a message signature placeholder is reserved in the form of a preset placeholder based on a dynamic template, and then after sending a message to the terminal through the message sending port, the message signature can be filled in the preset placeholder, so that after the message is sent to the terminal side, the message carrying the message signature can be displayed to the user. It can be seen that through this solution, a message signature that can represent the identity of the message sender can be efficiently generated, and then a message that can indicate the identity of the sender can be sent to the user side.

[0082] Optionally, the above step S33 specifically includes: calling the message sending interface of the Chatbot message processing platform to fill in the message signature of the message sending end in the preset placeholder in the message body message, and sending a second target message containing the message signature to the user receiving terminal associated with the message sending end.

[0083] Through the solution provided in the embodiment of the present application, the message sending interface of the Chatbot message processing platform is called to execute message sending, and the message signature is filled in the position of the preset placeholder, which can ensure that the second target message sent carries the message signature.

[0084] Based on the solution provided in the above embodiment, optionally, Figure 4 As shown, in the above step S11, obtaining the identity information of the message sender includes:

[0085] S41: Acquire the registration information of the message sender.

[0086] In this step, the message sender provides registration information to the platform. The registration information is used to indicate the identity of the message sender, and may specifically include the message sender's own information and the message sender's business information.

[0087] Optionally, the registration information includes at least one of a name, business classification information, authority information, and geographic location information.

[0088] For the registration information provided by the message sender, in order to facilitate the subsequent steps of word segmentation and keyword extraction, the registration information can be processed in advance. For example, the text information such as words, phrases, sentences, etc. in the registration information can be summarized into a set T = {S 1 ,S 2 ,...,S n}.

[0089] S42: extracting text information from the registration information and performing word segmentation.

[0090] In this step, the text information is segmented into multiple words, such as sentences and phrases, so that the part of speech of each word can be determined in the subsequent steps. Optionally, after obtaining multiple words, filtering is performed to remove stop words.

[0091] S43: extracting nouns from the segmented text information to obtain a plurality of candidate keywords.

[0092] In this step, part-of-speech tagging is performed on each word, and nouns are selected from multiple words according to the tagged part-of-speech to obtain multiple candidate keywords, which are represented as S = {t i1 ,t i2 ,...,t in}.

[0093] Through the scheme provided in the embodiment of the present application, text processing is performed on the registration information provided by the message sender, and noun keywords are extracted by word segmentation and part-of-speech tagging. Compared with words of other parts of speech, nouns are more likely to have identity characteristics and can better highlight the identity of the message sender. By screening noun keywords, it is helpful to determine a valid message signature in the subsequent steps.

[0094] Based on the solution provided in the above embodiment, optionally, Figure 5 As shown, before the above step S12, that is, before constructing a candidate keyword graph based on the multiple candidate keywords, it also includes:

[0095] S51: Obtain identity information of multiple message senders.

[0096] In the embodiment of the present application, identity information of multiple message senders is obtained to construct a preset word library, wherein the identity information of multiple message senders is authorized to provide to the platform by the multiple message senders, and the platform can construct a preset word library based on the authorized information.

[0097] In practical applications, the identity information provided by different message senders can express the identity from different perspectives. In this step, identity information from multiple different sources is aggregated and processed to obtain information expressing the identity from multiple different perspectives.

[0098] S52: extracting text information from the registration information and performing word segmentation to obtain a plurality of keywords to be marked.

[0099] If the registration information contains non-text information such as images and voices, the non-text information can be converted into text through image recognition, voice recognition, etc., so as to extract the text information.

[0100] Then, the text information is segmented, and sentences, phrases and other texts are processed into words, thereby obtaining multiple keywords to be marked.

[0101] Optionally, after performing word segmentation, the parts of speech may be marked for multiple word segments respectively, and noun-like words may be selected as keywords to be marked.

[0102] For example, multiple keywords to be labeled can be expressed in the form of a data set as D = {w 1 ,w 2 ,...,w n}, which may include information such as the company name, chatbot name (i.e. business name), province, etc. provided by the message sender.

[0103] S53: Annotate the multiple keywords to be annotated to obtain multiple positive sample keywords with identity characteristics and multiple negative sample keywords without identity characteristics.

[0104] In this step, multiple keywords to be annotated are annotated. For example, the keywords to be annotated are evaluated by labeling whether they have identity characteristics. Among them, keywords with identity characteristics can indicate the identity of the message sender and are used as positive sample keywords. On the contrary, negative sample keywords cannot indicate the identity of the message sender and do not have identity characteristics.

[0105] S54: constructing the preset word library based on the multiple positive sample keywords and the multiple negative sample keywords.

[0106] In practical applications, a certain number of positive sample keywords and negative sample keywords can be selected according to actual needs to build a preset vocabulary library, which stores keywords and corresponding attributes, that is, stores keywords and whether the keywords are positive samples or negative samples.

[0107] For example, the positive sample keyword can be expressed as p + , that is, the keywords with identity characteristics in D. Negative sample keywords can be expressed as p - That is, keywords in D that do not have identity characteristics. For example, "China Mobile", "10086", "Service", etc. in "China Mobile 10086 will serve you wholeheartedly" are meaningful positive sample keywords, and "Sincerely", "Your Service", etc. are meaningless negative sample keywords). Optionally, in actual applications, the identity information provided by the message sender often has identity characteristics. After processing through the above steps, there may be a large number of positive samples and too few negative samples. In order to balance the number of positive and negative samples in the preset vocabulary, keywords that do not have identity characteristics outside the data set D can be added to the preset vocabulary as negative samples to improve the quality of the preset vocabulary.

[0108] Based on the solution provided in the above embodiment, optionally, Figure 6a As shown, in the above step S14, at least one target node is selected from the candidate keyword graph according to the weight value of each node, and the message signature of the message sender is generated with the candidate keyword corresponding to the at least one target node, including:

[0109] S61: Generate a signature of a message to be verified using the candidate keyword corresponding to the at least one target node.

[0110] In this step, the signature of the message to be verified is generated based on the candidate keywords corresponding to the at least one target node. Specifically, multiple candidate keywords can be sequentially spliced ​​in descending order of the weights corresponding to the target nodes to generate the signature of the message to be verified.

[0111] S62: Verify the signature of the message to be verified based on preset verification rules.

[0112] In this step, the signature of the message to be verified is verified based on preset verification rules, and the actual verification rules can be flexibly set according to needs.

[0113] Optionally, the preset verification rule includes at least one of a message signature length rule, a message signature format rule, and a message signature word order rule.

[0114] For example, the signature length of the message to be verified is verified based on a preset number of candidate keywords or based on a preset number of words to ensure that the length of the generated message signature to be verified is appropriate. Whether there are illegal characters in the message signature to be verified is determined based on the preset message signature format rules. Whether the word order in the message signature is correct is identified based on the preset grammar rules. If there is a situation that does not comply with the rules, the reason for the non-compliance with the rules and the recommended modification method can be generated. Alternatively, the message signature that does not comply with the rules is modified based on the preset verification rules to obtain a message signature that complies with the rules.

[0115] For example, for SMS messages, this step can determine whether the number of signature characters exceeds the upper limit required by the SMS specification. If it exceeds, the merchant can manually modify the signature. If it does not exceed the upper limit, it will be submitted to CMP for review. At the same time, the system supports platform administrators to manually modify the signature to ensure that the signature is consistent with the facts and accurate.

[0116] S63: If the verification is successful, the message signature to be verified is determined as the message signature of the message sender.

[0117] In this step, the verified message signature can be associated and stored with the information of the message sender. If the merchant makes adjustments to the message signature later, it can also be modified and updated directly through the CSP or contact the platform administrator. The subsequent process is based on the latest signature to ensure that the message signature of the sent message is the latest message signature.

[0118] The present invention is further described below with reference to an example. Figure 6bA system architecture to which the method provided in the embodiment of the present application can be applied. In this system architecture, including CSP, CMP (Chatbot Management Platform), CMPP (Chatbot Message Processing Platform), MaaP (Messaging as a Platform), 5GMC (5G Message Center) and terminals, it can be used to implement sending 5G messages carrying message signatures to terminals.

[0119] based on Figure 6b The system shown in the figure, the solution provided in the embodiment of the present application can be implemented by the following process, see Figure 6c .

[0120] First, merchants register and create Chatbot through CSP, filling in information such as company name, chatbot name, service description, etc.

[0121] Then, through any of the solutions provided in the above embodiments of the present application, the message signature of the above merchant is obtained using the improved TextRank keyword extraction algorithm.

[0122] Next, based on the preset rules, it is determined whether the number of signature characters is greater than the upper limit required by the short color specification. If it exceeds, the review fails, and the merchant is supported to manually modify the signature. If it does not exceed, the review passes, and it is submitted to CMP for review.

[0123] During the CMP review phase, the signature can be manually modified to accurately express the merchant's identity. The CMP can provide feedback to the CSP to express the audit success or failure results.

[0124] If the review is passed, a static template is created, and a static template can be created according to the message type. The message type includes, for example, 5G message plain text, short 2.0, video SMS, text SMS, etc.

[0125] Subsequently, the AddSign field is added through the template interface to automatically fill in the merchant's Chatbot's latest signature to obtain a message carrying the message signature, and submit it to CMP for review of the above static template.

[0126] After CMP reviews the static template, it will provide feedback to CSP to express the review result of success or failure.

[0127] If the review is successful, the CMPP message sending interface is called to send the message, and then the message is sent to the terminal. The message received by the terminal carries the rear-end signature.

[0128] In addition, CSP can also send messages through dynamic templates. First, CSP creates a dynamic template, which can also be created according to the above message type. Then, the created dynamic template is submitted to CMP for review.

[0129] CMP feeds back the audit result indicating success or failure to CSP. If the audit is successful, the CMPP message sending interface is called to send the message. Among them, a blank placeholder AddSign is added after the Bodytext content to reserve a place to add the message signature. Subsequently, in the process of sending the message to the terminal based on the dynamic template, the merchant's Chatbot latest signature is filled in the placeholder, and a 5G message carrying the message signature is sent to the terminal, and the message with the rear-end signature is displayed to the user on the terminal side.

[0130] In the solution provided by the embodiment of the present application, the logic and technical implementation of the automatic tail-end signature are designed for the CSP to create static templates and dynamic templates with personalized differences. The blank placeholder AddSign field is added in the template encapsulation and message delivery temporary assembly, and the message body stage is combined with the algorithm to realize the automatic addition of signatures to the 5G messages and fallback SMS received by the terminal. This solution can realize a complete system based on CSP, CMP, CMPP and the whole process of the terminal to realize the automatic extraction and tail-end signature of 5G messages and fallback SMS, including the creation of chatbot to extract signatures, template creation, authentication and template review, message delivery, terminal reception, etc. Complete business logic and visual interface.

[0131] Through the solution provided by the embodiment of the present application, a linear layer function is added based on the TextRank keyword extraction algorithm to select limited noun keywords, and evaluate the keywords with identity characteristics through weights, so as to further narrow the scope of keyword hits and improve the accuracy of keyword extraction. For the identity information of the message sender, the embodiment of the present application summarizes and converges all the company names, chatbot names (i.e., business names), provinces and other content information submitted for certification and review through the text data set to build a preset vocabulary to achieve keyword weight evaluation. The linear layer function can score the importance of candidate keywords, so that the keywords in the hit data set D The candidate words with higher weights are prioritized to construct message signatures.

[0132] In practical applications, this solution can be widely used in signature extraction for example, such as 5G messages and fallback short messages 2.0, video text messages, text messages, etc., and can automatically extract signature forms that comply with the short message specifications, such as "province + company name", "company name + business name", etc., based on the original data such as the company name, business scope, uploaded business license, province, chatbot name (such as business name), and chatbot service description provided when the merchant is registered and the Chatbot is created.

[0133] The message signature generated by this solution can be a combination of abbreviations such as "province + company abbreviation", "company abbreviation + business abbreviation" formed by information such as company name, business scope, chatbot name, service description, and province. The signature supports editing and modification by merchants, and is used as the only signature of the real-name system after passing the certification review. In addition, in the process of sending 5G message rich media text and short 2.0, video SMS, and text SMS based on Chatbot, it can automatically read and append the latest signature of the chatbot without the need for merchants to enter it manually. The complete signature can be displayed when it is sent to different terminals. In the specific implementation method, CSP directly assembles the latest signature when creating a static template. After creating a dynamic template, CSP calls the CMPP message sending interface to add a blank placeholder AddSign after the Bodytext content to fill in the latest signature.

[0134] This solution adopts an unsupervised approach, without the need for additional public data samples to construct data set training. It only relies on the original text information provided by the merchant and makes full use of the differences and correlations between text nodes to start word segmentation, mark parts of speech, filter stop words, and improve the scores of meaningful keywords with identity characteristics from the same weight. It is used to separate and converge weights, and ultimately automatically select the optimal keyword combination suitable for 5G messages and fallback SMS signatures.

[0135] In order to solve the problems in the related art, the embodiment of the present application provides a device 70 for generating a message signature, such as Figure 7 As shown, including:

[0136] An acquisition module 71 acquires identity information of a message sender, wherein the identity information includes a plurality of candidate keywords;

[0137] A construction module 72 constructs a candidate keyword graph based on the multiple candidate keywords, the candidate keyword graph including nodes and edges connecting the nodes, the nodes corresponding to the candidate keywords in the identity information, and the edges representing the co-occurrence relationship between the connected nodes;

[0138] A determination module 73 determines the weight value of each node in the candidate keyword graph through a keyword extraction algorithm according to a preset linear layer function, wherein the preset linear layer function is used to determine the matching degree between the candidate keyword and the positive sample keyword in the preset vocabulary, wherein the positive sample keyword is a nominal keyword with identity characteristics;

[0139] A generation module 74 selects at least one target node from the candidate keyword graph according to the weight values ​​of each node, and generates a message signature of the message sender with the candidate keyword corresponding to the at least one target node, wherein the weight value of any selected target node in the candidate keyword graph is greater than the weight value of any unselected node.

[0140] Through the device provided by the embodiment of the present application, first, the identity information of the message sender is obtained, and the identity information includes multiple candidate keywords. Then, a candidate keyword graph is constructed based on the multiple candidate keywords, wherein the candidate keyword graph includes nodes and edges connecting the nodes, the nodes correspond to the candidate keywords in the identity information, and the edges represent the co-occurrence relationship between the connected nodes. Next, the weight values ​​of each node in the candidate keyword graph are determined by the keyword extraction algorithm according to the preset linear layer function, wherein the preset linear layer function is used to determine the matching degree between the candidate keywords and the positive sample keywords in the preset vocabulary, and the positive sample keywords are nominal keywords with identity characteristics. Subsequently, at least one target node is selected from the candidate keyword graph according to the weight value of each node, and the message signature of the message sender is generated with the candidate keywords corresponding to the at least one target node, wherein the weight value of any selected target node in the candidate keyword graph is greater than the weight value of any unselected node. Through the scheme provided by the embodiment of the present application, the weight values ​​of multiple candidate keywords in the identity information of the message sender can be determined by the keyword extraction algorithm according to the preset linear layer function, and the weight value can represent the degree to which the candidate keywords have identity characteristics. Then, the candidate keywords corresponding to the target node with a large weight value are selected to generate the message signature of the sender. The message signature generated by this scheme is generated by the keywords in the identity information of the message sender, and has authenticity. And the keywords of the target node selected by the weight value are keywords with relatively identity characteristics in the identity information, which can accurately express the identity of the message sender, thereby efficiently generating a message signature that can characterize the identity of the message sender.

[0141] Among them, the above modules in the device provided by the embodiment of the present application can also implement the method steps provided by the above method embodiment. Alternatively, the device provided by the embodiment of the present application can also include other modules in addition to the above modules to implement the method steps provided by the above method embodiment. And the device provided by the embodiment of the present application can achieve the technical effects that can be achieved by the above method embodiment.

[0142] Preferably, an embodiment of the present application also provides an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the various processes of the above-mentioned method embodiment for generating a message signature are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0143] The embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, each process of the above-mentioned method for generating a message signature is implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. The computer-readable storage medium is, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0144] An embodiment of the present application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program. The computer program can be operated to enable a computer to execute part or all of the steps of the above-mentioned method embodiment for generating a message signature, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0145] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0146] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0147] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0148] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0149] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0150] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0151] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0152] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0153] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0154] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for generating a message signature, characterized in that: include: Acquire identity information of a message sender, wherein the identity information includes a plurality of candidate keywords; constructing a candidate keyword graph based on the multiple candidate keywords, the candidate keyword graph comprising nodes and edges connecting the nodes, the nodes corresponding to the candidate keywords in the identity information, and the edges representing the co-occurrence relationship between the connected nodes; Determine the weight value of each node in the candidate keyword graph through a keyword extraction algorithm according to a preset linear layer function, wherein the preset linear layer function is used to determine the matching degree between the candidate keyword and the positive sample keyword in the preset vocabulary, wherein the positive sample keyword is a noun keyword with identity characteristics; At least one target node is selected from the candidate keyword graph according to the weight value of each node, and the message signature of the message sender is generated with the candidate keyword corresponding to the at least one target node, wherein the weight value of any selected target node in the candidate keyword graph is greater than the weight value of any unselected node.

2. The method according to claim 1, characterized in that After selecting at least one target node from the candidate keyword graph according to the weight values ​​of each node, and generating the message signature of the message sending end with the candidate keyword corresponding to the at least one target node, the method further includes: Acquire information of a message to be sent from the message sending end, wherein the information of the message to be sent includes first message content and a first message type; Generate a first target message of the first message type including the first message content and a message signature of the message sender; The first target message is sent to a user receiving terminal associated with the message sending end.

3. The method according to claim 2, characterized in that Sending the first target message to a user receiving terminal associated with the message sending end includes: The message sending interface of the Chatbot message processing platform is called to send the first target message to the user receiving terminal associated with the message sending end.

4. The method according to claim 1, characterized in that After selecting at least one target node from the candidate keyword graph according to the weight values ​​of each node, and generating the message signature of the message sending end with the candidate keyword corresponding to the at least one target node, the method further includes: Acquire information of a message to be sent from the message sending end, wherein the information of the message to be sent includes a second message content and a second message type; Generate a message body message of the second message type including the second message content and a preset placeholder, wherein the length of the preset placeholder is the length of the message signature of the message sender; A second target message including the message signature is sent to a user receiving terminal associated with the message sending end, wherein the message signature is located at the position of the preset placeholder.

5. The method according to claim 4, characterized in that Sending a second target message including the message signature to a user receiving terminal associated with the message sending end includes: The message sending interface of the Chatbot message processing platform is called to fill in the message signature of the message sending end in the preset placeholder in the message body message, and send a second target message containing the message signature to the user receiving terminal associated with the message sending end.

6. The method according to any one of claims 1 to 5, characterized in that: Get the identity information of the message sender, including: Obtaining registration information of the message sender; Extracting text information from the registration information and performing word segmentation; Nouns are extracted from the text information after word segmentation to obtain multiple candidate keywords.

7. The method according to claim 6, characterized in that The registration information includes at least one of a name, business classification information, authority information, and geographic location information.

8. The method according to any one of claims 1 to 5, characterized in that: Before constructing a candidate keyword graph based on the plurality of candidate keywords, the method further includes: Obtain identity information of multiple message senders; Extracting text information from the registration information and performing word segmentation to obtain a plurality of keywords to be marked; Performing labeling on the multiple keywords to be labeled to obtain multiple positive sample keywords with identity characteristics and multiple negative sample keywords without identity characteristics; The preset word library is constructed based on the multiple positive sample keywords and the multiple negative sample keywords.

9. The method according to any one of claims 1 to 5, characterized in that: Selecting at least one target node from the candidate keyword graph according to the weight value of each node, and generating a message signature of the message sending end with the candidate keyword corresponding to the at least one target node, including: Generating a signature of a message to be verified using a candidate keyword corresponding to the at least one target node; Performing verification on the signature of the message to be verified based on a preset verification rule; If the verification is successful, the message signature to be verified is determined as the message signature of the message sender.

10. The method according to claim 9, characterized in that The preset verification rule includes at least one of a message signature length rule, a message signature format rule, and a message signature word order rule.