Automatic short message content classification processing method based on intelligent algorithm

Abstract

The invention relates to the technical field of deep learning, and discloses a short message content automatic classification processing method based on an intelligent algorithm, and the method comprises the following steps: collecting user short message comprehensive information and to-be-recognized short message content, and carrying out the sorting based on the user short message comprehensive information, and obtaining first one-dimensional structure data; inputting the first one-dimensional structure data into a short message classification processing model, wherein the short message classification model comprises a first intermediate layer, a second intermediate layer, a third intermediate layer, a first intermediate fusion layer, a first output layer and a second output layer; and outputting the propagation link of the to-be-identified short message content and the classification processing result of the short message content. According to the method, the short message content to be processed can be identified and classified, and the information propagation link is predicted in combination with the communication behavior mode of the user.

Description

Technical Field

[0001] The present invention relates to the field of deep learning technology, and more specifically, to an automatic classification and processing method for text message content based on an intelligent algorithm. Background Art

[0002] With the popularity of mobile phone communications and the rapid development of social media, SMS has become one of the important ways for people to communicate and disseminate information in their daily lives. Traditional SMS processing methods often rely on manual review and simple filtering rules, which are not only inefficient but also unable to cope with complex content classification requirements and dynamic dissemination patterns. An automatic SMS content classification processing method based on intelligent algorithms has emerged. By introducing artificial intelligence and machine learning technology, this method aims to automatically extract and analyze SMS content and its dissemination information, and realize intelligent classification and processing of SMS. The intelligent algorithm can accurately classify SMS into specific categories by autonomously learning SMS features. Summary of the invention

[0003] The present invention provides an automatic classification and processing method for SMS content based on an intelligent algorithm. Through a set of complex data collection and processing mechanisms, the comprehensive information of user SMS and the information of propagation links are effectively organized and extracted. The orderly sorting of user SMS information, combined with the analysis of time series, can construct one-dimensional and two-dimensional structured data. This structured data is not only convenient for application in intelligent classification models, but also conforms to the characteristics of SMS information flow in real scenarios. In the model, the multi-level middle layer design ensures the layer-by-layer extraction of information flow and the fusion of complex features, and finally outputs accurate SMS classification processing results and propagation links. Under this framework, the system makes full use of the advantages of intelligent algorithms, realizes the automatic processing of SMS content classification and its propagation link prediction, and solves the technical problems of efficiency and accuracy of SMS management in related technologies.

[0004] The present invention provides a method for automatically classifying and processing short message content based on an intelligent algorithm, comprising the following steps:

[0005] Step 100, collecting comprehensive information of user SMS, which includes: SMS content information, SMS sending platform information, user information, and SMS rule information; obtaining the SMS content to be identified;

[0006] Step 200, based on the user's short message comprehensive information sorting, obtain the first-order one-dimensional structure data, the first-order one-dimensional structure data includes data items sorted by time, the fth data item represents the first two-dimensional structure data generated by the comprehensive information of user SMS collected at the fth moment;

[0007] Step 300, input the one-dimensional structure data No. 1 into the SMS classification processing model, the SMS classification processing model includes a first intermediate layer, a second intermediate layer, a third intermediate layer, a first intermediate fusion layer, a first output layer, and a second output layer; the one-dimensional structure data No. 1 is input into the first intermediate layer, and the first intermediate representation data is output to the second intermediate; the second intermediate layer outputs the second intermediate representation data to the first intermediate fusion layer; the SMS content to be identified is input into the third intermediate layer, and the third intermediate representation data is output to the first intermediate fusion layer; the first intermediate fusion layer outputs the first fusion representation data to the first output layer and the second output layer; the first output layer outputs the propagation link of the SMS content to be identified, and the second output layer outputs the classification processing result of the SMS content to be identified.

[0008] In a preferred embodiment, the SMS content information includes: text content, image content, link content;

[0009] The SMS sending platform information is represented as a description text of the SMS sending platform consisting of operator categories;

[0010] User information includes: user description text consisting of user's phone number, user's location, and the size of the user's social network;

[0011] SMS rules include: SMS security rules and SMS compliance rules, which are composed of SMS rule description text.

[0012] In a preferred embodiment, the user's short message comprehensive information is subjected to feature engineering to obtain corresponding features;

[0013] The text describing the SMS sending platform, the user description, and the SMS rule description in the comprehensive SMS information of the user are used as a feature engineering method to generate features of the SMS sending platform, user features, and SMS rule features;

[0014] SMS content information includes text content, image content, and link content. The text content and link content are used as a feature engineering method through the WordEmbedding algorithm to generate SMS text content features and SMS link content features. The expression image content is used as a feature engineering method through the convolutional neural network CNN algorithm to generate SMS image content features. The SMS text content features, SMS link content features, and SMS image content features are spliced ​​to obtain the SMS content features.

[0015] In a preferred embodiment, the constructed two-dimensional structure data No. 1 includes:

[0016] The No. 1 two-dimensional structure data includes a data matrix and a relationship matrix. A unit of the data matrix represents the comprehensive information of user SMS of an independent object. The independent object includes SMS content, SMS sending platform, user, and SMS rules. A unit of the data matrix only contains the comprehensive information of user SMS of the independent object it represents.

[0017] The element in the i-th row and j-th column of the relationship matrix represents the association between the i-th cell of the data matrix and the independent object represented by the j-th cell. If there is an association, the value of this element of the relationship matrix is ​​1, otherwise it is 0.

[0018] In a preferred embodiment, the association relationship between independent objects is expressed as follows:

[0019] The existence of association between users means: there is a record of communication between two users, where communication includes SMS communication and other communication, where other communication refers to communication through commonly used social platforms;

[0020] The association between a user and the content of a text message means: the user is the sender or receiver of the text message;

[0021] The association between SMS and SMS rules means that: the content of SMS is subject to the information of SMS rules;

[0022] The association between SMS and SMS sending platform means that the SMS content is sent by the SMS platform.

[0023] In a preferred embodiment, the classification processing result of the SMS content to be identified includes:

[0024] Whether the content of the SMS to be identified is compliant, with 1 indicating compliance and 0 indicating non-compliance;

[0025] Functional classification of the SMS content to be identified, including: notification SMS, marketing SMS, verification SMS, transaction processing SMS, harmful SMS, news and information SMS, and social interaction SMS;

[0026] The sentiment classification of the SMS content to be identified, sentiment classification represents negative, neutral, and positive.

[0027] In a preferred embodiment, the meanings of positive, neutral, and negative in the sentiment classification are different in different business scenarios. In a sales scenario, positive means that the customer has a high interest in the product promoted by the salesperson, neutral means that the customer has an indifferent attitude towards the product promoted by the salesperson, and negative means that the customer is not interested in the product promoted by the salesperson; in a customer service scenario, positive means that the customer is satisfied with the service experience, neutral means that the customer feels average about the service experience, and negative means that the customer is dissatisfied with the service experience.

[0028] In a preferred embodiment, the second output layer is trained separately, and the training model has the ability to accurately judge the compliance of the text messages to be identified, the functional category of the text messages to be identified, and the emotional category of the text messages to be identified. The loss function calculation formula is as follows:

[0029] ;

[0030] Among them, LOSS is the total loss value, is the functional classification loss of the SMS content to be identified, is the compliance classification loss of the SMS content to be identified, is the sentiment classification loss of the text message content to be identified.

[0031] ;

[0032] in represents the function classification loss value of the SMS content to be identified, M represents the number of categories of the SMS content to be identified, Indicates the total number of SMS contents to be identified. It is The symbol value of the SMS content to be identified. If The actual classification of the content of the SMS to be identified is the cth functional category, and its value is 1, otherwise it is 0. is the first output layer in the output The probability that the content of the SMS to be identified belongs to the cth functional category, Represents the logarithmic function with the natural constant e as the base.

[0033] ;

[0034] in, Indicates the compliance classification loss value of the SMS content to be identified. Indicates the total number of SMS contents to be identified. It is The real classification label of the SMS content to be identified is 0 or 1, 1 means that the SMS content of the user is compliant, and 0 means that the SMS content of the user is not compliant. is the first output layer in the output Compliance classification labels for the SMS content to be identified, Represents the logarithmic function with the natural constant e as the base.

[0035] ;

[0036] in Represents the sentiment classification loss value of the SMS content to be identified, Indicates the number of categories of emotions in the text message content to be identified, Indicates the total number of SMS contents to be identified. It is The symbol value of the SMS content to be identified. The true emotional classification of the SMS content to be identified is The value of sentiment category is 1, otherwise it is 0. is the first output layer in the output The content of the SMS to be identified belongs to The probability of sentiment categories, Represents the logarithmic function with the natural constant e as the base.

[0037] In a preferred embodiment, the first output layer is trained, and the training model has the ability to accurately predict the SMS propagation link. The loss function calculation formula is as follows:

[0038] ;

[0039] in, Indicates the predicted loss value of the transmission link of the SMS content to be identified, Indicates the total number of SMS contents to be identified. is the true label of the transmission of the SMS content to be identified between the i-th and j-th users, and its value is 0 or 1. 1 indicates that there is the transmission of the SMS content to be identified between the i-th and j-th users, and 0 indicates that there is no transmission of the SMS content to be identified between the i-th and j-th users. is the probability value of the transmission of the SMS content to be identified between the i-th and j-th users in the output of the first output layer. The probability value is expressed as the element value of the i-th row and j-th column in the first output matrix. Represents the logarithmic function with the natural constant e as the base.

[0040] The present invention provides a storage medium storing non-temporary computer-readable instructions. When the non-temporary computer-readable instructions are executed by a computer, the steps in the above-mentioned method for automatic classification and processing of SMS content based on an intelligent algorithm can be executed.

[0041] The beneficial effects of the present invention are:

[0042] Improve classification accuracy: By introducing intelligent algorithms and machine learning technology, the present invention can automatically learn and identify the features in the text message content, greatly improving the accuracy and precision of text message classification;

[0043] Diversified information processing capabilities: The model can not only process text information, but also analyze it in combination with multiple information formats such as images and links. In addition, it can respond to different types of SMS content in a dynamic manner, which improves the flexibility and adaptability of overall information processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 The present invention is a flowchart of a method for automatically classifying and processing short message content based on an intelligent algorithm. DETAILED DESCRIPTION

[0045] The subject matter described herein will now be discussed with reference to example implementations. It should be understood that the discussion of these implementations is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the present specification. Various examples may omit, replace, or add various processes or components as needed. In addition, the features described in some examples may also be combined in other examples.

[0046] At least one embodiment of the present invention discloses a method for automatically classifying and processing short message content based on an intelligent algorithm, such as Figure 1 As shown, the following steps are included:

[0047] Step 100, collecting comprehensive information of user SMS, which includes: SMS content information, SMS sending platform information, user information, and SMS rule information; obtaining the SMS content to be identified;

[0048] In one embodiment of the present invention, the SMS content information includes: text content, image content, link content;

[0049] The SMS sending platform information is represented as a description text of the SMS sending platform consisting of operator categories;

[0050] User information includes: user description text consisting of user's phone number, user's location, and the size of the user's social network;

[0051] SMS rules include: SMS security rules and SMS compliance rules, which are composed of SMS rule description text.

[0052] In one embodiment of the present invention, the user's short message comprehensive information is subjected to feature engineering to obtain corresponding features;

[0053] The text describing the SMS sending platform, the user description, and the SMS rule description in the comprehensive SMS information of the user are used as a feature engineering method to generate features of the SMS sending platform, user features, and SMS rule features;

[0054] SMS content information includes text content, image content, and link content. The text content and link content are used as a feature engineering method through the WordEmbedding algorithm to generate SMS text content features and SMS link content features. The expression image content is used as a feature engineering method through the convolutional neural network CNN algorithm to generate SMS image content features. The SMS text content features, SMS link content features, and SMS image content features are spliced ​​to obtain the SMS content features.

[0055] Step 200, based on the user's short message comprehensive information sorting, obtain the first-order one-dimensional structure data, the first-order one-dimensional structure data includes data items sorted by time, the fth data item represents the first two-dimensional structure data generated by the comprehensive information of user SMS collected at the fth moment;

[0056] The No. 1 two-dimensional structure data includes a data matrix and a relationship matrix. A unit of the data matrix represents the comprehensive information of user SMS of an independent object. The independent object includes SMS content, SMS sending platform, user, and SMS rules. A unit of the data matrix only contains the comprehensive information of user SMS of the independent object it represents.

[0057] The element in the i-th row and j-th column of the relationship matrix represents the association between the i-th unit of the data matrix and the independent object represented by the j-th unit. If there is an association, the value of this element of the relationship matrix is ​​1, otherwise it is 0;

[0058] The existence of association between users means: there is a record of communication between two users, where communication includes SMS communication and other communication, where other communication refers to communication through commonly used social platforms;

[0059] The association between a user and the content of a text message means: the user is the sender or receiver of the text message;

[0060] The association between SMS and SMS rules means that: the content of SMS is subject to the information of SMS rules;

[0061] The association between SMS and SMS sending platform means that the SMS content is sent by the SMS platform.

[0062] Step 300, input the one-dimensional structure data No. 1 into the SMS classification processing model, the SMS classification processing model includes a first intermediate layer, a second intermediate layer, a third intermediate layer, a first intermediate fusion layer, a first output layer, and a second output layer; the one-dimensional structure data No. 1 is input into the first intermediate layer, and the first intermediate representation data is output to the second intermediate; the second intermediate layer outputs the second intermediate representation data to the first intermediate fusion layer; the SMS content to be identified is input into the third intermediate layer, and the third intermediate representation data is output to the first intermediate fusion layer; the first intermediate fusion layer outputs the first fusion representation data to the first output layer and the second output layer; the first output layer outputs the propagation link of the SMS content to be identified, and the second output layer outputs the classification processing result of the SMS content to be identified.

[0063] In one embodiment of the present invention, the calculation formula of the first intermediate layer is as follows:

[0064] ;

[0065] ;

[0066] ;

[0067] in Indicates the first one-dimensional structure data The first The first intermediate representation data of the unit, and Respectively represent the first one-dimensional structure data The first and The comprehensive information of user SMS messages represented by independent objects in each unit. Indicates the first one-dimensional structure data The first and The attention coefficient of each unit, Indicates the first one-dimensional structure data The first and The attention coefficient of each unit, Indicates the first one-dimensional structure data The first and The normalized attention coefficient of each unit, Indicates the first one-dimensional structure data The first The set of cells of the data matrix where there are associations between cells, represents an exponential function with a natural constant as base, represents the trainable attention vector parameter, CONCAT represents the concatenation operation, represents transpose, represents the two-dimensional recognition weight parameter, and LeakyReLU is the LeakyReLU activation function.

[0068] In one embodiment of the present invention, the calculation formula of the second intermediate layer is as follows:

[0069] ;

[0070] in Indicates A second intermediate representation data, Indicates A second intermediate representation data, , The total number of data items representing the one-dimensional structure data. , Represents the first one-dimensional structure data The set of cells of all data matrices with data items, and are the first and second weight parameters, is the first bias parameter and tanh is the hyperbolic tangent function.

[0071] In one embodiment of the present invention, the calculation formula of the third intermediate layer is as follows:

[0072] ;

[0073] in represents the third intermediate representation data, represents the third weight parameter, Indicates the content features of the input SMS to be identified. represents the second bias parameter, Represents a multi-layer perceptron.

[0074] In one embodiment of the present invention, the calculation formula of the first data fusion layer is as follows:

[0075] ;

[0076] in, represents the fusion representation data, represents the fusion function, Indicates A second intermediate representation data, represents the third intermediate representation data, represents the sum weight matrix, Represents the summed bias parameter.

[0077] In one embodiment of the present invention, the calculation formula of the first output layer is as follows:

[0078] ;

[0079] in represents the first output matrix. If the element in the i-th row and j-th column of the matrix is ​​greater than 0.5, it means that there is a transmission of the SMS content to be identified between the i-th and j-th users. Otherwise, it does not exist. is the first output weight parameter, is the first output bias parameter, Represents the sigmoid function.

[0080] In one embodiment of the present invention, the calculation formula of the second output layer is as follows:

[0081] ;

[0082] in Represents the second output matrix, the first row of the matrix Column represents the Whether the SMS content to be identified is compliant, 1 means compliant, 0 means non-compliant. Column represents the The third row of the matrix is ​​the function classification of the SMS content to be identified. Column represents the The sentiment classification of the SMS content to be identified is expressed as positive, neutral, and negative. is the second output weight parameter, is the second output bias parameter, Represents the sigmoid function.

[0083] It should be noted that the positive, neutral, and negative sentiments in the sentiment classification have different meanings in different business scenarios. For example, in a sales scenario, positive means that customers have a high interest in the products sold by the salesperson, neutral means that customers have an indifferent attitude towards the products sold by the salesperson, and negative means that customers are not interested in the products sold by the salesperson. In a customer service scenario, positive means that customers are satisfied with the service experience, neutral means that customers feel average about the service experience, and negative means that customers are dissatisfied with the service experience.

[0084] In one embodiment of the present invention, the categories of functions of the SMS content to be identified are represented as notification SMS, marketing SMS, verification SMS, transaction processing SMS, harmful SMS, news information SMS, and social interaction SMS;

[0085] Notification SMS messages include: information notifications, reminders, and announcements; marketing SMS messages include: product promotion, advertising, and sales promotion; verification SMS messages include: identity authentication and verification codes; transaction processing SMS messages include: bill notifications, business confirmations, and payment reminders; harmful SMS messages contain deceptive or unwanted content; news and information SMS messages include: news summaries and information sharing; social interaction SMS messages include: personal communication, greetings, and social interaction.

[0086] In one embodiment of the present invention, the second output layer is trained separately, and the training model has the ability to accurately judge the compliance of the text messages to be identified, the functional category of the text messages to be identified, and the emotional category of the text messages to be identified. The loss function calculation formula is as follows:

[0087] ;

[0088] Among them, LOSS is the total loss value, is the functional classification loss of the SMS content to be identified, is the compliance classification loss of the SMS content to be identified, is the sentiment classification loss of the text message content to be identified.

[0089] ;

[0090] in represents the function classification loss value of the SMS content to be identified, M represents the number of categories of the SMS content to be identified, Indicates the total number of SMS contents to be identified. It is The symbol value of the SMS content to be identified. If The actual classification of the content of the SMS to be identified is the cth functional category, and its value is 1, otherwise it is 0. is the first output layer in the output The probability that the content of the SMS to be identified belongs to the cth functional category, Represents the logarithmic function with the natural constant e as the base.

[0091] ;

[0092] in, Indicates the compliance classification loss value of the SMS content to be identified. Indicates the total number of SMS contents to be identified. It is The real classification label of the SMS content to be identified is 0 or 1, 1 means that the SMS content of the user is compliant, and 0 means that the SMS content of the user is not compliant. is the first output layer in the output Compliance classification labels for the SMS content to be identified, Represents the logarithmic function with the natural constant e as the base.

[0093] ;

[0094] in Represents the sentiment classification loss value of the SMS content to be identified, Indicates the number of categories of emotions in the text message content to be identified, Indicates the total number of SMS contents to be identified. It is The symbol value of the SMS content to be identified. The true emotional classification of the SMS content to be identified is The value of sentiment category is 1, otherwise it is 0. is the first output layer in the output The content of the SMS to be identified belongs to The probability of sentiment categories, Represents the logarithmic function with the natural constant e as the base.

[0095] In one embodiment of the present invention, the first output layer is trained, and the training model has the ability to accurately predict the SMS propagation link. The loss function calculation formula is as follows:

[0096] ;

[0097] in, Indicates the predicted loss value of the transmission link of the SMS content to be identified, Indicates the total number of SMS contents to be identified. is the true label of the transmission of the SMS content to be identified between the i-th and j-th users, and its value is 0 or 1. 1 indicates that there is the transmission of the SMS content to be identified between the i-th and j-th users, and 0 indicates that there is no transmission of the SMS content to be identified between the i-th and j-th users. is the probability value of the transmission of the SMS content to be identified between the i-th and j-th users in the output of the first output layer. The probability value is expressed as the element value of the i-th row and j-th column in the first output matrix. Represents the logarithmic function with the natural constant e as the base.

[0098] A storage medium stores non-temporary computer-readable instructions. When the non-temporary computer-readable instructions are executed by a computer, the steps in the above-mentioned method for automatic classification and processing of SMS content based on an intelligent algorithm can be executed.

[0099] The above describes an embodiment of the present invention, but this embodiment is not limited to the above-mentioned specific implementation mode. The above-mentioned specific implementation mode is merely illustrative and not restrictive. Under the guidance of this embodiment, ordinary technicians in this field can also make more forms of equivalent embodiments, all of which are within the protection of this embodiment.

Claims

1. A method for automatic classification and processing of SMS content based on intelligent algorithm, characterized in that: The following steps are involved: Step 100, collecting comprehensive information of user SMS, which includes: SMS content information, SMS sending platform information, user information, and SMS rule information; obtaining the SMS content to be identified; Step 200, based on the user's short message comprehensive information sorting, obtain the first-order one-dimensional structure data, the first-order one-dimensional structure data includes data items sorted by time, the fth data item represents the first two-dimensional structure data generated by the comprehensive information of user SMS collected at the fth moment; Step 300, input the one-dimensional structure data No. 1 into the SMS classification processing model, the SMS classification processing model includes a first intermediate layer, a second intermediate layer, a third intermediate layer, a first intermediate fusion layer, a first output layer, and a second output layer; the one-dimensional structure data No. 1 is input into the first intermediate layer, and the first intermediate representation data is output to the second intermediate; the second intermediate layer outputs the second intermediate representation data to the first intermediate fusion layer; the SMS content to be identified is input into the third intermediate layer, and the third intermediate representation data is output to the first intermediate fusion layer; the first intermediate fusion layer outputs the first fusion representation data to the first output layer and the second output layer; the first output layer outputs the propagation link of the SMS content to be identified, and the second output layer outputs the classification processing result of the SMS content to be identified.

2. According to claim 1, a method for automatic classification and processing of SMS content based on intelligent algorithm is characterized in that: SMS content information includes: text content, image content, link content; The SMS sending platform information is represented as a description text of the SMS sending platform consisting of operator categories; User information includes: user description text consisting of user's phone number, user's location, and the size of the user's social network; SMS rules include: SMS security rules and SMS compliance rules, which are composed of SMS rule description text.

3. The method for automatic classification and processing of SMS content based on intelligent algorithm according to claim 1, characterized in that: Obtain corresponding features from comprehensive information of user SMS through feature engineering; The text describing the SMS sending platform, the user description, and the SMS rule description in the comprehensive SMS information of the user are used as a feature engineering method through the WordEmbedding algorithm to generate features of the SMS sending platform, user features, and SMS rule features; SMS content information includes text content, image content, and link content. The text content and link content are used as a feature engineering method through the WordEmbedding algorithm to generate SMS text content features and SMS link content features. The expression image content is used as a feature engineering method through the convolutional neural network CNN algorithm to generate SMS image content features. The SMS text content features, SMS link content features, and SMS image content features are spliced ​​to obtain the SMS content features.

4. The method for automatic classification and processing of SMS content based on intelligent algorithm according to claim 1, characterized in that: The constructed No. 1 two-dimensional structure data includes: The No. 1 two-dimensional structure data includes a data matrix and a relationship matrix. A unit of the data matrix represents the comprehensive information of user SMS of an independent object. The independent object includes SMS content, SMS sending platform, user, and SMS rules. A unit of the data matrix only contains the comprehensive information of user SMS of the independent object it represents. The element in the i-th row and j-th column of the relationship matrix represents the association between the i-th cell of the data matrix and the independent object represented by the j-th cell. If there is an association, the value of this element of the relationship matrix is ​​1, otherwise it is 0.

5. The method for automatic classification and processing of SMS content based on intelligent algorithm according to claim 4, characterized in that: The association relationship between independent objects is expressed as follows: The existence of association between users means: there is a record of communication between two users, where communication includes SMS communication and other communication, where other communication refers to communication through commonly used social platforms; The association between a user and the content of a text message means: the user is the sender or receiver of the text message; The association between SMS and SMS rules means that: the content of SMS is subject to the information of SMS rules; The association between SMS and SMS sending platform means that the SMS content is sent by the SMS platform.

6. The method for automatic classification and processing of SMS content based on intelligent algorithm according to claim 1, characterized in that: The classification results of the SMS content to be identified include: Whether the content of the SMS to be identified is compliant, with 1 indicating compliance and 0 indicating non-compliance; Functional classification of the SMS content to be identified, including: notification SMS, marketing SMS, verification SMS, transaction processing SMS, harmful SMS, news and information SMS, and social interaction SMS; The sentiment classification of the SMS content to be identified, sentiment classification represents negative, neutral, and positive.

7. The method for automatic classification and processing of SMS content based on intelligent algorithm according to claim 6, characterized in that: The positive, neutral, and negative sentiments in the sentiment classification have different meanings in different business scenarios. In the sales scenario, positive means that the customer has a high interest in the product promoted by the salesperson, neutral means that the customer has an indifferent attitude towards the product promoted by the salesperson, and negative means that the customer is not interested in the product promoted by the salesperson. In the customer service scenario, positive means that the customer is satisfied with the service experience, neutral means that the customer feels average about the service experience, and negative means that the customer is dissatisfied with the service experience.

8. The method for automatic classification and processing of SMS content based on intelligent algorithm according to claim 1, characterized in that: The second output layer is trained separately. The training model has the ability to accurately judge the compliance of the SMS to be identified, the functional category of the SMS to be identified, and the emotional category of the SMS to be identified. The loss function calculation formula is as follows: ； Among them, LOSS is the total loss value, is the functional classification loss of the SMS content to be identified, is the compliance classification loss of the SMS content to be identified, is the sentiment classification loss of the text message content to be identified; ； in represents the function classification loss value of the SMS content to be identified, M represents the number of categories of the SMS content to be identified, Indicates the total number of SMS contents to be identified. It is The symbol value of the SMS content to be identified. If The actual classification of the content of the SMS to be identified is the cth functional category, and its value is 1, otherwise it is 0. is the first output layer in the output The probability that the content of the SMS to be identified belongs to the cth functional category, It represents the logarithmic function with the natural constant e as the base; ； in, Indicates the compliance classification loss value of the SMS content to be identified. Indicates the total number of SMS contents to be identified. It is The real classification label of the SMS content to be identified is 0 or 1, where 1 means that the SMS content of the user is compliant, and 0 means that the SMS content of the user is not compliant. is the first output layer in the output Compliance classification labels for the SMS content to be identified, It represents the logarithmic function with the natural constant e as the base; ； in Represents the sentiment classification loss value of the text message content to be identified, Indicates the number of categories of emotions in the text message content to be identified, Indicates the total number of SMS contents to be identified. It is The symbol value of the SMS content to be identified. The true classification of the emotion of the SMS content to be identified is The value of sentiment category is 1, otherwise it is 0. is the first output layer in the output The content of the SMS to be identified belongs to The probability of sentiment categories, Represents the logarithmic function with the natural constant e as the base.

9. The method for automatic classification and processing of SMS content based on intelligent algorithm according to claim 1, characterized in that: The first output layer is trained, and the training model has the ability to accurately predict the SMS propagation link. The loss function calculation formula is as follows: ； in, Indicates the predicted loss value of the transmission link of the SMS content to be identified, Indicates the total number of SMS contents to be identified. is the true label of the transmission of the SMS content to be identified between the i-th and j-th users, and its value is 0 or 1. 1 indicates that there is the transmission of the SMS content to be identified between the i-th and j-th users, and 0 indicates that there is no transmission of the SMS content to be identified between the i-th and j-th users. is the probability value of the transmission of the SMS content to be identified between the i-th and j-th users in the output of the first output layer. The probability value is expressed as the element value of the i-th row and j-th column in the first output matrix. Represents the logarithmic function with the natural constant e as the base.

10. A storage medium storing non-transitory computer-readable instructions, characterized in that: When the non-transitory computer-readable instructions are executed by a computer, the steps in the method for automatic classification and processing of SMS content based on an intelligent algorithm as described in any one of claims 1 to 9 can be executed.

Images