A mail management method, device, storage medium and program product

Abstract

This application provides an email management method, apparatus, storage medium, and program product. The method includes: upon receiving an email to be processed for a target object, performing feature extraction processing on the email to be processed to obtain first feature representation information of the email to be processed; determining a matching degree between the email to be processed and the target object based on the first feature representation information and second feature representation information of the target object; wherein the second feature representation information is obtained by performing feature extraction processing on emails included in a first email set associated with the target object; and performing management operations on the email to be processed according to the matching degree, the management operations including rejecting receipt or adding it to a corresponding email list. Using the method provided in this application, emails can be personalized according to the matching degree between the email to be processed and the target object, effectively improving the accuracy of email management.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to an email management method, an email management device, a computer-readable storage medium, and a computer program product. Background Technology

[0002] With the development of internet technology, email has become a crucial communication tool for people in their studies and work. Email is a communication method that uses electronic means to exchange information. People can quickly send text, pictures, audio, and video messages to other users, and also obtain a large amount of news, special reports, and other information via email. However, the emergence of spam causes people to waste time and energy sorting through emails, and can even lead to financial losses due to deceptive or misleading emails.

[0003] Traditional spam detection technology uses trained models to identify and block new emails. However, this method can lead to false positives or false negatives, resulting in poor accuracy in email management. For example, advertising emails may not be spam for users who need the advertised product, but they may be spam for users who don't need the advertised product, and therefore need to be blocked. Therefore, improving the accuracy of email management is a pressing issue that needs to be addressed. Summary of the Invention

[0004] This application provides an email management method, apparatus, storage medium, and program product, which can manage emails to be processed based on the matching degree of feature representation information between the email to be processed and the emails associated with the target object, effectively improving the accuracy of email management.

[0005] On one hand, embodiments of this application provide an email management method, wherein the method includes:

[0006] When a pending email is received from the target object, feature extraction processing is performed on the pending email to obtain the first feature representation information of the pending email;

[0007] Based on the first feature representation information and the second feature representation information of the target object, the matching degree between the email to be processed and the target object is determined; wherein, the second feature representation information is obtained by feature extraction processing of emails included in the first email set associated with the target object;

[0008] The pending emails are managed based on the matching degree, and the management operations include rejecting them or adding them to the corresponding email list.

[0009] On one hand, embodiments of this application provide an email management device, wherein the device includes:

[0010] The processing module is used to perform feature extraction processing on the email to be processed when it receives the email to be processed from the target object, so as to obtain the first feature representation information of the email to be processed.

[0011] The determining module is used to determine the matching degree between the email to be processed and the target object based on the first feature representation information and the second feature representation information of the target object; wherein, the second feature representation information is obtained by feature extraction processing of emails included in the first email set associated with the target object;

[0012] The processing module is also used to perform management operations on the emails to be processed based on the matching degree, including rejecting them or adding them to the corresponding email list.

[0013] On one hand, this application provides a server, which includes a processor, a communication interface, and a memory. The processor, the communication interface, and the memory are interconnected. The memory stores executable program code, and the processor is used to call the executable program code to implement the email management method provided in this application.

[0014] Accordingly, embodiments of this application also provide a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program that is adapted to be loaded by a processor and implement the email management method provided in embodiments of this application.

[0015] Accordingly, this application also provides a computer program product, wherein the computer program product includes a computer program that, when executed by a processor, implements the email management method provided in this application.

[0016] In this embodiment, when the server receives an email to be processed from a target object, it performs feature extraction processing on the email to obtain first feature representation information of the email to be processed. Based on the first feature representation information and the second feature representation information of the target object, it determines the matching degree between the email to be processed and the target object, and performs management operations on the email to be processed according to the matching degree. The second feature representation information is obtained by extracting features from emails included in a first email set associated with the target object. The management operations include rejecting the email or adding it to the corresponding email list. Using the method provided in this embodiment, feature extraction processing can be performed on the email to be processed to obtain first feature representation information, and feature extraction can be performed on emails included in a first email set associated with the target object to obtain second feature representation information. The matching degree between the first and second feature representation information can be determined, and management operations can be performed on the email to be processed according to the matching degree, effectively improving the accuracy of email management. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0018] Figure 1 This is a schematic diagram of the structure of an email management system provided in an embodiment of this application;

[0019] Figure 2 This is a flowchart illustrating an email management method provided in an embodiment of this application;

[0020] Figure 3 This is a schematic diagram of a process for determining second feature representation information provided in an embodiment of this application;

[0021] Figure 4 This is a flowchart illustrating another email management method provided in an embodiment of this application;

[0022] Figure 5 This is a schematic diagram of a process for extracting text feature representation information provided in an embodiment of this application;

[0023] Figure 6 This is a schematic diagram of the structure of a text feature extraction model provided in an embodiment of this application;

[0024] Figure 7This is a schematic diagram of a text feature vector extraction process provided in an embodiment of this application;

[0025] Figure 8 This is a schematic diagram of the structure of an email management device provided in an embodiment of this application;

[0026] Figure 9 This is a schematic diagram of the structure of a server provided in an embodiment of this application. Detailed Implementation

[0027] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0028] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, components, features, and elements with the same names in different embodiments of this application may have the same meaning or different meanings, the specific meaning of which must be determined by its interpretation in that specific embodiment or further in conjunction with the context of that specific embodiment.

[0029] It should be understood that although the terms first, second, third, etc., may be used herein to describe various information, such information should not be limited to these terms. These terms are used only to distinguish information of the same type from one another. For example, without departing from the scope of this document, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if," as used herein, may be interpreted as "when," "when," or "in response to determination." Furthermore, as used herein, the singular forms "a," "an," and "the" are intended to also include the plural forms unless the context indicates otherwise. It should be further understood that the terms "comprising," "including," indicate the presence of the stated feature, step, operation, element, component, item, kind, and / or group, but do not exclude the presence, occurrence, or addition of one or more other features, steps, operations, elements, components, items, kinds, and / or groups. The terms "or," "and / or," "including at least one of the following," etc., as used in this application, may be interpreted as inclusive, or mean any one or any combination thereof. For example, "including at least one of the following: A, B, C" means "any one of the following: A; B; C; A and B; A and C; B and C; A and B and C." Similarly, "A, B, or C" or "A, B, and / or C" means "any one of the following: A; B; C; A and B; A and C; B and C; A and B and C." Exceptions to this definition only occur when the combination of elements, functions, steps, or operations is inherently mutually exclusive in some way.

[0030] It should be understood that although the steps in the flowcharts of this application's embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in the figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0031] Depending on the context, the words “if” or “suppose” as used here can be interpreted as “when” or “in response to determination” or “in response to detection.” Similarly, depending on the context, the phrases “if determination” or “if detection (of the stated condition or event)” can be interpreted as “when determination” or “in response to determination” or “when detection (of the stated condition or event)” or “in response to detection (of the stated condition or event).”

[0032] The following describes an email management system provided by an embodiment of this application.

[0033] Please see Figure 1 , Figure 1 This is a schematic diagram of the structure of an email management system provided in an embodiment of this application. For example... Figure 1 As shown, the email management system includes a server 101, a first terminal device 102, and a second terminal device 103. The first terminal device 102 and the second terminal device 103 can communicate with the server 101. The server 101 may include a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. The first terminal device 102 and the second terminal device 103 may include smartphones (such as Android phones, iOS phones, Windows Phones, etc.), desktop computers, tablets, PDAs, laptops, mobile internet devices (MIDs), and wearable devices. The above communication methods, servers, and terminal devices are merely examples and not exhaustive, and include, but are not limited to, the aforementioned communication methods, servers, and terminal devices. Server 101 is used to manage emails to be processed, and may include rejecting or accepting and adding them to the corresponding email list; the first terminal device 102 and the second terminal device 103 can be used to edit emails, send emails, and receive emails, and the content of the emails may include pictures, text, and videos.

[0034] In one feasible embodiment, the number of terminal devices may include multiple devices, and may also include third terminal devices and fourth terminal devices. Terminal devices may send emails to each other using a server, or one terminal device may send the same email to multiple terminal devices. The number of terminal devices may be adjusted according to the actual situation, and this application embodiment does not limit it.

[0035] In this embodiment, when server 101 receives an email to be processed from a sender to a target object, it performs feature extraction processing on the email to be processed to obtain first feature representation information of the email to be processed; it performs feature extraction processing on emails included in a first email set associated with the target object to obtain second feature representation information; it determines the matching degree between the email to be processed and the target object based on the first feature representation information and the second feature representation information, and performs management operations on the email to be processed according to the matching degree; wherein, the management operations include rejecting receipt or adding to the corresponding email list. Through this embodiment, the matching degree between the email to be processed and the target object can be determined based on the feature representation information of emails in the email set associated with the target object and the feature representation information of the email to be processed, and the email to be processed can be managed according to the matching degree. The email to be processed can be managed according to the email set associated with the target object, realizing personalized management of emails to be processed and effectively improving the accuracy of email management.

[0036] The following describes an email management method provided by an embodiment of this application.

[0037] Please see Figure 2 , Figure 2 This is a flowchart illustrating an email management method provided in an embodiment of this application. The email management method provided in this embodiment can be applied to the above-mentioned... Figure 1 The email management system shown below will be illustrated using the example of applying this email management method to the server within the system. Figure 2 As shown, this email management method includes:

[0038] S201. When a pending email is received from the target object, feature extraction processing is performed on the pending email to obtain the first feature representation information of the pending email.

[0039] In this embodiment of the application, when the server receives an email to be processed from the sender to the target object, it can obtain the email to be processed, extract and process the features of the email to be processed, and obtain the first feature representation information of the email to be processed.

[0040] In a feasible embodiment, feature extraction processing is performed on the email to be processed to obtain the first feature representation information of the email to be processed. Specifically, this may include: extracting feature information of the email to be processed; wherein, the feature information includes content feature information and additional feature information, the content feature information includes one or more of text information, image information, audio information and video information, and the additional feature information includes one or more of the following: the number of emails received by the target object, the network feature information of the sender of the email to be processed, and the ratio information of text content to character content; the feature information of the email to be processed is input into a feature extraction model for feature extraction processing to obtain the first feature representation information of the email to be processed.

[0041] Specifically, the server can extract the feature information of the email to be processed and call the feature extraction model to perform feature extraction processing on the feature information of the email to be processed, thereby obtaining the first feature representation information of the email to be processed. The feature information of the email to be processed can include content features and additional features. Content features can include one or more of text information, image information, audio information, and video information. Text information can include text information in the email body and text information in email attachments; image information can be images in the email body and images in email attachments; audio and video information can be obtained from audio files and video files in email attachments. Auxiliary feature information can include the number of emails received by the target object; feature information can include the number and type of emails received by the target object in a recent period; network feature information of the sender of the email to be processed can include the sender's Internet Protocol (IP) address and domain name; and the ratio information of text content to character content can refer to the proportion of English letters or Chinese characters in the email. Calling the feature extraction model to extract the feature information of the email to be processed can obtain the first feature representation information of the email to be processed, effectively improving the accuracy of the first feature representation information.

[0042] In one feasible embodiment, the feature extraction model includes a text feature extraction model, an image feature extraction model, and an auxiliary feature extraction model. The feature information of the email to be processed is input into the feature extraction model for feature extraction processing to obtain the first feature representation information of the email to be processed. Specifically, this may include: inputting the text information of the email to be processed into the text feature extraction model for feature extraction processing to obtain the text feature representation information of the email to be processed; inputting the image information of the email to be processed into the image feature extraction model for feature extraction processing to obtain the image feature representation information of the email to be processed; inputting the audio information, video information, and auxiliary feature information of the email to be processed into the auxiliary feature extraction model for feature extraction processing to obtain the auxiliary feature representation information of the email to be processed; and obtaining the first feature representation information of the email to be processed based on the text feature representation information, image feature representation information, and auxiliary feature representation information of the email to be processed.

[0043] Specifically, the server obtains the feature information of the email to be processed, and can call a text feature extraction model to extract features from the text information included in the feature information to obtain the text feature representation information of the email to be processed; call an image feature extraction model to extract features from the image information included in the feature information to obtain the image feature representation information of the email to be processed; call an audio feature extraction model to extract features from the audio information included in the feature information to obtain the audio feature representation information of the email to be processed; call a video feature extraction model to extract features from the video information included in the feature information to obtain the video feature representation information of the email to be processed; call an auxiliary feature extraction model to extract features from the additional feature information included in the feature information to obtain the auxiliary feature representation information of the email to be processed; and fuse the text feature representation information, image feature representation information, audio feature representation information, video feature representation information, and auxiliary feature representation information to obtain the first feature representation information, which effectively improves the diversity of the first feature representation information and further improves the accuracy of the first feature representation information.

[0044] In a feasible embodiment, each of the above feature extraction models is obtained by training the corresponding initial model based on the corresponding training samples. For example, the image feature extraction model is obtained by training the initial feature extraction model using email images as training samples.

[0045] In one feasible embodiment, the text information of the email to be processed is input into a text feature extraction model for feature extraction processing to obtain the text feature representation information of the email to be processed. Specifically, this may include: performing word segmentation processing on the text information of the email to be processed to obtain at least one text word; inputting each text word in the at least one text word into the text feature extraction model for feature extraction processing to obtain the word segmentation feature representation information of each text word; and fusing the word segmentation feature representation information of the at least one text word to obtain the text feature representation information of the email to be processed.

[0046] Specifically, the extraction of text feature representation information can be achieved by segmenting the text information of the email to be processed into at least one text segment, calling a text feature extraction model to perform feature extraction processing on each text segment in the at least one text segment to obtain the segmentation feature representation information of each segment, and then fusing the segmentation feature representation information of each text segment to obtain the text feature representation information of the email to be processed.

[0047] It should be noted that the collection and processing of relevant data (such as text information and content feature information of emails) in this application should strictly comply with the requirements of relevant laws and regulations, obtain the informed consent or separate consent of the personal information subject, and carry out subsequent data use and processing within the scope of laws and regulations and the authorization of the personal information subject.

[0048] In one feasible embodiment, please refer to Figure 5 , Figure 5 This is a schematic diagram illustrating a process for extracting text feature representation information according to an embodiment of this application. For example... Figure 5 As shown, text feature representation information can include text vectors, and the word segmentation feature representation information for each text word can include word vectors for each text word. Word vectors can be extracted from the email to be processed. The word vector extraction process can include email sampling and word segmentation extraction to obtain word vectors. Word vectors can be calculated from the text words to obtain the corresponding word vectors, and the obtained word vectors can be fused to obtain the text feature vector corresponding to the email to be processed. The word vector extraction can utilize either the CBOW (Continuous Bag-of-Words Model) or Skip-Gram model structure based on the word2vec algorithm.

[0049] Please see Figure 6 , Figure 6This is a schematic diagram of the structure of a text feature extraction model provided in this application embodiment. The CBOW model mainly predicts the current word based on the words in the context within a sliding window, calculated using the weight matrix of the projection layer. By inputting word segmentation 0, word segmentation 1, word segmentation 3, and word segmentation 4 into the CBOW model, the projection layer predicts word segmentation 2 at the current position and outputs word segmentation 2. The Skip-Gram model predicts the words in the context based on the current word, calculated using the weight matrix of the projection layer. By inputting word segmentation 7 into the Skip-Gram model, the projection layer predicts word segmentation 5, word segmentation 6, word segmentation 8, and word segmentation 9 in the context at the current position and outputs word segmentation 5, word segmentation 6, word segmentation 8, and word segmentation 9. Word segmentation 0, word segmentation 1, word segmentation 2, word segmentation 3, and word segmentation 4 can be sequentially combined to form a complete sentence, and word segmentation 5, word segmentation 6, word segmentation 7, word segmentation 8, and word segmentation 9 can also be sequentially combined to form a complete sentence. Since word2vec does not require constructing labeled datasets, it can learn the rich semantic relationships between words in the email context and better fit the expression of emails by expanding the data scale and increasing the diversity of training samples. After training word2vec on email samples and obtaining word vectors, for a piece of email text, the text can be segmented into word vectors, and the feature vector representation of the email text can be obtained by summing each dimension of all word vectors and taking the average. This effectively improves the accuracy of text feature representation information, thereby improving the accuracy of subsequent management of emails using this text feature representation information.

[0050] In a feasible embodiment, the model for extracting content feature information may also include word vector models, LDA (Latent Dirichlet Allocation) models, GPT (Generative Pre-Trained Transformer) large models, and LLM (Large Language Model) language models. Content-based recognition technologies are limited to content recognition and cannot perform personalized recognition; tree classification models based solely on email features are limited to feature recognition and cannot perform personalized recognition. This embodiment of the application can extract both content feature information and additional feature information from emails separately, and combine this with emails from the user's historical activity to construct a preference feature vector. Emails that the user is interested in are allowed to pass, while emails that are not interested are blocked, thereby achieving personalized blocking for each user, effectively reducing the probability of false blocking and missed blocking, and improving user experience.

[0051] S202. Based on the first feature representation information and the second feature representation information of the target object, determine the matching degree between the email to be processed and the target object; wherein, the second feature representation information is obtained by feature extraction processing of emails included in the first email set associated with the target object.

[0052] In this embodiment, the first email set includes at least one email with historical behavior from the target object's email list. Historical behavior may include reporting, reading, deleting, marking, and moving to the spam list. The second feature representation information can be obtained by feature extraction based on at least one historical email of the target object. The second feature representation information can be updated according to a preset period or calculated in real time after receiving the email to be processed. After obtaining the first and second feature representation information, the server can use the first and second feature representation information to determine the matching degree between the email to be processed and the target object. Determining the matching degree between the email to be processed and the target object based on the second feature representation information generated from emails with historical behavior of the target object and the first feature representation information of the email to be processed effectively improves the accuracy of the matching degree between the email to be processed and the target object. Using emails with historical behavior of the target object to extract the second feature representation information makes managing the emails to be processed more in line with the target object's preferences, effectively reducing the probability of emails being mistakenly blocked or misplaced.

[0053] In a feasible embodiment, the various feature representation information mentioned in the embodiments of this application (such as the aforementioned first feature representation information, second feature representation information, text feature representation information, image feature representation information, audio feature representation information, video feature representation information, and auxiliary feature representation information) can specifically be feature vectors with a certain dimension, such as 64-dimensional vectors.

[0054] In one feasible embodiment, please refer to Figure 7 , Figure 7 This is a schematic diagram of a text feature vector extraction process provided in an embodiment of this application, such as... Figure 7As shown, during offline training, the text feature extraction model, taking the extraction of the text feature vector of "Good morning, today's weather is sunny" as an example, can sample new emails to obtain the sampled text "Good morning, today's weather is sunny". Then, it can segment the sampled text "Good morning, today's weather is sunny" into five words: "morning", "good", "today", "weather", and "sunny". Finally, it can extract word vectors from each word, resulting in the word vector for "morning" being (4, 5, 1, 2, -1) and the word vector for "good" being (2, 5, 1, 2, -1). The word vectors for "today" (3, 1, -1, 0, 1), "weather" (7, -1, 0, 15, 2), and "sunny" (0, 4, 7, -3, 2) are given by the text feature extraction model. The model segments the content of the new email into words and extracts word vectors for each segment. A weighted average of these word vectors is then performed, with each word having the same weight, resulting in the text feature vector of the new email (5, 0, 4, 2, -1). By fusing the feature representation information from the text segmentation, the accuracy of the text feature representation information for the new email is effectively improved.

[0055] In one feasible embodiment, the matching degree between the email to be processed and the target object is determined based on the first feature representation information and the second feature representation information of the target object. Specifically, this may include: calculating the similarity between the first feature representation information and the second feature representation information of the target object to obtain the similarity between the first feature representation information and the second feature representation information; and determining the matching degree between the email to be processed and the target object based on the similarity between the first feature representation information and the second feature representation information.

[0056] In this embodiment, the server can calculate the similarity between the first feature representation information and the second feature representation information to obtain their similarity score. Based on this similarity score, the matching degree between the email to be processed and the target object is determined. The similarity calculation between the first and second feature representation information can utilize softmax, logistic regression, XGBoost, or deep learning classification models. Taking feature representation information as a vector as an example, the similarity between the first and second feature representation information can be obtained by calculating vector similarity, effectively improving the accuracy of the similarity calculation.

[0057] Please see Figure 3 , Figure 3 This is a schematic diagram of a process for determining second feature representation information provided in an embodiment of this application, such as... Figure 3 As shown, determining the second feature representation information can specifically include:

[0058] S301. Obtain a first set of emails associated with the target object; wherein the emails in the first set of emails contain historical operation records of the target object, and the historical operation records include operation types.

[0059] In this embodiment, the server can obtain a preset number (e.g., 100) of emails from emails containing historical operation records of the target object, and combine the obtained emails into a first email set.

[0060] In a feasible embodiment, the operation time in the historical operation record of the target object's emails can be within a preset time period (e.g., within a month or a week). Emails with more recent operation times in the historical operation record can be obtained, which effectively improves the accuracy of the second feature representation information.

[0061] S302. Based on the historical operation type of the target object, the emails included in the first email set are cleaned to obtain the second email set.

[0062] In this embodiment, the cleaning process may include deleting emails with erroneous historical operations from the first email set to obtain a second email set for determining the second feature representation information, which effectively improves the accuracy of the second feature representation information determined by the email set.

[0063] In one feasible embodiment, the emails included in the first email set are cleaned based on the historical operation types of the target object to obtain a second email set. Specifically, this may include: obtaining the historical operation types of the target object on each email in the first email set; determining the target emails in the first email set whose historical operation type is one-click operation; and determining the second email set based on the emails in the first email set other than the target emails.

[0064] Specifically, historical operation types can include reading emails, deleting emails, reporting, marking, and one-click operations. Marking can include important marking, and one-click operations can include one-click reading and one-click deleting. The server can obtain the historical operation types of each email from the first email set. One-click operations usually do not reflect the user's preference for each email and are arbitrary. Therefore, emails in the first email set with the historical operation type of one-click operation can be identified and deleted from the first email set to obtain the second email set. This achieves data cleaning of the email set and effectively improves the accuracy of the second feature representation information.

[0065] S303. Extract feature information of the emails included in the second email set; wherein, the feature information includes content feature information and additional feature information, the content feature information includes one or more of text information, image information, audio information and video information, and the additional feature information includes one or more of the following: the number of emails received by the target object, the network feature information of the sender of the email to be processed, and the ratio information of text content to character content.

[0066] Specifically, content feature information may include one or more of text, image, audio, and video information. Additional feature information may include one or more of the following: the number of emails received by the target object, the network characteristics of the sender of the email to be processed, and the ratio of text content to character content. The server can perform feature extraction processing on each email in the second email set to obtain the content feature information and additional feature information for each email. The type of content feature information extracted can be determined based on the content feature information included in the email to be processed. For example, if the content feature information of the email to be processed only includes text and image information, then when extracting content feature information from emails in the second email set, only text and image information will be extracted, achieving consistency in content feature information between the email to be processed and emails in the second email set.

[0067] S304. Input the feature information of the emails included in the second email set into the feature extraction model for feature extraction processing to obtain the second feature representation information of the target object.

[0068] Specifically, the server can call the feature extraction model to perform feature extraction processing on the content feature information and additional feature information of the emails included in the second email set, and obtain the second feature representation information of the target object.

[0069] In one feasible embodiment, the feature information of the emails included in the second email set is input into a feature extraction model for feature extraction processing to obtain the second feature representation information of the target object. Specifically, this may include: inputting the feature information of each email included in the second email set into a feature extraction model for feature extraction processing to obtain the feature representation information of each email; and performing a fusion processing on the feature representation information of each email to obtain the second feature representation information of the target object.

[0070] Specifically, taking feature representation information as feature vectors as an example, the server can input the feature information of each email in the second email set into the feature vector extraction model to obtain the feature vector of each email. The server then adds the feature vectors of the same dimensions from each email and takes the average to obtain the second feature vector of the target object (i.e., the second feature representation information). By using the email set associated with the target object to obtain the second feature representation information of the target object, different feature representation information can be obtained for different target objects, effectively improving the accuracy of the feature representation information.

[0071] S203. Perform management operations on the emails to be processed according to the matching degree. The management operations include rejecting the emails or adding them to the corresponding email list.

[0072] In this embodiment, rejecting an email can mean blocking it, and adding it to the corresponding email list can mean placing it in the inbox's corresponding email list or in the spam / junk mail list. Managing emails based on their match with the target audience allows for personalized management of the same emails for different target audiences, effectively improving the accuracy of email management.

[0073] In one feasible embodiment, please refer to Figure 4 , Figure 4 This is a flowchart illustrating another email management method provided in an embodiment of this application, such as... Figure 4As shown, the email management method includes: extracting features from emails to be processed to obtain content features and additional features; processing the content features and additional features of the emails to be processed to obtain text feature representation information, image feature representation information, and other feature representation information; fusing the text feature representation information, image feature representation information, and other feature representation information to obtain the fused feature representation information corresponding to the emails to be processed; cleaning the first email set to obtain a second email set; extracting features from the second email set to obtain the content features and additional features of the emails in the second email set; and processing the content features and additional features of the emails in the second email set to obtain the fused feature representation information corresponding to the emails to be processed. Feature information processing is performed on the content features and additional features to obtain text feature representation information, image feature representation information, and other feature representation information (such as the aforementioned additional feature representation information). The text feature representation information, image feature representation information, and other feature representation information are then fused to obtain the feature representation information corresponding to the emails in the fused second email set. The matching degree of the feature representation information corresponding to the emails in the second email set and the feature representation information corresponding to the emails to be processed is calculated to obtain the matching degree. Based on the matching degree, email management operations are performed on the emails to be processed, realizing the management of emails to be processed according to the email set of the target object, which effectively improves the accuracy of email management.

[0074] In one feasible embodiment, management operations are performed on the emails to be processed based on the matching degree. The management operations include rejecting receipt or adding to the corresponding email list. This includes: if the matching degree is less than or equal to a first matching degree threshold, the management operation for the emails to be processed is to reject receipt; if the matching degree is greater than the first matching degree threshold, the management operation for the emails to be processed is to add to the corresponding email list; and the management operation is performed on the emails to be processed.

[0075] Specifically, a matching threshold can be set, and emails to be processed can be managed based on the relationship between the matching degree of the email to be processed and the target object and the matching threshold. The server can reject emails to be processed when the matching degree is less than or equal to the threshold; and add emails to the email list when the matching degree is greater than the threshold. This achieves personalized email management for each individual and effectively improves the accuracy of email management.

[0076] In one feasible embodiment, determining that the management operation for the email to be processed is to add it to the corresponding email list includes: if the matching degree is within a first matching degree range, then determining that the management operation for the email to be processed is to add it to the email list corresponding to the spam folder; if the matching degree is within a second matching degree range, then determining that the management operation for the email to be processed is to add it to the email list corresponding to the inbox; wherein, the minimum value in the second matching degree range is greater than the maximum value in the first matching degree range. For example, the first matching degree threshold is 60%, the first matching degree range can be 60%-75%, and the second matching degree range can be 76%-100%.

[0077] Specifically, for emails added to corresponding email lists, the type of email list to which an email to be processed is placed can be determined by setting a matching score range. The range exceeding the matching score threshold can be divided into multiple sub-ranges, with no overlap between them. Each sub-range can correspond to a specific email list type. For example, if the email list types include inbox and spam, the range exceeding the matching score threshold can be divided into two sub-ranges. Emails to be processed can then be placed in the email list corresponding to the matching score sub-range. This categorization of email lists to which emails to be processed are added makes the management of emails to be processed more precise and effectively improves the accuracy of email management.

[0078] In a feasible embodiment, the present application embodiment may further include an email classification model, which can be used to classify emails to be processed. Specifically, it may include: inputting the first feature representation information into the email classification model for classification processing to obtain the classification result of the emails to be processed; if the classification result of the emails to be processed is to allow the emails, then the step of determining the management operation of the emails to be processed is to add them to the corresponding email list is executed; if the classification result of the emails to be processed is to reject the emails, then the management operation of the emails to be processed is to refuse to receive them is determined.

[0079] Specifically, the email classification model is trained by using the feature representation information of emails as training samples. After obtaining the matching degree between the email to be processed and the target object, if the matching degree is greater than the first matching degree threshold, the server can further classify the email to be processed using the email classification model to obtain the email type of the email to be processed in order to reduce the misjudgment or false positive of the email. Based on the matching degree and the email type result, the email to be processed is managed, which effectively reduces the probability of misjudgment or false positive of the email and further improves the accuracy of email management.

[0080] In this embodiment, when a target object receives an email to be processed, the server performs feature extraction processing on the email to be processed to obtain first feature representation information of the email to be processed; performs feature extraction processing on emails included in a first email set associated with the target object to obtain second feature representation information; determines the matching degree between the email to be processed and the target object based on the first feature representation information and the second feature representation information, and performs management operations on the email to be processed according to the matching degree; wherein, the management operations include rejecting receipt or adding to the corresponding email list. Through this embodiment, the matching degree between the email to be processed and the target object can be determined based on the feature representation information of emails in the email set associated with the target object and the feature representation information of the email to be processed, and the email to be processed can be managed according to the matching degree. This allows for personalized management of emails to be processed, effectively improving the accuracy of email management.

[0081] The following describes an email management device provided by an embodiment of this application.

[0082] Please see Figure 8 , Figure 8 This is a schematic diagram of the structure of an email management device provided in an embodiment of this application. Figure 8 As shown, the email management device includes:

[0083] Processing module 801 is used to perform feature extraction processing on the email to be processed when it receives the email to be processed from the target object, and obtain the first feature representation information of the email to be processed.

[0084] The determining module 802 is used to determine the matching degree between the email to be processed and the target object based on the first feature representation information and the second feature representation information of the target object; wherein, the second feature representation information is obtained by feature extraction processing of emails included in the first email set associated with the target object;

[0085] The processing module 801 is further configured to perform management operations on the emails to be processed based on the matching degree, the management operations including rejecting receipt or adding to the corresponding email list.

[0086] In a feasible embodiment, when the determining module 802 determines the matching degree between the email to be processed and the target object based on the first feature representation information and the second feature representation information of the target object, it is specifically used to: perform similarity calculation on the first feature representation information and the second feature representation information of the target object to obtain the similarity between the first feature representation information and the second feature representation information; and determine the matching degree between the email to be processed and the target object based on the similarity between the first feature representation information and the second feature representation information.

[0087] In a feasible embodiment, when the processing module 801 performs feature extraction processing on the email to be processed to obtain the first feature representation information of the email to be processed, it is specifically used to: extract feature information of the email to be processed; wherein, the feature information includes content feature information and additional feature information, the content feature information includes one or more of text information, image information, audio information and video information, and the additional feature information includes one or more of the following: the number of emails received by the target object, the network feature information of the sender of the email to be processed, and the ratio information of text content to character content; input the feature information of the email to be processed into a feature extraction model for feature extraction processing to obtain the first feature representation information of the email to be processed.

[0088] In a feasible embodiment, the feature extraction model includes a text feature extraction model, an image feature extraction model, and an auxiliary feature extraction model. When the processing module 801 inputs the feature information of the email to be processed into the feature extraction model for feature extraction processing to obtain the first feature representation information of the email to be processed, it is specifically used for: inputting the text information of the email to be processed into the text feature extraction model for feature extraction processing to obtain the text feature representation information of the email to be processed; inputting the image information of the email to be processed into the image feature extraction model for feature extraction processing to obtain the image feature representation information of the email to be processed; inputting one or more of the audio information, video information, and additional feature information of the email to be processed into the auxiliary feature extraction model for feature extraction processing to obtain the auxiliary feature representation information of the email to be processed; and performing fusion processing on the text feature representation information, image feature representation information, and auxiliary feature representation information of the email to be processed to obtain the first feature representation information of the email to be processed.

[0089] In a feasible embodiment, when the processing module 801 inputs the text information of the email to be processed into the text feature extraction model for feature extraction processing to obtain the text feature representation information of the email to be processed, it is specifically used for: performing word segmentation processing on the text information of the email to be processed to obtain at least one text word; inputting each text word in the at least one text word into the text feature extraction model for feature extraction processing to obtain the word segmentation feature representation information of each text word; and performing fusion processing on the word segmentation feature representation information of the at least one text word to obtain the text feature representation information of the email to be processed.

[0090] In a feasible embodiment, the processing module 801 is further configured to: acquire a first email set associated with a target object; wherein the emails in the first email set contain historical operation records of the target object, and the historical operation records include operation types; perform cleaning processing on the emails in the first email set based on the historical operation types of the target object to obtain a second email set; extract feature information of the emails in the second email set; wherein the feature information includes content feature information and additional feature information, the content feature information includes one or more of text information, image information, audio information, and video information, and the additional feature information includes one or more of the following: the number of emails received by the target object, the network feature information of the sender of the email to be processed, and the ratio information of text content to character content; input the feature information of the emails in the second email set into a feature extraction model for feature extraction processing to obtain second feature representation information of the target object.

[0091] In a feasible embodiment, when the processing module 801 cleans the emails included in the first email set based on the historical operation type of the target object to obtain the second email set, it is specifically used to: obtain the historical operation type of the target object on each email in the first email set; determine the target email in the first email set whose historical operation type is a one-click operation; and determine the second email set based on the emails in the first email set other than the target email.

[0092] In a feasible embodiment, when the processing module 801 inputs the feature information of the emails included in the second email set into the feature extraction model for feature extraction processing to obtain the second feature representation information of the target object, it is specifically used to: input the feature information of each email included in the second email set into the feature extraction model for feature extraction processing to obtain the feature representation information of each email; and perform fusion processing on the feature representation information of each email to obtain the second feature representation information of the target object.

[0093] In a feasible embodiment, the processing module 801 performs management operations on the email to be processed based on the matching degree. The management operations include rejecting receipt or adding to the corresponding email list. Specifically, it is used to: if the matching degree is less than or equal to a first matching degree threshold, determine that the management operation for the email to be processed is to reject receipt; if the matching degree is greater than the first matching degree threshold, determine that the management operation for the email to be processed is to add to the corresponding email list; and execute the management operation on the email to be processed.

[0094] In a feasible embodiment, when the processing module 801 determines that the management operation of the email to be processed is to add it to the corresponding email list, it is specifically configured to: if the matching degree is in a first matching degree range, determine that the management operation of the email to be processed is to add it to the email list corresponding to the spam folder; if the matching degree is in a second matching degree range, determine that the management operation of the email to be processed is to add it to the email list corresponding to the inbox; wherein, the minimum value in the second matching degree range is greater than the maximum value in the first matching degree range.

[0095] In a feasible embodiment, the processing module 801 is further configured to: input the first feature representation information into the email classification model for classification processing to obtain the classification result of the email to be processed; if the classification result of the email to be processed is an approved email, then perform the step of determining that the management operation of the email to be processed is to add it to the corresponding email list; if the classification result of the email to be processed is a rejected email, then determine that the management operation of the email to be processed is to refuse to receive it.

[0096] In a feasible embodiment, the email management device provided in this application can be implemented in software. The email management device can be stored in a memory and can be software in the form of programs and plug-ins, and includes a series of modules, including a processing module and a determination module; wherein, the processing module and the determination module are used to implement the email management method provided in this application.

[0097] In other feasible embodiments, the email management device provided in this application can also be implemented in a combination of hardware and software. As an example, the email management device provided in this application can be a processor in the form of a hardware decoding processor, which is programmed to execute the email management method provided in this application. For example, the processor in the form of a hardware decoding processor can be one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), or other electronic components.

[0098] The following describes a server provided by an embodiment of this application.

[0099] Please see Figure 9 , Figure 9 This is a schematic diagram of the structure of a server provided in an embodiment of this application. Figure 9 The server in this embodiment may include one or more processors 901, one or more communication interfaces 902, and a memory 903. The processors 901, communication interfaces 902, and memory 903 are connected via a bus 904. The memory 903 stores computer programs, including program instructions, and the processors 901 execute the program instructions stored in the memory 903. By running the executable program code in the memory 903, the processor 901 performs the following operations:

[0100] When a pending email is received from the target object, feature extraction processing is performed on the pending email to obtain the first feature representation information of the pending email;

[0101] Based on the first feature representation information and the second feature representation information of the target object, the matching degree between the email to be processed and the target object is determined; wherein, the second feature representation information is obtained by feature extraction processing of emails included in the first email set associated with the target object;

[0102] The pending emails are managed based on the matching degree, and the management operations include rejecting them or adding them to the corresponding email list.

[0103] In a feasible embodiment, when the processor 901 determines the matching degree between the email to be processed and the target object based on the first feature representation information and the second feature representation information of the target object, it is specifically configured to: perform similarity calculation on the first feature representation information and the second feature representation information of the target object to obtain the similarity between the first feature representation information and the second feature representation information; and determine the matching degree between the email to be processed and the target object based on the similarity between the first feature representation information and the second feature representation information.

[0104] In a feasible embodiment, when the processor 901 performs feature extraction processing on the email to be processed to obtain the first feature representation information of the email to be processed, it is specifically used to: extract feature information of the email to be processed; wherein, the feature information includes content feature information and additional feature information, the content feature information includes one or more of text information, image information, audio information and video information, and the additional feature information includes one or more of the following: the number of emails received by the target object, the network feature information of the sender of the email to be processed, and the ratio information of text content to character content; input the feature information of the email to be processed into a feature extraction model for feature extraction processing to obtain the first feature representation information of the email to be processed.

[0105] In a feasible embodiment, the feature extraction model includes a text feature extraction model, an image feature extraction model, and an auxiliary feature extraction model. When the processor 901 inputs the feature information of the email to be processed into the feature extraction model for feature extraction processing to obtain the first feature representation information of the email to be processed, it specifically performs the following steps: inputting the text information of the email to be processed into the text feature extraction model for feature extraction processing to obtain the text feature representation information of the email to be processed; inputting the image information of the email to be processed into the image feature extraction model for feature extraction processing to obtain the image feature representation information of the email to be processed; inputting one or more of the audio information, video information, and additional feature information of the email to be processed into the auxiliary feature extraction model for feature extraction processing to obtain the auxiliary feature representation information of the email to be processed; and performing fusion processing on the text feature representation information, image feature representation information, and auxiliary feature representation information of the email to be processed to obtain the first feature representation information of the email to be processed.

[0106] In a feasible embodiment, when the processor 901 inputs the text information of the email to be processed into the text feature extraction model for feature extraction processing to obtain the text feature representation information of the email to be processed, it specifically performs the following: performs word segmentation processing on the text information of the email to be processed to obtain at least one text word; inputs each of the at least one text word into the text feature extraction model for feature extraction processing to obtain the word segmentation feature representation information of each text word; and performs fusion processing on the word segmentation feature representation information of the at least one text word to obtain the text feature representation information of the email to be processed.

[0107] In a feasible embodiment, the processor 901 is further configured to: acquire a first email set associated with a target object; wherein the emails in the first email set contain historical operation records of the target object, and the historical operation records include operation types; perform cleaning processing on the emails in the first email set based on the historical operation types of the target object to obtain a second email set; extract feature information of the emails in the second email set; wherein the feature information includes content feature information and additional feature information, the content feature information includes one or more of text information, image information, audio information, and video information, and the additional feature information includes one or more of the following: the number of emails received by the target object, the network feature information of the sender of the email to be processed, and the ratio information of text content to character content; input the feature information of the emails in the second email set into a feature extraction model for feature extraction processing to obtain second feature representation information of the target object.

[0108] In a feasible embodiment, when the processor 901 cleans the emails included in the first email set based on the historical operation type of the target object to obtain the second email set, it is specifically used to: obtain the historical operation type of the target object on each email in the first email set; determine the target email in the first email set whose historical operation type is a one-click operation; and determine the second email set based on the emails in the first email set other than the target email.

[0109] In a feasible embodiment, when the processor 901 inputs the feature information of the emails included in the second email set into a feature extraction model for feature extraction processing to obtain the second feature representation information of the target object, it is specifically used to: input the feature information of each email included in the second email set into the feature extraction model for feature extraction processing to obtain the feature representation information of each email; and perform fusion processing on the feature representation information of each email to obtain the second feature representation information of the target object.

[0110] In a feasible embodiment, the processor 901 performs management operations on the email to be processed based on the matching degree. The management operations include rejecting reception or adding to the corresponding email list. Specifically, it is used to: if the matching degree is less than or equal to a first matching degree threshold, determine that the management operation for the email to be processed is to reject reception; if the matching degree is greater than the first matching degree threshold, determine that the management operation for the email to be processed is to add to the corresponding email list; and execute the management operation on the email to be processed.

[0111] In a feasible embodiment, when the processor 901 determines that the management operation of the email to be processed is to add it to the corresponding email list, it is specifically configured to: if the matching degree is in a first matching degree range, determine that the management operation of the email to be processed is to add it to the email list corresponding to the spam folder; if the matching degree is in a second matching degree range, determine that the management operation of the email to be processed is to add it to the email list corresponding to the inbox; wherein, the minimum value in the second matching degree range is greater than the maximum value in the first matching degree range.

[0112] In a feasible embodiment, the processor 901 is further configured to: input the first feature representation information into an email classification model for classification processing to obtain the classification result of the email to be processed; if the classification result of the email to be processed is an approved email, then execute the step of determining that the management operation of the email to be processed is to add it to the corresponding email list; if the classification result of the email to be processed is a rejected email, then determine that the management operation of the email to be processed is to refuse to receive it.

[0113] The method steps in the embodiments of this application can be adjusted, combined, or deleted according to actual needs.

[0114] The units in the embodiments of this application can be merged, divided, and deleted according to actual needs.

[0115] In this application embodiment, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0116] In this application, the same or similar terms, concepts, technical solutions and / or application scenario descriptions are generally described in detail only when they appear for the first time. When they appear again, they are generally not repeated for the sake of brevity. When understanding the technical solutions and other contents of this application, the same or similar terms, concepts, technical solutions and / or application scenario descriptions that are not described in detail later can be referred to their previous relevant detailed descriptions.

[0117] In this application, the descriptions of the various embodiments have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0118] The technical features of the present application can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of the present application.

[0119] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, storage disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0120] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. An email management method, characterized in that, The method includes: When a pending email is received from the target object, feature extraction processing is performed on the pending email to obtain the first feature representation information of the pending email; Based on the first feature representation information and the second feature representation information of the target object, the matching degree between the email to be processed and the target object is determined; wherein, the second feature representation information is obtained by feature extraction processing of emails included in the first email set associated with the target object; The pending emails are managed based on the matching degree, and the management operations include rejecting them or adding them to the corresponding email list.

2. The method according to claim 1, characterized in that, The step of determining the matching degree between the email to be processed and the target object based on the first feature representation information and the second feature representation information of the target object includes: The similarity between the first feature representation information and the second feature representation information of the target object is calculated to obtain the similarity between the first feature representation information and the second feature representation information; Based on the similarity between the first feature representation information and the second feature representation information, the matching degree between the email to be processed and the target object is determined.

3. The method according to claim 1, characterized in that, The step of performing feature extraction processing on the email to be processed to obtain the first feature representation information of the email to be processed includes: Extract the feature information of the email to be processed; wherein, the feature information includes content feature information and additional feature information, the content feature information includes one or more of text information, image information, audio information and video information, and the additional feature information includes one or more of the following: the number of emails received by the target object, the network feature information of the sender of the email to be processed, and the ratio information of text content to character content; The feature information of the email to be processed is input into the feature extraction model for feature extraction processing to obtain the first feature representation information of the email to be processed.

4. The method according to claim 3, characterized in that, The feature extraction model includes a text feature extraction model, an image feature extraction model, and an auxiliary feature extraction model; the step of inputting the feature information of the email to be processed into the feature extraction model for feature extraction processing to obtain the first feature representation information of the email to be processed includes: The text information of the email to be processed is input into the text feature extraction model for feature extraction processing to obtain the text feature representation information of the email to be processed. The image information of the email to be processed is input into the image feature extraction model for feature extraction processing to obtain the image feature representation information of the email to be processed. The auxiliary feature extraction model is input into one or more of the audio information, video information and additional feature information of the email to be processed to perform feature extraction processing, thereby obtaining the auxiliary feature representation information of the email to be processed. The text feature representation information, image feature representation information, and auxiliary feature representation information of the email to be processed are fused to obtain the first feature representation information of the email to be processed.

5. The method according to claim 4, characterized in that, The step of inputting the text information of the email to be processed into the text feature extraction model for feature extraction processing to obtain the text feature representation information of the email to be processed includes: The text information of the email to be processed is segmented into words to obtain at least one text segment; Each text segment in the at least one text segmentation is input into a text feature extraction model for feature extraction processing to obtain the segmentation feature representation information of each text segment; The word segmentation feature representation information of the at least one text segmentation is fused to obtain the text feature representation information of the email to be processed.

6. The method according to any one of claims 1-5, characterized in that, The method further includes: Obtain a first set of emails associated with a target object; wherein the emails in the first set of emails contain historical operation records of the target object, and the historical operation records include operation types; Based on the historical operation types of the target object, the emails included in the first email set are cleaned to obtain the second email set; Extract feature information of the emails included in the second email set; wherein, the feature information includes content feature information and additional feature information, the content feature information includes one or more of text information, image information, audio information and video information, and the additional feature information includes one or more of the following: the number of emails received by the target object, the network feature information of the sender of the email to be processed, and the ratio information of text content to character content; The feature information of the emails included in the second email set is input into the feature extraction model for feature extraction processing to obtain the second feature representation information of the target object.

7. The method according to claim 6, characterized in that, The process of cleaning the emails in the first email set based on the historical operation types of the target object yields a second email set, including: Obtain the historical operation types of the target object for each email in the first email set; Identify the target emails in the first email set whose historical operation type is one-click operation; A second email set is determined based on emails in the first email set other than the target email.

8. The method according to claim 6, characterized in that, The step of inputting the feature information of the emails included in the second email set into a feature extraction model for feature extraction processing to obtain the second feature representation information of the target object includes: The feature information of each email in the second email set is input into the feature extraction model for feature extraction processing to obtain the feature representation information of each email. The feature representation information of each email is fused to obtain the second feature representation information of the target object.

9. The method according to claim 1, characterized in that, The step of managing the pending emails based on the matching degree includes rejecting them or adding them to the corresponding email list, including: If the matching degree is less than or equal to the first matching degree threshold, then the management operation for the email to be processed is to refuse to receive it. If the matching degree is greater than the first matching degree threshold, then the management operation for the email to be processed is to add it to the corresponding email list. Perform the management operation on the email to be processed.

10. The method according to claim 9, characterized in that, The step of determining that the management operation for the email to be processed is to add it to the corresponding email list includes: If the matching degree is within the first matching degree range, then the management operation for the email to be processed is to add it to the email list corresponding to the spam folder. If the matching degree is within the second matching degree range, then the management operation for the email to be processed is determined to be adding it to the email list corresponding to the inbox; wherein, the minimum value in the second matching degree range is greater than the maximum value in the first matching degree range.

11. The method according to claim 9, characterized in that, The method further includes: The first feature representation information is input into the email classification model for classification processing to obtain the classification result of the email to be processed; If the classification result of the email to be processed is an email to be allowed, then the step of determining the management operation of the email to be processed is to add it to the corresponding email list is executed; If the classification result of the email to be processed is rejected, then the management operation of the email to be processed is determined to be rejected.

12. An email management device, characterized in that, The device includes: The processing module is used to perform feature extraction processing on the email to be processed when it receives the email to be processed from the target object, so as to obtain the first feature representation information of the email to be processed. The determining module is used to determine the matching degree between the email to be processed and the target object based on the first feature representation information and the second feature representation information of the target object; wherein, the second feature representation information is obtained by feature extraction processing of emails included in the first email set associated with the target object; The processing module is also used to perform management operations on the emails to be processed based on the matching degree, including rejecting them or adding them to the corresponding email list.

13. A server, characterized in that, include: The system includes a processor, a communication interface, and a memory, which are interconnected. The memory stores executable program code, and the processor is used to call the executable program code to implement the email management method as described in any one of claims 1-11.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed on a computer, cause the computer to implement the email management method as described in any one of claims 1-11.

15. A computer program product, characterized in that, The computer program product includes a computer program or computer instructions, which, when executed by a processor, are used to implement the email management method as described in any one of claims 1-11.

Images