Short message management method and device, electronic equipment and storage medium

Abstract

The present disclosure provides a short message management method and device, electronic equipment and storage medium, and relates to the technical field of artificial intelligence. The method comprises: acquiring a short message text; determining a first attribution type corresponding to the short message text; determining a time range corresponding to the first attribution type; if there is a generated image within the time range, determining a second attribution type corresponding to the image; and generating a schedule event based on the short message text and the image according to the first attribution type and the second attribution type. Thus, the schedule event can be generated according to the short message text in combination with the corresponding image, the intelligent management of the short message can be performed, the personalized needs of the user can be met, and the user experience can be improved.

Description

SMS management methods, devices, electronic equipment and storage media Technical Field

[0001] This disclosure relates to the field of artificial intelligence technology, and in particular to a method, apparatus, electronic device and storage medium for SMS management. Background Technology

[0002] SMS, an indispensable service for users, boasts advantages such as no data consumption, fast message reception, and real-time performance. However, current technologies that manage SMS messages solely based on their text content offer only simplistic management methods and fail to meet users' personalized needs, a problem that urgently requires resolution. Summary of the Invention

[0003] This disclosure provides a method, apparatus, electronic device, and storage medium for managing text messages.

[0004] In a first aspect, this disclosure provides a text message management method, comprising: acquiring text message text; determining a first attribution type corresponding to the text message text; determining a time range corresponding to the first attribution type; if an image is generated within the time range, determining a second attribution type corresponding to the image; and generating a schedule event based on the text message text and the image according to the first attribution type and the second attribution type.

[0005] In some embodiments, determining the first category corresponding to the SMS text includes: inputting the SMS text into a text processing model to generate a text semantic feature vector; and performing feature vector matching between the text semantic feature vector and a preset text type feature vector to determine the first category corresponding to the SMS text.

[0006] In some embodiments, determining the second attribution type corresponding to an image includes: inputting the image into an image processing model to generate an image semantic feature vector; and performing feature vector matching between the image semantic feature vector and a preset image type feature vector to determine the second attribution type corresponding to the image.

[0007] In some embodiments, generating a schedule event based on SMS text and an image includes: inputting the SMS text into a text processing model to generate a text semantic feature vector; inputting the image into an image processing model to generate an image semantic feature vector; inputting the text semantic feature vector and the image semantic feature vector into an event generation model to generate an event text description; and generating a schedule event based on the SMS text's receiving time, the image's generation time, the time range, the image, and the event text description.

[0008] In some embodiments, the method further includes: if no image is generated within the time range, generating a schedule event based on the SMS text.

[0009] In some embodiments, the method further includes: determining a historical schedule event that belongs to the same event as the schedule event; obtaining address information corresponding to the schedule event and the historical schedule event respectively; and associating the schedule event with the historical schedule event if the address information is the same.

[0010] In some embodiments, determining a historical event that belongs to the same event as the scheduled event includes: obtaining a first sending entity corresponding to a text message; determining a second sending entity and a third attribution type corresponding to a candidate historical text message; and performing a similarity calculation based on the first sending entity and the first attribution type of the text message, and the second sending entity and the third attribution type of the candidate historical text message, to determine the historical event that belongs to the same event as the scheduled event.

[0011] In some embodiments, the method further includes: generating a schedule record based on the schedule events corresponding to SMS texts obtained during a specific time period; obtaining historical schedule records; and determining recommendation information based on the historical schedule records and the schedule record.

[0012] In some embodiments, determining recommended information based on historical calendar entries and calendar entries includes: determining a target calendar entry that matches the calendar entry in the historical calendar entries; and determining recommended information based on the target calendar entry and calendar entries.

[0013] Secondly, this disclosure provides a text message management device, comprising: a text acquisition unit for acquiring text messages; a first type determination unit for determining a first attribution type corresponding to the text messages; a time determination unit for determining a time range corresponding to the first attribution type; a second type determination unit for determining a second attribution type corresponding to an image if an image is generated within the time range; and an event generation unit for generating a schedule event based on the text messages and the image according to the first attribution type and the second attribution type.

[0014] Thirdly, this disclosure provides an electronic device, including: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the SMS management method disclosed in the embodiments of this disclosure.

[0015] Fourthly, this disclosure provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the SMS management method disclosed in the embodiments of this disclosure.

[0016] Fifthly, this disclosure provides a computer program product, including a computer program that, when executed by a processor, implements the SMS management method disclosed in the embodiments of this disclosure.

[0017] The technical solutions provided by the embodiments of this disclosure have at least the following beneficial effects:

[0018] In this embodiment of the disclosure, a schedule event can be generated based on the SMS text and combined with the corresponding image to perform intelligent SMS management, meet users' personalized needs, and improve user experience. Attached Figure Description

[0019] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0020] Figure 1 is a flowchart illustrating a text message management method according to an exemplary embodiment;

[0021] Figure 2 is a schematic diagram of a text message management system according to an exemplary embodiment;

[0022] Figure 3 is a schematic diagram of another SMS management system according to another exemplary embodiment;

[0023] Figure 4 is a structural diagram of a text message management device according to an exemplary embodiment;

[0024] Figure 5 is a structural diagram of a text message management device according to another exemplary embodiment;

[0025] Figure 6 is a structural diagram of a text message management device according to another exemplary embodiment;

[0026] Figure 7 is a schematic diagram of the structure of the electronic device provided in the embodiments of this disclosure.

[0027] The accompanying drawings have illustrated specific embodiments of this disclosure, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concepts of this disclosure to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0028] 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 numerals 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 disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0029] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.) and signals involved in the embodiments of this disclosure are all authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0030] The technical solutions of this disclosure and how they solve the aforementioned technical problems will be described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of this disclosure will now be described with reference to the accompanying drawings.

[0031] Among related technologies, SMS management methods are relatively simple and cannot meet users' personalized needs.

[0032] Based on this, embodiments of this disclosure provide a text message management method, apparatus, electronic device, and storage medium. The method includes: acquiring text message text; determining a first attribution type corresponding to the text message text; determining a time range corresponding to the first attribution type; if an image exists within the time range, determining a second attribution type corresponding to the image; and generating a schedule event based on the text message text and the image according to the first and second attribution types. Thus, schedule events can be generated based on the text message text and corresponding images, enabling intelligent management of text messages, meeting personalized user needs, and improving user experience.

[0033] Figure 1 is a flowchart illustrating a text message management method according to an exemplary embodiment.

[0034] As shown in Figure 1, the SMS management method provided in this embodiment includes, but is not limited to, the following steps:

[0035] S1, retrieve SMS text.

[0036] In this embodiment of the disclosure, the SMS management method is executed by an SMS management device, which can be an electronic device, such as a mobile phone, computer, or tablet.

[0037] In this embodiment of the disclosure, the electronic device can execute the SMS management method by installing a specific application. For example, if the electronic device is a mobile phone, a specific application can be installed on the mobile phone to execute the SMS management method.

[0038] In this embodiment of the disclosure, the SMS text can be obtained at the same time as the SMS is received, or it can be obtained at a specific time from SMS messages received within a specific time range before, or it can be obtained from SMS text imported by the user. This embodiment of the disclosure does not impose specific limitations on this.

[0039] In some embodiments, the SMS message is received and the SMS text is retrieved after obtaining the user's authorization.

[0040] In some embodiments, the obtained SMS text is the text of a single SMS message or multiple SMS messages.

[0041] Understandably, there are various types of text messages: verification messages, such as verification codes; time-sensitive messages, such as those from hospitals and government agencies; and messages from multiple service providers, such as courier and food delivery information messages, etc.

[0042] In some embodiments, the obtained SMS text includes at least one of the following: the sender of the SMS, the time of receipt, and the content of the SMS.

[0043] It is understandable that the time interval required for sending and receiving text messages is very short. The time to receive a message can also be called the sending time, which is the time it takes to receive the text message.

[0044] S2, determine the first attribution type corresponding to the SMS text.

[0045] In this embodiment of the disclosure, after obtaining the SMS text, the first attribution type corresponding to the SMS text can be determined.

[0046] In some embodiments, the SMS text includes SMS content, and then a first attribution type corresponding to the SMS text is determined based on the SMS content.

[0047] For example, the first category can be appointment, notification, verification, or event. For instance: if the SMS message contains a date and address, such as a museum reservation, hospital appointment, or household registration appointment, then the SMS message is an appointment; if the SMS message contains content related to "food delivery" or "express delivery," then the SMS message is a notification; if the SMS message contains a verification code, then the SMS message is a verification; if the SMS message contains information such as bills (consumption, water, electricity, gas, telecommunications, etc.), marketing activities (credit cards, online stores, brand advertisements, promotional information, etc.), or itineraries (accommodation, flights, trains, etc.), then the SMS message is an event.

[0048] For example, the first attribution type can be related to what the user does, such as food delivery, dining out, hospital reservations, museum reservations, tourist attractions, hotel accommodations, flight itineraries, train journeys, shopping mall shopping, and so on.

[0049] In some embodiments, determining the first category corresponding to the SMS text includes: inputting the SMS text into a text processing model to generate a text semantic feature vector; and performing feature vector matching between the text semantic feature vector and a preset text type feature vector to determine the first category corresponding to the SMS text.

[0050] In this embodiment of the disclosure, after obtaining the SMS text, the SMS text can be input into a text processing model to generate a text semantic feature vector. Then, the text semantic feature vector is matched with a preset text type feature vector to determine the first type to which the SMS text belongs.

[0051] In some embodiments, the text processing model is a large language model (e.g., GPT (Generative Pre-trained Transformer) or LLM (Large Language Model)).

[0052] In some embodiments, an initial text processing model is trained in advance using a text message dataset to obtain the text processing model.

[0053] In some embodiments, the SMS text dataset includes sample SMS texts and their corresponding sample text speech feature vectors and sample attribution types.

[0054] S3, determine the time range corresponding to the first attribution type.

[0055] In this embodiment of the disclosure, when the first attribution type corresponding to the SMS text is determined, the time range corresponding to the first attribution type can be determined.

[0056] In some embodiments, the time range is related to the time when the SMS text message is received.

[0057] For example, if the first attribution type of the SMS text is determined to be a notification, the time range is determined to be one hour before and after the time the message is received.

[0058] It is understandable that different time ranges may correspond to different primary attribution types.

[0059] For example, if the first attribution type is a reservation type, the corresponding time range is the date indicated in the SMS message; if the first attribution type is a notification type, the corresponding time range is one hour before and after the message receiving time; if the first attribution type is a verification type, the corresponding time range is half an hour before and after the message receiving time; if the first attribution type is an event type, the corresponding time range is half an hour before and after the message receiving time.

[0060] It should be noted that the above examples are for illustrative purposes only and are not intended to limit the specific embodiments of this disclosure. The time range corresponding to the first attribution type may also be other times.

[0061] S4. If an image exists within the time frame, determine the second attribution type corresponding to the image.

[0062] In this embodiment of the disclosure, after determining the first attribution type corresponding to the SMS text and the time range corresponding to the first attribution type, it is determined whether an image is generated within the time range.

[0063] In this embodiment of the disclosure, the generated images include: images taken by the user using the terminal camera and stored in the user terminal, images captured by the user and stored in the user terminal, images shared by other users and stored in the user terminal, etc.

[0064] It is understandable that the image generation time is the time it is stored in the user terminal. Determining whether an image exists within a given time frame means determining whether the image's generation time falls within that time frame.

[0065] It is also understandable to determine the primary category of the SMS text and its time range, and to obtain images generated within that time range that may be related to the SMS text.

[0066] For example, when a user receives a food delivery text message, they may take a screenshot while selecting the food and discuss with other users whether it tastes good, or they may take a picture after receiving the food and share it with other users. The generation time of images that may be related to the food delivery text message is within about one hour before and after the time the text message is received. Based on this, in this embodiment of the disclosure, after determining the first attribution type corresponding to the text message, the corresponding time range can be determined, and images that may be related to the text message within the corresponding time range can be obtained.

[0067] In this embodiment of the disclosure, after obtaining an image that may be related to the SMS text, a second attribution type corresponding to the image is determined.

[0068] For example, the second attribution type can be an appointment class, a notification class, a verification class, or an event class.

[0069] For example, the second attribution type can be related to what the user does, such as food delivery, dining out, hospital reservations, museum reservations, tourist attractions, hotel accommodations, flight itineraries, train journeys, shopping mall shopping, and so on.

[0070] In some embodiments, determining the second attribution type corresponding to an image includes: inputting the image into an image processing model to generate an image semantic feature vector; and performing feature vector matching between the image semantic feature vector and a preset image type feature vector to determine the second attribution type corresponding to the image.

[0071] In this embodiment of the disclosure, to determine the second attribution type of an image, the image can be input into an image processing model to generate an image semantic feature vector, and then the image semantic feature vector can be matched with a preset image type feature vector to determine the second attribution type of the image.

[0072] In some embodiments, the image-text processing model is a large language model (e.g., CLIP (Contrastive Language-Image Pre-training, a cross-modal pre-trained model based on contrastive image-text learning) and BLIP (Bootstrapping Language-Image Pre-training, a bootstrapping image-text pre-trained model)).

[0073] In some embodiments, an initial image processing model is trained in advance using an image and text dataset to obtain an image and text processing model.

[0074] In some embodiments, the image and text dataset includes sample SMS images and their corresponding sample image semantic feature vectors and sample attribution types.

[0075] It should be noted that when determining the second attribution type of an image, one or more types can be determined. For example, the second attribution type of an image can be determined to be appointment type and notification type.

[0076] In some embodiments, both the text processing model and the image processing model are pre-trained AI models, which can be divided into two main categories. One category refers to a fine-tuned large language model trained using a collected SMS dataset, capable of recognizing preset classification types. Specifically, the collected SMS dataset is labeled with preset types, and this labeled SMS text dataset is used to fine-tune a pre-trained large language model (GPT, LLM) to obtain a fine-tuned large language model that recognizes SMS semantics and determines classification types. The other category refers to image-text pre-trained models (CLIP, BLIP), which, by collecting and training text-image pair data, enable the AI ​​model to master the matching relationship between text and image pairs.

[0077] The message types are categorized into four main types: appointment, notification, verification, and event. These are further subdivided according to the message subject and type, such as private conversation messages, historical appointment messages, verification message sets, notification message sets, billing messages, travel message sets, and marketing messages. The specific subject of the message is determined based on the message body and semantic classification to identify the specific subject within that category.

[0078] The specific matching process is as follows: The SMS content information and the word vectors parsed by the client are synchronized to the cloud. Semantic parsing is performed using large language model technologies (GPT, LLM) to extract semantic feature vectors from the SMS content. These semantic feature vectors include SMS topic word vectors and text semantic feature vectors. The SMS topic word vectors mainly summarize the SMS content and determine the specific event type. The text semantic feature vectors are then matched with preset category semantics during the large language model calculation process to output the category and category. Images backed up by the user within a specified time range are parsed, and image semantic feature vectors are extracted using image-text pre-trained models (CLIP, BLIP).

[0079] For example, based on the feature word vectors associated with content such as "takeout" and "express delivery," the message can be identified as a notification message; based on the date and other semantic features in the message, it can be identified as an appointment message, such as museum reservations, hospital registration appointments, or household registration appointments; and based on the message subject and semantic feature vectors, it can be identified as an event message, such as bills (consumption, water, electricity, gas, telecommunications, etc.), marketing activities (credit cards, shopping malls, brand advertisements, promotional information, etc.), itineraries (accommodation, flights, trains, etc.), or fraudulent messages. The specific names of the museums, hospitals, and shopping malls mentioned above are the specific subjects belonging to the classification.

[0080] S5 generates a schedule event based on SMS text and images, according to the first and second attribution types.

[0081] In this embodiment of the disclosure, when the first attribution type corresponding to the SMS text and the second attribution type corresponding to the image are determined, it can be determined whether to generate a schedule event based on the SMS text and the image according to the first attribution type and the second attribution type.

[0082] In some embodiments, if the first attribution type and the second attribution type are the same, a calendar event is generated based on the SMS text and the image. If the first attribution type and the second attribution type are different, it is determined that different calendar events will be generated based on the SMS text and the image respectively.

[0083] It is understandable that the first attribution type corresponding to the text message may include one type, and the second attribution type corresponding to the image may include one or more types. If there is a type in the second attribution type that is the same as the first attribution type, then it can be determined that the first attribution type and the second attribution type are consistent.

[0084] In this embodiment of the disclosure, a schedule event is generated based on SMS text and an image. The schedule event includes at least one of the following: SMS text, image, message receipt time, time range, image generation time, and event text description determined based on the image and SMS text.

[0085] In some embodiments, generating a schedule event based on SMS text and an image includes: inputting the SMS text into a text processing model to generate a text semantic feature vector; inputting the image into an image processing model to generate an image semantic feature vector; inputting the text semantic feature vector and the image semantic feature vector into an event generation model to generate an event text description; and generating a schedule event based on the SMS text's receiving time, the image's generation time, the time range, the image, and the event text description.

[0086] In this embodiment of the disclosure, a schedule event is generated based on SMS text and an image. The SMS text can be input into a text processing model to generate a text semantic feature vector, and the image can be input into an image processing model to generate an image semantic feature vector. The text semantic feature vector and the image semantic feature vector can be input into an event generation model to generate an event text description. Then, a schedule event is generated based on the SMS text's receiving time, the image's generation time, the time range, the image, and the event text description.

[0087] In some embodiments, if no image is generated within the time frame, a schedule event is generated based on the SMS text.

[0088] In this embodiment of the disclosure, if no image is generated within the time range, a schedule event can be generated solely based on the SMS text.

[0089] In some embodiments, a schedule event is generated based on the SMS text, including: a user receives an SMS from the operator's SMS service center, and after obtaining user authorization, a new SMS is parsed and managed. This involves: retrieving shallow information from the SMS to obtain the SMS text (including the sender, receipt time, and SMS content); identifying the sender using a public knowledge base and a user-defined contact database, and extracting the SMS subject word vector. For example, if the sender of the SMS received by the user is a bank, insurance company, operator, or courier, the SMS subject word vector can be identified by matching the number database in the public knowledge base; extracting the SMS content using a text information encoder to obtain time-limited word vectors and SMS subject word vectors; matching keywords such as "login," "pickup code," "verification code," and "appointment" to obtain the SMS subject word vector, and then determining the sender based on the SMS subject word vector; and finally, extracting the event description from the SMS text and generating a schedule event based on the sender and event description.

[0090] In some embodiments, if a new notification SMS message is added, it is merged with the existing SMS messages and listed based on the extracted semantic feature vector and the time of receipt. For example, for multiple express delivery information pickup codes, within a time limit, the time information, express locker number, pickup code, pickup location, and other relevant information from the existing and new SMS messages are merged into a list for display. Notification SMS messages that exceed a certain time limit are first hidden and then deleted.

[0091] In some embodiments, if the newly added SMS message is an event-related message, it is displayed according to the event type. If it is a bill-related message, it is associated with previous monthly bills and merged, and then displayed in a table format based on the bill type, bill time, and bill cost, using feature vectors. If it is a marketing campaign-related message, the newly added SMS message is displayed with the main body of the message, while other existing SMS messages with the same main body and type are hidden as sub-modules. If it is a trip-related SMS message, it is matched with existing SMS messages based on the SMS message feature word vectors, and different trips are displayed in chronological order, while existing SMS messages with different main bodies but for the same trip are hidden.

[0092] In this embodiment of the disclosure, users' text messages can be managed, displaying only valuable messages on the user's terminal while hiding and deleting duplicate, worthless, meaningless, and expired messages, thus facilitating user focus and improving text message viewing efficiency.

[0093] In this embodiment of the disclosure, if the first attribution type corresponding to the SMS text is consistent with the second attribution type corresponding to the image, that is, the SMS and the image belong to the same event, then the SMS and the image are associated, and a schedule event including a text description is generated using the surplus semantic feature values ​​of the SMS and the image. The implementation process is as follows:

[0094] (1) Based on the first attribution type corresponding to the obtained SMS text, the time range corresponding to the attribution type is obtained, and then SMS text messages of the same event and the same attribution type within the time range are merged to obtain all SMS events of the batch of SMS text messages.

[0095] (2) The text semantic feature vector of the event is matched with the image semantic feature vector within the time range to calculate the attribution type, and all associated images of the SMS event are obtained.

[0096] (3) Use the first category corresponding to the SMS text as the topic of the event, and input the text semantic feature vector and the image semantic feature vector as semantic generative task into the large model to obtain the event text description.

[0097] If the text message has no associated image, a description of the corresponding calendar event is generated based on the text message content; the remaining image set is then checked to see if it has event attributes, and if so, a description of the calendar event is generated based on the image content. The implementation process is as follows:

[0098] (1) If the text message has no associated image, then the event to which the text message belongs will be the topic of the schedule.

[0099] (2) Obtain the event time period based on the semantic feature vector in the text message, and then input it into the large model as a semantic generative task to obtain the text description of the event.

[0100] (3) Sequentially judge the images of the remaining unrelated SMS events. If there is a classification, use the semantic feature vector of the image as the input of the semantic generative task into the large model to obtain the text description of the event.

[0101] Finally, based on the time range, SMS receipt time, image generation time, and event text description mentioned above, the content is integrated to generate the corresponding schedule event.

[0102] In some embodiments, after generating a calendar event based on SMS text and an image, a historical calendar event belonging to the same event as the calendar event is determined; address information corresponding to the calendar event and the historical calendar event is obtained respectively; if the address information is the same, the calendar event is associated with the historical calendar event.

[0103] In this embodiment of the disclosure, if it is determined that there is a historical schedule event that belongs to the same event as the scheduled event, the address information corresponding to the scheduled event and the historical schedule event are obtained respectively. If the address information is the same, the scheduled event and the historical schedule event are associated.

[0104] In this embodiment of the disclosure, candidate historical calendar events composed of historical text messages and images are searched and judged based on the calendar events and the user's proprietary knowledge base.

[0105] In some embodiments, determining a historical event that belongs to the same event as the scheduled event includes: obtaining a first sending entity corresponding to a text message; determining a second sending entity and a third attribution type corresponding to a candidate historical text message; and performing a similarity calculation based on the first sending entity and the first attribution type of the text message, and the second sending entity and the third attribution type of the candidate historical text message, to determine the historical event that belongs to the same event as the scheduled event.

[0106] In this embodiment of the disclosure, to determine a historical event that belongs to the same event as the scheduled event, the first sending entity corresponding to the SMS text can be obtained, and the second sending entity and third attribution type corresponding to the candidate historical SMS text can be determined. Based on the first sending entity and first attribution type of the SMS text, and the second sending entity and third attribution type of the candidate historical SMS text, a similarity calculation is performed to determine the historical event that belongs to the same event as the scheduled event.

[0107] In some embodiments, a correlation search and judgment is performed on candidate historical calendar events composed of historical SMS messages and images, based on calendar events and a user-specific knowledge base. The correlation search and judgment process is as follows:

[0108] (1) Using the first attribution type and the first sender of the SMS text, perform similarity matching calculations on the second sender and the third attribution type of the candidate historical SMS text respectively.

[0109] (2) Identify historical calendar events that belong to the same event as the calendar event through similarity matching calculation results.

[0110] (3) Combine the address information corresponding to the scheduled event and historical scheduled events with the user's proprietary knowledge base to determine whether it is the user's activity area or permanent residence. If the address information is the same, then associate the scheduled event with the historical scheduled event.

[0111] (4) If a connection is determined, the previous connection history is marked. Specifically, based on the matching result, the historical SMS text is located, and its corresponding time range, schedule topic, and schedule event are queried. The above information of the historical SMS text is then associated with the schedule event and displayed.

[0112] The process involves analyzing and identifying the specific subject of newly added SMS messages. If the sender of an incremental SMS message is an unknown subject in the knowledge base, the sender is labeled based on the semantic feature vector of the SMS message, classifying it as the corresponding specific subject of the SMS message. Information such as the sender, user activity region, and place of residence are standardized and labeled, and the results are synchronized to the user's proprietary knowledge base. For example, even if the delivery or takeout order originates from different courier companies or uses different numbers for notification, it can still be identified and synchronized to the user's proprietary knowledge base after matching the SMS feature vectors. When users associate events with their schedules, they can search the knowledge base to obtain related SMS messages or schedule information.

[0113] In some embodiments, a schedule record is generated based on the schedule events corresponding to SMS texts obtained during a specific time period; historical schedule records are obtained; and recommended information is determined based on the historical schedule records and the schedule record.

[0114] In some embodiments, a specific time period can be the current day, the current month, the current quarter, the current year, etc.

[0115] For example, all the scheduled events of the day are sorted according to the order in which the specific events occurred to form a daily schedule.

[0116] In this embodiment of the disclosure, a schedule record is generated based on the schedule events corresponding to SMS texts obtained during a specific time period, and historical schedule records are obtained. Recommendation information is determined based on the historical schedule records and the schedule record.

[0117] In some embodiments, determining recommended information based on historical calendar entries and calendar entries includes: determining a target calendar entry that matches the calendar entry in the historical calendar entries; and determining recommended information based on the target calendar entry and calendar entries.

[0118] In this embodiment, the analysis of daily schedules is used as a classification task. Based on the potential time span, location information, trend change information, and event completion status, the task types are divided into remote recommendation, daily recommendation, to-do list recommendation, and behavioral habit recommendation. The semantic information of the current and past daily schedules, along with semantic information from the user's knowledge base such as historical residence and activity areas, is combined to form an input feature vector. This vector is then used as input to the large model for the classification task to obtain the classification task type. After determining the task type for this event recommendation, relevant information is retrieved from the internet to recommend and suggest potential related arrangements.

[0119] For example, for daily commutes or appointment-based services, notifications can be provided regarding weather and traffic congestion changes at the corresponding location. For out-of-town travel, notifications can be provided based on the location, including weather, expected road congestion, recommended modes of transportation, and suggested attractions. For consumption bills, users' past spending habits can be quantitatively compared based on their behavior. For consumer goods, relevant product recommendations and similar brand recommendations can be provided.

[0120] In this embodiment of the disclosure, SMS messages, as low-volume text information, do not require a large amount of cloud space even if they are considered permanent assets of the user. As a supplement to the user's cloud storage intelligent service, it provides users with multi-terminal synchronization and cloud collaboration, and performs full storage of relevant information of historical SMS messages, including historical SMS text, historical calendar notes, and historical calendar events, thereby reducing the user's terminal storage costs and avoiding data loss and recovery costs caused by the failure and loss of local storage devices.

[0121] In this embodiment of the disclosure, the management method for classifying SMS content, managing schedules, and related searches provides users with a more efficient means of searching for SMS messages or information, facilitating user information retrieval and historical record review.

[0122] In this embodiment, the optimized and trained large language model technology is used to parse the semantics of SMS text, which is more accurate and has better quality assurance compared with related technologies.

[0123] SMS has advantages such as no data consumption, fast message delivery, and strong real-time capability. Although SMS is an indispensable service for users, compared with other instant messaging social applications such as WeChat and DingTalk, SMS has problems such as message clutter, unclear categories, and unknown sender.

[0124] In this disclosed embodiment, cloud storage technology can be incorporated to provide users with both asset cloud storage services and intelligent SMS management services. As SMS is a primary revenue stream for telecom operators, cloud collaboration and intelligent features can help users solve problems related to SMS viewing, management, and synchronization backup after phone replacement, providing better SMS value-added services and improving the user experience.

[0125] In this embodiment, users can use this system to assist in managing text messages on their electronic terminals. By optimizing and training a large semantic model, the system performs semantic parsing and type analysis on user-backed images and text messages. It provides intelligent text message management services by combining various aspects such as message subject, message topic, message attributes, message semantics, and event time limits, assisting users in managing their text messages, meeting their personalized needs, and improving user experience.

[0126] The SMS management method disclosed in this embodiment provides terminal SMS and cloud-based full SMS management. This method uses large language model technology (GPT, LLM) to identify text content, and then uses an optimized and trained semantic model to perform semantic parsing and type analysis on user-backed images and SMS messages. Furthermore, it designs key information content parsing methods for various aspects, including message subject, message theme, message attributes, message semantics, and event time limits. This enables intelligent management services for user SMS messages.

[0127] Figure 2 is a schematic diagram of a text message management system according to an exemplary embodiment.

[0128] As shown in Figure 2, the SMS management system provided in this embodiment mainly involves an SMS service center, multi-terminal clients, and a service platform. The SMS service center is the SMS sender. Multi-terminals include mobile phones, computers, tablets, and other electronic devices. Users can view and manage SMS information through web interfaces, H5 pages, and mobile clients adapted to these electronic devices.

[0129] As shown in Figure 3, users receive SMS messages via their mobile phones. After authorization, users can manage and view their SMS messages through clients on multiple terminals. The business service platform is the intelligent SMS management platform, which consists of a storage service platform, a content parsing platform, and an intelligent management platform. The storage service platform provides SMS cloud storage services, ensuring synchronization across multiple terminals; the content parsing platform provides users with SMS content parsing functions, analyzing the sender, receiving time, and SMS content; the intelligent management platform classifies, organizes, and manages various SMS elements such as message subject, message theme, message semantics, and event time limit, organizing SMS messages on users' mobile phones and aggregating and categorizing cloud-based SMS messages; the knowledge base is divided into a public knowledge base, a private knowledge base, and a proprietary knowledge base, used to store, identify, and synchronize sending sources, including but not limited to sending sources identified by operators, enterprise sending sources with publicly available numbers, sending sources marked by users, and sending sources engaged in established businesses.

[0130] Figure 4 is a structural diagram of a text message management device according to an exemplary embodiment.

[0131] As shown in Figure 4, the SMS management device 1 includes: a text acquisition unit 11, a first type determination unit 12, a time determination unit 13, a second type determination unit 14, and an event generation unit 15.

[0132] The text acquisition unit 11 is used to acquire SMS text.

[0133] The first type determination unit 12 is used to determine the first attribution type corresponding to the SMS text.

[0134] The time determination unit 13 is used to determine the time range corresponding to the first attribution type.

[0135] The second type determination unit 14 is used to determine the second attribution type of an image if an image is generated within a time range.

[0136] Event generation unit 15 is used to generate schedule events based on SMS text and images according to a first attribution type and a second attribution type.

[0137] In some embodiments, the first type determination unit 12 is used to input the SMS text into a text processing model to generate a text semantic feature vector; and to perform feature vector matching between the text semantic feature vector and a preset text type feature vector to determine the first type to which the SMS text belongs.

[0138] In some embodiments, the second type determination unit 14 is used to input an image into a text-image processing model to generate an image semantic feature vector; and to perform feature vector matching between the image semantic feature vector and a preset image type feature vector to determine the second type to which the image belongs.

[0139] In some embodiments, the event generation unit 15 is configured to input SMS text into a text processing model to generate a text semantic feature vector; input an image into an image processing model to generate an image semantic feature vector; input the text semantic feature vector and the image semantic feature vector into an event generation model to generate an event text description; and generate a schedule event based on the SMS text's receiving time, the image's generation time, time range, the image, and the event text description.

[0140] In some embodiments, the event generation unit 15 is further configured to generate a schedule event based on the SMS text if no image is generated within the time range.

[0141] As shown in Figure 5, in some embodiments, the SMS management device 1 further includes: a historical schedule determination unit 16, an information acquisition unit 17, and an association processing unit 18.

[0142] Historical schedule determination unit 16 is used to determine historical schedule events that belong to the same event as the scheduled event.

[0143] The information acquisition unit 17 is used to acquire the address information corresponding to scheduled events and historical scheduled events, respectively.

[0144] The association processing unit 18 is used to associate the schedule event with the historical schedule event if the address information is the same.

[0145] In some embodiments, the historical schedule determination unit 16 is used to obtain the first sending entity corresponding to the SMS text; determine the second sending entity and the third attribution type corresponding to the candidate historical SMS text; and perform similarity calculation based on the first sending entity and the first attribution type of the SMS text, and the second sending entity and the third attribution type of the candidate historical SMS text, to determine the historical schedule event that belongs to the same event as the schedule event.

[0146] As shown in Figure 6, in some embodiments, the SMS management device 1 further includes: a note generation unit 19, a historical note acquisition unit 20, and an information determination unit 21.

[0147] The note generation unit 19 is used to generate schedule notes based on the schedule events corresponding to SMS text obtained during a specific time period.

[0148] Historical record acquisition unit 20 is used to acquire historical schedule records.

[0149] Information determination unit 21 is used to determine recommended information based on historical schedule records and schedule records.

[0150] In some embodiments, the information determining unit 21 is used to determine a target schedule record that matches a schedule record in historical schedule records; and to determine recommended information based on the target schedule record and the schedule record.

[0151] According to embodiments of this disclosure, an electronic device is also provided, including: a processor; and a memory for storing processor-executable instructions, wherein the processor is configured to: implement the SMS management method disclosed in embodiments of this disclosure.

[0152] To implement the above embodiments, this disclosure also proposes a storage medium.

[0153] When the instructions in the storage medium are executed by the processor, the processor is able to execute the SMS management method disclosed in the embodiments of this disclosure.

[0154] To implement the above embodiments, this disclosure also provides a computer program product.

[0155] When the computer program product is executed by the processor of the electronic device, it enables the electronic device to execute the SMS management method disclosed in the embodiments of this disclosure.

[0156] Figure 7 is a structural block diagram of an electronic device according to an exemplary embodiment. The electronic device shown in Figure 7 is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0157] As shown in Figure 7, the electronic device 1000 includes a processor 111, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 112 or a program loaded from memory 116 into a random access memory (RAM) 113. The RAM 113 also stores various programs and data required for the operation of the electronic device 1000. The processor 111, ROM 112, and RAM 113 are interconnected via a bus 114. An input / output (I / O) interface 115 is also connected to the bus 114.

[0158] The following components are connected to I / O interface 115: memory 116 including hard disks, etc.; and communication section 117 including network interface cards such as local area network (LAN) cards, modems, etc., communication section 117 performs communication processing via a network such as the Internet; and driver 118 is also connected to I / O interface 115 as needed.

[0159] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 117. When the computer program is executed by processor 111, it performs the functions defined in the methods of this disclosure.

[0160] In an exemplary embodiment, a storage medium including instructions is also provided, such as a memory including instructions, which can be executed by the processor 111 of the electronic device 1000 to perform the above-described method. Optionally, the storage medium may be a non-transitory computer-readable storage medium, such as a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device.

[0161] In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can transmit, propagate, or transfer a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wireline, optical fiber, RF, etc., or any suitable combination thereof.

[0162] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the following claims.

[0163] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

Claims

1. A method for managing text messages, characterized in that, include: Get the SMS text; Determine the first attribution type corresponding to the SMS text; Determine the time range corresponding to the first attribution type; if an image is generated within the time range, determine the second attribution type corresponding to the image; Based on the first attribution type and the second attribution type, a schedule event is generated based on the SMS text and the image, including: inputting the SMS text into a text processing model to generate a text semantic feature vector; inputting the image into an image-text processing model to generate an image semantic feature vector; inputting the text semantic feature vector and the image semantic feature vector into an event generation model to generate an event text description; generating the schedule event based on the SMS text's receiving time, the image's generation time, the time range, the image, and the event text description; the method further includes: determining historical schedule events belonging to the same event as the schedule event; obtaining address information corresponding to the schedule event and the historical schedule event respectively; judging the address information corresponding to the schedule event and the historical schedule event respectively based on a user-specific knowledge base, and associating the schedule event with the historical schedule event.

2. The method as described in claim 1, characterized in that, The step of determining the historical schedule event that belongs to the same event as the scheduled event includes: obtaining the first sending entity corresponding to the SMS text; determining the second sending entity and the third attribution type corresponding to the candidate historical SMS text; and performing similarity calculation based on the first sending entity and the first attribution type of the SMS text, and the second sending entity and the third attribution type of the candidate historical SMS text, to determine the historical schedule event that belongs to the same event as the scheduled event.

3. The method as described in claim 1, characterized in that, The method further includes: generating a schedule record based on the schedule event corresponding to the SMS text obtained in a specific time period; obtaining historical schedule records; and determining recommendation information based on the historical schedule records and the schedule record.

4. The method as described in claim 3, characterized in that, The step of determining recommended information based on the historical schedule entries and the schedule entries includes: determining a target schedule entry in the historical schedule entries that matches the schedule entry; and determining the recommended information based on the target schedule entry and the schedule entry.

5. A text message management device, characterized in that it comprises: The text acquisition unit is used to acquire SMS text. The first type determination unit is used to determine the first attribution type corresponding to the SMS text. A time determination unit is used to determine the time range corresponding to the first attribution type; a second type determination unit is used to determine the second attribution type corresponding to the image if an image is generated within the time range. An event generation unit is configured to generate a schedule event based on the SMS text and the image according to the first attribution type and the second attribution type, including: inputting the SMS text into a text processing model to generate a text semantic feature vector; inputting the image into an image-text processing model to generate an image semantic feature vector; inputting the text semantic feature vector and the image semantic feature vector into an event generation model to generate an event text description; and generating the schedule event according to the SMS text's receiving time, the image's generation time, the time range, the image, and the event text description. The device further includes: a historical schedule determination unit, an information acquisition unit, and an association processing unit; the historical schedule determination unit is configured to determine historical schedule events belonging to the same event as the schedule event; the information acquisition unit is configured to acquire address information corresponding to the schedule event and the historical schedule event respectively; and the association processing unit is configured to judge the address information corresponding to the schedule event and the historical schedule event according to a user-specific knowledge base, and associate the schedule event with the historical schedule event.

6. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the method as described in any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 4.

8. A computer program product, characterized in that, When the computer program in the computer program product is executed by a processor, the method as described in any one of claims 1 to 4 is performed.

Images