To-do list reminding method and device and electronic equipment

Abstract

The invention discloses a to-do list reminding method and device and electronic equipment, and belongs to the technical field of artificial intelligence. The method comprises the following steps: inputting a chat message between a home terminal user and a contact into a language large model, and displaying first prompt information under the condition that the language large model recognizes a to-do list in the chat message; the first prompt information comprises item description information and item completion deadline of the identified to-do item; and creating a to-do prompt, wherein the prompt content of the to-do prompt is the to-do list.

Description

Technical Field

[0001] The present application belongs to the field of artificial intelligence technology, and specifically relates to a to-do list reminder method, device and electronic device. Background Art

[0002] Nowadays, in daily work and life, users generate many to-do items when communicating with others. These to-do items may come from other people's requirements of the users, or from the users' active commitments to others. After generating multiple to-do items, in order to avoid forgetting the to-do items, users want to set to-do reminders for each to-do item.

[0003] However, in related art, to-do reminders typically require users to summarize the contents of the to-do items they need to perform from chat messages after receiving them, and then manually create to-do reminders for each to-do item based on the contents of the to-do items. This to-do reminder method requires users to spend a lot of time and manpower, resulting in low efficiency. Summary of the Invention

[0004] The purpose of the embodiments of the present application is to provide a to-do list reminder method, device and electronic device, which reduce the user's operation steps and improve the efficiency of to-do list reminders.

[0005] In a first aspect, an embodiment of the present application provides a to-do reminder method, the method comprising:

[0006] Obtaining location information and activity information of at least two users, as well as chat messages between the local user and at least one contact, wherein the at least two users include the local user and the contact;

[0007] Determining N to-do items of the local user based on chat messages between the local user and at least one contact, where N is a positive integer;

[0008] determining the relationship strength and commitment sensitivity of the local user to each of the contacts based on the location information, the activity information, and the chat messages;

[0009] The execution order of the N to-do items is determined according to the relationship strength, commitment sensitivity, the association relationship between the N to-do items, and the deadline of each to-do item.

[0010] In a second aspect, an embodiment of the present application provides a to-do list reminder device, the device comprising:

[0011] an acquisition module, configured to acquire location information and activity information of at least two users, and chat messages between the local user and at least one contact, wherein the at least two users include the local user and the contact;

[0012] an identification module, configured to determine N to-do items of the local user based on chat messages between the local user and at least one contact, where N is a positive integer;

[0013] a determination module, determining the relationship strength and commitment sensitivity of the local user to each of the contacts based on the location information, the activity information, and the chat messages;

[0014] A sorting module is used to determine the execution order of the N to-do items according to the relationship strength, commitment sensitivity, the association relationship between the N to-do items and the deadline of each to-do item.

[0015] In a third aspect, an embodiment of the present application provides an electronic device comprising a processor and a memory, wherein the memory stores programs or instructions that can be run on the processor, and when the programs or instructions are executed by the processor, the steps of the method provided in the first aspect are implemented.

[0016] In a fourth aspect, an embodiment of the present application provides a readable storage medium, on which a program or instruction is stored, and when the program or instruction is executed by a processor, the steps of the method provided in the first aspect are implemented.

[0017] In a fifth aspect, an embodiment of the present application provides a chip, which includes a processor and a communication interface, the communication interface and the processor are coupled, and the processor is used to run programs or instructions to implement the method provided in the first aspect.

[0018] In a sixth aspect, an embodiment of the present application provides a computer program product, which is stored in a storage medium and executed by at least one processor to implement the method provided in the first aspect.

[0019] In the to-do list reminder method, device, and electronic device of the present application, the to-do items in the chat message can be automatically identified through the language model, and the prompt information of each to-do item can be displayed, and then a to-do reminder for the to-do item can be created. Compared with the manual identification of to-do items and manual creation of to-do reminders in the related art, the process of item identification and reminder creation is greatly simplified. In this way, the automatic identification of to-do items and generation of to-do reminders through the language model reduces the user's operation steps and improves the efficiency of to-do item reminders. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a flowchart of a to-do reminder method provided in some embodiments of the present application;

[0021] Figure 2 is a schematic diagram of a display interface of an electronic device provided in some embodiments of the present application;

[0022] Figure 3 is a schematic diagram of a display interface of an electronic device provided in some embodiments of the present application;

[0023] Figure 4 is a schematic structural diagram of a to-do list reminder device provided in some embodiments of the present application;

[0024] Figure 5 is a schematic structural diagram of an electronic device provided by some embodiments of the present application;

[0025] Figure 6 It is a schematic diagram of the hardware structure of the electronic device provided in some embodiments of the present application. DETAILED DESCRIPTION

[0026] The following will be combined with the accompanying drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the embodiments described are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field are within the scope of protection of this application.

[0027] The terms "first," "second," and the like in the specification and claims of this application are used to distinguish similar objects, and are not used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of this application can be implemented in an order other than that illustrated or described herein, and that the objects distinguished by "first," "second," and the like are generally of the same type, and do not limit the number of objects; for example, the first object can be one or more. In addition, the term "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0028] The terms used in the embodiments of this application are only used to explain the specific embodiments of this application and are not intended to limit this application. The following is an explanation of the terms involved in the embodiments of this application.

[0029] To-do items: Specific tasks or committed behaviors that users need to complete before a certain point in the future.

[0030] Relationship strength: The closeness of the social connection between the local user and the contact.

[0031] Commitment sensitivity: The priority of the tasks that the user commits to the contact.

[0032] Interaction intensity: represents the activity of social connections between the local user and the contacts during the reference time period.

[0033] Comprehensive vector: Represents the social relationship characteristics between the local user and the contact from at least two perspectives.

[0034] Chat Rate: The ratio of the number of days in which actual chats occurred between the local user and the contact during the reference time period to the total number of days in the reference time period.

[0035] Location overlap rate: The ratio of the number of days during a reference time period that the local user and the contact appear at the same geographic location to the total number of days during the reference time period.

[0036] Joint activity rate: The ratio of the number of days during the reference time period that the local user and the contact jointly participated in the same project or event to the total number of days during the reference time period.

[0037] Multidimensional relationship network model: A graph structure model with the local user as the core and based on the multidimensional interaction characteristics of the local user and each contact. It is used to characterize the strength and structural distribution of the social relationships between the local user and all contacts.

[0038] Spatial vector: A feature vector that represents the frequency of geographic contact between the local user and the contact within a reference time period.

[0039] Semantic feature vector: A feature vector obtained by encoding multiple chat contents between the local user and the contact at the semantic level.

[0040] Time series vector: A feature vector that represents the temporal distribution of the interaction between the local user and the contact over the entire time range.

[0041] Commitment keywords: participles that express the tone of commitment in chat messages.

[0042] Communication Depth: This value represents the communication density and information richness between the user and each contact in the conversation.

[0043] Commitment word frequency: used to represent the frequency of words with a commitment tone in conversations between users and contacts.

[0044] Commitment message: A chat message that includes indicator variables for the commitment sentence.

[0045] Commitment strength: characterizes the degree to which the commitment tone is expressed in a commitment message.

[0046] Average message frequency: The average frequency of chat messages generated by the local user and contacts per unit time during the reference time period.

[0047] Average Message Length: The average number of characters in each chat message between the user and the contact during the reference time period.

[0048] Timely message reply rate: The proportion of all chat messages that receive a reply within the preset time threshold.

[0049] Collaboration: This refers to the strength of the collaborative weight coefficient between a specific to-do item and other to-do items. It can reflect factors such as whether the to-do item is close in time to other to-do items, whether they come from the same contact person, and whether they can be combined for execution. It is an auxiliary weight that reflects the composability and overall execution efficiency of to-do items.

[0050] To solve the above technical problems, the present application provides a to-do list reminder method. The following describes the to-do list reminder method provided by the present application in detail through specific embodiments and application scenarios in conjunction with the accompanying drawings.

[0051] It should be noted that the execution subject of the to-do list reminder method provided in the embodiments of the present application can be an electronic device such as a mobile phone, tablet computer, laptop computer, PDA, in-vehicle electronic device, etc. In some embodiments of the present application, the execution subject of the to-do list reminder method is an electronic device as an example to illustrate the to-do list reminder method provided in the embodiments of the present application.

[0052] The to-do list reminder method provided in the embodiment of the present application can be applied to at least two specific application scenarios, wherein one specific application scenario is a work communication scenario, and the other specific application scenario is a family life scenario.

[0053] like Figure 1 As shown, Figure 1 : is a flowchart of a to-do list reminder method provided by an embodiment of the present application. The embodiment of the present application provides a to-do list reminder method, which may include:

[0054] Step 101: Input a chat message between a local user and a contact into a language macro model. If the language macro model identifies a to-do item in the chat message, display a first prompt message; the first prompt message includes: a description of the identified to-do item and a deadline for completion of the item;

[0055] In this embodiment, during a chat between a local user and each contact, the electronic device's system can continuously monitor the chat messages between the local user and the contact. For any chat message between the local user and the contact, the system can input the chat message, along with at least one previous chat message between the local user and the contact, into a large language model, which then performs semantic understanding and contextual reasoning on the chat message. The large language model can determine, based on the input chat message, whether the local user has expressed an intention to complete a certain task before a certain point in the future. Once it is determined that the local user has expressed an intention to complete a certain task before a certain point in the future, the system can determine the to-do items that the user wants to complete based on the intention expressed by the local user.

[0056] After the system identifies a to-do item, the electronic device may display first prompt information of the to-do item in the form of a pop-up window. The first prompt information may include item description information and the deadline for completion of the to-do item.

[0057] Specifically, the first prompt for a to-do item will only be displayed if the user can identify the to-do item's description, deadline, and associated contacts from the chat message. If the to-do item's description, deadline, or associated contacts cannot be extracted, the first prompt for the to-do item will not be displayed. The to-do item description is the specific task action expressed by the user in the chat message, such as "write a report" or "book a hotel." The deadline refers to the agreed or implied completion time limit for the to-do item. The associated contacts can be the initiator, recipient, or collaborator of the item.

[0058] For example, in a work communication scenario, a user is discussing work matters with a colleague. A chat message input into the language model includes the user sending the colleague "I'll finish the PPT tomorrow afternoon." The language model can recognize that the user's intention in this chat message is to complete a to-do item by a certain time in the future. It can then extract the to-do item's description as "complete the PPT," the deadline as "tomorrow afternoon," and the contact person as the colleague with whom the user is chatting. The completion of the PPT can then be listed as a to-do item for the user, and the first prompt for this to-do item can be displayed. In a family life scenario, the user is chatting with a family member. A chat message input into the language model includes the user sending the family member "The weather is nice today, I want to go for a walk." Although this chat message also includes the user's intention to go for a walk, the model cannot recognize the deadline for this intention. Therefore, the language model can ignore this input and not display the first prompt for this to-do item.

[0059] In the above scheme, only when the item description information and the deadline for completion of a to-do item are identified through the chat message, the first prompt information of the to-do item will be displayed to remind the user of the item description information and the deadline for completion of the to-do item. In this way, only the to-do items with complete information can be prompted, which can effectively improve the accuracy of the to-do item prompts and ensure that the prompted to-do items are executable.

[0060] Step 102: Create a to-do reminder, where the reminder content of the to-do reminder is the to-do item.

[0061] In this embodiment, after the first prompt information corresponding to the to-do item is displayed on the electronic device, the system can create a to-do reminder. The to-do reminder refers to a reminder task automatically generated by the system based on the identified to-do item, which is used to remind the user at a set time point to prompt the user to complete the to-do item in a timely manner before the deadline.

[0062] In an embodiment of the present application, a language model can be used to automatically identify to-do items in chat messages, display prompt information for each to-do item, and then create a to-do reminder for the to-do item. Compared with the related art of manually identifying to-do items and manually creating to-do reminders, this greatly simplifies the process of item identification and reminder creation. In this way, the automatic identification of to-do items and generation of to-do reminders through a language model reduces the user's operation steps and improves the efficiency of to-do reminders.

[0063] Specifically, in some embodiments, the first prompt information further includes: to-do reminder creation prompt information, the to-do reminder creation prompt information is used to prompt that a to-do reminder with the reminder content being the to-do item has been created;

[0064] When the language model identifies a to-do item in the chat message, the method further includes:

[0065] Show cancel control;

[0066] Creating a to-do reminder includes:

[0067] When the language model identifies a to-do item in the chat message, creating a to-do reminder;

[0068] After creating the to-do reminder, it also includes:

[0069] In a case where a selection input of the cancel control is received within a first duration of displaying the cancel control, the to-do reminder is deleted.

[0070] In this embodiment, while the first prompt information of the to-do item is displayed on the electronic device, a cancel control may also be displayed, where the cancel control is used to instruct the user to cancel the to-do reminder corresponding to the created to-do item.

[0071] In some Figure 2 In the embodiment shown, in the application interface 201 of the social application, the system receives a new chat record "Our group will do the next year summary at 3 pm next Tuesday, everyone prepare the PPT", then the system can identify the to-do item "Prepare the annual summary PPT", and then can display the first prompt information 202 of the to-do item and the cancel control 203.

[0072] Specifically, if the user on the local end believes that the to-do item has been misidentified or that the user on the local end does not want to execute the to-do item, the user can select and input the cancel control within the first duration of the first prompt information being displayed. The first duration can be set to any duration between 1 minute and 5 minutes. In this way, the system will respond to the selection input and delete the to-do reminder corresponding to the to-do item, and the system will not remind the user of the to-do item. If the user on the local end approves the to-do item, the user will not select and input the cancel control within the first duration of the to-do item. In this way, the system will remind the user of the to-do item.

[0073] The selection input for the cancel control may be a touch input performed by the user on the cancel control displayed on the electronic device using a finger or a touch device such as a stylus. For example, the touch input may be a touch click operation, a slide operation, or other operations. The touch input for the cancel control displayed on the electronic device may also be an input operation performed by the user using other input devices connected to the electronic device, such as a mouse, remote control device, keyboard, etc.

[0074] Through the above method, a confirmation mechanism for the user to the to-do reminder corresponding to the to-do item can be introduced through the cancel control. If the user believes that the to-do item identified by the system is incorrect, the to-do reminder of the to-do item can be canceled manually, thereby avoiding the interference caused by the system's misidentification and improving the accuracy of the to-do reminder.

[0075] Specifically, in some embodiments, the first prompt information further includes: to-do reminder creation confirmation information: the to-do reminder creation confirmation information is used to prompt the user whether to create a to-do reminder based on the to-do item;

[0076] Creating a to-do reminder includes:

[0077] In a case where a confirmation input for the to-do reminder creation confirmation information is received, a to-do reminder is created in response to the confirmation input.

[0078] In this embodiment, the first prompt information displayed on the electronic device for a to-do item may include the item description information and the deadline for completion of the identified to-do item, and may also include to-do reminder creation confirmation information. The to-do reminder creation confirmation information is used to guide the user to decide whether to generate a to-do reminder based on the identified to-do item.

[0079] Exemplarily, the to-do reminder creation confirmation information may be a displayed "Add" control and a "Don't Add" control, wherein the "Add" control is used to indicate that a to-do reminder corresponding to the to-do item is created, and the "Don't Add" control is used to indicate that a to-do reminder corresponding to the to-do item is not created. The to-do reminder creation confirmation information may also be a checkbox, where if the checkbox is selected, it indicates that a to-do reminder corresponding to the to-do item is created, and if the checkbox is unselected, it indicates that a to-do reminder corresponding to the to-do item is not created.

[0080] Specifically, if the local user wants to create a to-do reminder corresponding to the to-do item, he or she can confirm the displayed to-do reminder creation confirmation information. The system receives and responds to the confirmation input, creates a to-do reminder corresponding to the to-do item, and reminds the to-do item through the to-do reminder.

[0081] If the to-do reminder creation confirmation information is a displayed "add" control, the confirmation input for the displayed to-do reminder creation confirmation information can be a touch input by the local user to the "add" control displayed in the first prompt information using a finger or a touch device such as a stylus. For example, the touch input can be a touch click operation, a sliding operation, or other operations. The touch input for the cancel control displayed on the electronic device can also be an input operation performed by the user using other input devices connected to the electronic device, such as a mouse, remote control device, keyboard, etc.

[0082] If the to-do reminder creation confirmation information is a checkbox, the confirmation input for the displayed to-do reminder creation confirmation information may be a touch input on the displayed to-do reminder creation confirmation information to switch the state of the checkbox to a selected state. For example, the touch input may be a touch click operation, a slide operation, or other operation. The first input may also be an input operation performed by the user using another input device connected to the electronic device, such as a mouse, remote control device, or keyboard.

[0083] Through the above method, a confirmation mechanism can be used for users to create confirmation information for to-do reminders, so that users can decide whether to create to-do reminders for to-do items. If the user believes that the to-do items identified by the system are incorrect, then the to-do reminders for the to-do items can be not created, thereby avoiding the interference caused by system misidentification and improving the accuracy of to-do reminder creation.

[0084] In some embodiments, inputting the chat messages between the local user and the contact into the language macro model includes:

[0085] Inputting chat messages between the local user and at least one contact into the language model;

[0086] When the language macro model identifies N to-do items in the chat message and N is an integer greater than 1, the method further includes:

[0087] Obtain location information and activity information of the local user and at least one contact;

[0088] An order of executing the N to-do items is determined according to the location information, the activity information, and chat messages between the local user and the at least one contact.

[0089] In this embodiment, location information refers to the specific geographic coordinates of each user at a specific point in time, which is used to reflect the spatial contact behavior of each user; activity information refers to the specific projects or events in which the user participates, such as meetings, training, and outings, and the activity information may include attributes such as the time when the activity occurs.

[0090] Since the chat messages between the local user and each contact contain the plans, commitments or action intentions expressed by the local user, the local user's to-do items can be identified from the chat messages. If the identified to-do items are N items greater than 1, the location information and activity information of the local user and at least one contact can be obtained. Specifically, with the help of the semantic understanding ability of the language large model for text content and context content, the content with task attributes in the chat messages can be identified, and the local user's N to-do items can be determined based on these contents. The location information, activity information and chat messages can be obtained separately through the data acquisition module of the electronic device. For example, the chat messages can be obtained through the system interface of the social application in the electronic device, the location information can be obtained by the GPS positioning module in the electronic device, and the activity information can be obtained through the calendar application or meeting record application in the electronic device.

[0091] After obtaining location information, activity information, and chat messages between the local user and at least one contact, the importance, execution intention, logical sequence, and time pressure of each to-do item can be weighed based on the above information, so as to reasonably sort the N to-do items and generate an execution order that is closer to the user's intention and actual constraints.

[0092] In this embodiment, after identifying multiple to-do items, the execution order of each to-do item can be determined by combining the location information, activity information and chat information of the user and the contact through comprehensive analysis of multi-source information, thereby avoiding the stacking and conflict of multiple to-do items and improving the rationality and execution efficiency of the arrangement of multiple to-do items.

[0093] In some embodiments, determining the execution order of the N to-do items based on the location information, the activity information, and chat messages between the local user and the at least one contact includes:

[0094] Determining, based on the location information, the activity information, and chat messages between the local user and each contact, the relationship strength and commitment sensitivity between the local user and each contact, wherein the relationship strength is used to represent the closeness of the social connection between the local user and each contact, and the commitment sensitivity is used to represent the priority of the task committed by the local user to the contact;

[0095] An order of executing the N to-do items is determined according to the relationship strength, commitment sensitivity, the degree of association among the N to-do items, and the deadline for completing each to-do item.

[0096] In the embodiments of the present application, relationship strength refers to the closeness of the social connection between the local user and each contact, and is used to measure the importance of the contact to the local user and the frequency of interaction. The stronger the relationship strength between the local user and the contact, the closer the social connection and the higher the trust between the two. Consequently, the likelihood and priority of completing the to-do items promised by the local user to the contact are also higher. Commitment sensitivity, on the other hand, refers to the priority of the items promised by the local user to the contact. It is used to assess the user's subjective intention and willingness to complete the items. Therefore, the stronger the local user's commitment sensitivity to the contact, the higher the likelihood and priority of completing the to-do items promised by the local user to the contact.

[0097] In this embodiment, the relationship strength and commitment sensitivity between the user in this segment and each contact may be comprehensively determined based on the user's location information, activity information, and chat messages between the user in this segment and at least one contact.

[0098] For example, the strength of the relationship between the user and a contact can be determined by counting the frequency of their appearances at the same geographic location within the same time period, the number and duration of events they jointly participated in, and the frequency of their chats. For example, if two people frequently check in at the same office, frequently attend meetings together, and frequently chat privately, the relationship strength is high. The commitment sensitivity between the user and a contact can be determined by the language features in their chat messages. For example, if a user uses words with a strong commitment tone, such as "must" or "definitely," the commitment sensitivity is high.

[0099] Since relationship strength reflects the degree of social closeness between the user and their contacts, commitment sensitivity reflects the user's willingness to execute to-do items, the relationships between to-do items describe their dependencies or collaboration, and deadlines reflect their urgency. By comprehensively considering these four types of information, the system can weigh each to-do's importance, willingness to execute, logical precedence, and time pressure, thereby rationally sorting N to-do items and generating an execution order that better aligns with user intent and practical constraints.

[0100] For example, the relationship strength, commitment sensitivity, the degree of association between N to-do items, and the deadline for completion of each to-do item can be input into a pre-trained sorting model, and the sorting model outputs the execution order of the N to-do items.

[0101] In an embodiment of the present application, the location information, activity information, and chat messages of multiple users can be obtained, and then based on the chat messages between the local user and each contact, multiple to-do items of the local user can be identified. Based on the location information, activity information, and chat messages of the multiple users, the relationship strength and commitment sensitivity of the local user to each contact can be determined. Then, the execution order of the multiple to-do items can be generated by combining the relationship strength, commitment sensitivity, the association between each to-do item, and the deadline of each to-do item. In this way, compared with the related art that requires manual identification of to-do items and manual planning of the execution order of each to-do item, the present application can identify to-do items from chat messages between users, and evaluate the relationship strength and commitment sensitivity between users based on the user's activities, location, and interactions between users, and plan and sort the user's to-do items based on the relationship strength and commitment sensitivity between users and the inherent attributes of each to-do item. While ensuring the accuracy of the to-do item planning and sorting, it can also improve the efficiency of to-do item reminders.

[0102] In some embodiments, determining the relationship strength between the local user and each of the contacts based on the location information, the activity information, and the chat message includes:

[0103] Determining the interaction intensity between the local user and each contact during the reference time period based on chat messages between the local user and each contact, activity information, and location information of the local user and each contact during the reference time period, wherein the interaction intensity is used to represent the activity of the social connection between the local user and the contact during the reference time period;

[0104] Determining, based on chat messages between the local user and each contact, location information, activity information, and interaction intensity of the local user and each contact during a reference time period, a comprehensive vector of the local user and each contact during the reference time period, the comprehensive vector being used to characterize social relationship characteristics between the local user and each contact from at least two perspectives;

[0105] The relationship strength between the local user and each contact is determined according to the comprehensive vector.

[0106] In this embodiment, the interaction intensity between the local user and the contact can be used to quantify the level of social connection activity between the local user and the contact during a reference time period. A higher interaction intensity indicates a higher frequency of verbal interaction, participation in offline joint activities, and geographic overlap between the local user and the contact. Therefore, interaction intensity can be used to measure the closeness of the social connection between the local user and the contact.

[0107] Specifically, the interaction intensity between the local user and the contact during a reference time period can be comprehensively calculated based on the chat messages, activity information, and location information between the local user and the contact. The chat messages may include factors such as chat frequency, message length, response speed, and semantic density between the local user and the contact. The higher these factors are, the more active the interaction between the local user and the contact. Activity information is used to determine whether the local user and the contact jointly participate in activities such as meetings and group projects. If the local user and the contact appear in the same interaction, it indicates that the relationship between the two is closer. Location information is used to detect whether the two appear in the same geographical area during the same time period, such as an office or a public place. The more times the two people overlap, the greater the possibility of offline contact between the two people. Therefore, the system can determine the interaction intensity between the local user and the contact based on chat messages, activity information, and location information.

[0108] In some embodiments, determining the interaction intensity between the local user and each contact within the reference time period based on chat messages between the local user and each contact, activity information, and location information of the local user and each contact within the reference time period includes:

[0109] determining, based on the chat messages, a chat rate between the local user and each contact within the reference time period;

[0110] Determining, based on the location information, a location overlap rate between the local user and each contact within the reference time period;

[0111] determining, based on the activity information, a common activity rate between the local user and each contact within the reference time period;

[0112] The interaction intensity between the local user and each contact within a reference time period is determined according to the location overlap rate, the chat rate, and the common activity rate.

[0113] In this embodiment, the chat rate refers to the ratio of the number of days during a reference time period that the local user and the contact actually chatted with each other to the total number of days during that reference time period. For example, user i represents the local user, user j represents the contact, and t represents the reference time period. C(i, j, t) represents the chat rate between the local user and the contact during the reference time period. This represents the ratio of the number of days during which user i and user j chatted with each other during reference time period t to the total number of days. If the reference time period is the 90 days prior to today, and user i and user j chatted on 45 of those 90 days, the chat rate is 0.5.

[0114] The location overlap rate refers to the percentage of days during a reference time period that the local user and the contact appear at the same geographic location. For example, if user i represents the local user, user j represents the contact, and t represents the reference time period, G(i, j, t) represents the location overlap rate between the local user and the contact during the reference time period. This represents the percentage of days during which user i and user j overlapped geographically relative to the total number of days during reference time period t. For example, if the reference time period is the 90 days prior to today, and user i and user j overlapped geographically for 63 days within that 90-day period, the location overlap rate is 0.7.

[0115] The co-activity rate refers to the ratio of the number of days during a reference time period that the local user and the contact jointly participated in the same project or event to the total number of days during that reference time period. For example, user i represents the local user, user j represents the contact, and t represents the reference time period. A(i, j, t) represents the co-activity rate between the local user and the contact during the reference time period. This represents the ratio of the number of days during which user i and user j participated in the same project or event to the total number of days during reference time period t. If the reference time period is the 90 days prior to today, and user i and user j participated in the same project or event on 9 days during that 90-day period, the co-activity rate is 0.1.

[0116] After determining the location overlap rate, chat rate, and common activity rate of the local user and the contact in the reference time period, the interaction intensity in the reference time period can be calculated according to the following formula (1):

[0117] W(i,j,t)=1 / [1+exp(-[α·C(i,j,t)+β·G(i,j,t)+γ·A(i,j,t)])] (1)

[0118] Among them, user i is used to represent the local user, user j can be used to represent the contact, t is used to represent the reference time period, A(i, j, t) is used to represent the common activity rate of the local user and the contact in the reference time period, C(i, j, t) is used to represent the chat rate of the local user and the contact in the reference time period, G(i, j, t) is used to represent the position overlap rate of the local user and the contact in the reference time period, W(i, j, t) is used to represent the interaction intensity between the local user and the contact in the reference time period, and α, β, and γ are the amplification factors corresponding to the common activity rate, position overlap rate, and chat rate, respectively.

[0119] If the chat rate C(i,j,t) is 0.5, the position overlap rate G(i,j,t) is 0.7, the common activity rate A(i,j,t) is 0.1, and α, β, and γ are 2, 3, and 4 respectively, then the interaction intensity can be calculated as 0.9707 according to the above formula (1).

[0120] In the above scheme, by integrating the three key features of chat rate, location overlap rate and common activity rate, the interaction between users at the language, space and behavior levels can be measured, thereby more comprehensively evaluating the overall activity of the communication between the local user and the contact during the reference time period, and accurately calculating the interaction intensity between the local user and the contact during the reference time period.

[0121] In some embodiments, after determining the interaction intensity between the local user and each contact within the reference time period based on the chat messages between the local user and each contact, and the activity information and location information of the local user and each contact within the reference time period, the method further includes:

[0122] Construct a multi-dimensional relationship network model based on the interaction intensity between the local user and each contact within the reference time period;

[0123] The local user corresponds to the central node in the multi-dimensional relationship network model, each contact corresponds to a contact node in the multi-dimensional relationship network model, and each interaction strength corresponds to a connection line between the central node and a contact node.

[0124] In this embodiment, the multidimensional relationship network model is a graph-structured model centered around the user and built on the multidimensional interaction characteristics between the user and their contacts. It represents the strength and structural distribution of the social relationships between the user and all their contacts. This model structures social relationships through nodes and links, facilitating analysis and subsequent reasoning.

[0125] Since there is an association between the local user and each contact, the central node of this multi-dimensional relationship network model refers to the main node representing the local user in the model; the contact node refers to the node representing each specific contact in the graph, and each contact node corresponds to a contact with whom the local user has an interaction record. Each contact node is connected to the central node through a weighted connection line. The weight of the weighted connection line is determined by the interaction strength between the user and the contact corresponding to the contact node.

[0126] Specifically, we can first extract the chat messages, location information, and activity information between the local user and all contacts, and use this information to calculate the interaction intensity between the local user and each contact; then, with the local user as the central node, add the contact node corresponding to each contact to the graph, and establish weighted connection lines between the central node and the contact node based on the interaction intensity, ultimately forming a complete multi-dimensional relationship network model.

[0127] In the above scheme, a multidimensional relationship network model centered on the local user can be constructed based on the multidimensional interaction characteristics. The multidimensional relationship network model can structuredly present the relationship strength distribution between the local user and each contact, thereby providing accurate social basic support for the priority sorting of to-do items related to each contact.

[0128] In some embodiments, determining a comprehensive vector of the local user and each contact during the reference time period based on chat messages between the local user and each contact, location information, activity information, and interaction intensity of the local user and each contact during the reference time period includes:

[0129] determining, based on chat messages between the local user and each contact within the reference time period, a semantic feature vector between the central node and a contact node corresponding to each contact;

[0130] Determining a time series vector between the central node and the contact node corresponding to the contact according to the number of chats between the local user and each contact in the reference time period;

[0131] Determining a spatial vector between the central node and the contact node corresponding to each contact based on the number of times the local user and each contact overlap within the reference time period and the density of common activities; the common activity density is used to represent the frequency with which two users participate in the same activity together;

[0132] Performing vector splicing on the semantic feature vector, the time series vector, and the spatial vector to obtain a spliced vector;

[0133] The concatenated vector is multiplied by the product of the interaction strength to obtain a comprehensive vector of the local user and each contact in the reference time period.

[0134] In this embodiment, the spatial vector is a feature vector representing the frequency of geographic contact between the local user and the contact within a reference time period, reflecting the degree of spatial overlap between the local user and the contact. The time series vector is a feature vector representing the temporal distribution of interactions between the local user and the contact over the entire timeframe. The semantic feature vector is a feature vector obtained by semantically encoding multiple chat messages between the local user and the contact, reflecting the emotional type and commitment intent in the conversation between the local user and the contact.

[0135] Specifically, within the reference time period, the system can obtain at least one chat message between the local user and the contact, extract the semantic features of each message through the text encoding model, and combine the importance of the message and the weighted aggregation of commitment keywords to generate a semantic feature vector.

[0136] Exemplarily, in some embodiments, determining, based on chat messages between the local user and each contact within the reference time period, a semantic feature vector between the central node and a contact node corresponding to each contact includes:

[0137] Perform vector encoding on each of the at least one chat message to obtain the encoding vector e corresponding to each chat message. u ;

[0138] Use the language model to perform sentiment inference on each chat message and obtain the sentiment type of each chat message;

[0139] According to the emotion type of each chat message, determine the emotion vector q(i, j, u) of each chat message;

[0140] Determine the word segmentation vector r(i, j, u) of each chat message based on the commitment keyword in each chat message, where the commitment keyword is a word segmentation that expresses the tone of commitment;

[0141] Determine a text feature vector for each chat message based on the encoding vector, the word segmentation vector, and the sentiment vector of each chat message;

[0142] The text feature vector corresponding to the at least one chat message is subjected to vector fusion to obtain a semantic feature vector between the central node and the contact node corresponding to each contact.

[0143] In this embodiment, the encoding vector corresponding to each chat message is a feature vector obtained by semantically modeling the original chat message using a vector encoding model. This feature vector retains the overall semantic structure of the chat message, including word order, syntax, and contextual semantic relationships. It is itself a high-dimensional feature vector.

[0144] The emotion vector corresponding to a chat message is a feature vector used to characterize the emotion expressed in the chat message. The emotion vector can indicate whether the emotion expressed in the chat message is positive, negative, or neutral. For example, an emotion recognition model can first be used to perform emotion inference on the chat message, outputting an emotion label or intensity score based on the emotional state expressed in the chat message. This emotion label or intensity score is then mapped and converted into a vector of the same dimension as the encoding vector to obtain the emotion vector, facilitating subsequent vector fusion. The emotion recognition model can be a large language model, and the emotion label or intensity score can be used to characterize the emotion type.

[0145] The word segmentation vector corresponding to a chat message is a feature vector used to characterize the degree of commitment intent in the chat message. For example, a commitment keyword recognition model can first be used to identify whether a chat message contains commitment keywords with a commitment tone, such as "I will definitely," "must complete," and "handle it immediately." Based on the type, number, and semantic strength of the commitment keywords, the current chat message is assigned a commitment importance score. This importance score is then converted into a word segmentation vector of the same dimension as the encoding vector through mapping, which is used to characterize the degree of commitment intent expressed in the chat message.

[0146] Based on the above scheme, the encoding vector, word segmentation vector and sentiment vector corresponding to each chat message can be obtained. First, the text feature vector corresponding to each chat message can be determined based on the encoding vector, word segmentation vector and sentiment vector corresponding to each chat message, and then the text feature vectors corresponding to at least one chat message can be fused to obtain a semantic feature vector.

[0147] Specifically, the semantic feature vector can be generated as described in Formula 2:

[0148] S(i,j)=∑(u=1→U)[ω1·q(i,j,u)+ω2·r(i,j,u)]·e u (2)

[0149] Among them, user i is used to represent the local user, user j can be used to represent the contact, U represents the total number of chat messages, u represents the uth chat message, e u It is used to represent the encoding vector corresponding to the u-th chat message, q(i,j,u) is used to represent the sentiment vector corresponding to the u-th chat message, r(i,j,t) is used to represent the word segmentation vector corresponding to the u-th chat message, S(i,j) is used to represent the semantic feature vector of the local user and the contact in the reference time period, ω1 and ω2 are the weight coefficients corresponding to the sentiment vector and the word segmentation vector respectively.

[0150] In this embodiment, semantic encoding, emotion type, and commitment intent are extracted from each chat message. Based on these three factors, a corresponding text feature vector is determined for each chat message. The text feature vectors corresponding to multiple chat messages are then integrated to generate semantic feature vectors for the local user and the contact. This semantic feature vector not only highlights key information with emotional and commitment significance but also improves the accuracy of understanding the user's true intent, thereby enhancing the effectiveness of subsequent to-do item sorting.

[0151] After determining the semantic feature vector between the local user and the contact, the time series vector between the local user and the contact may be determined based on the number of chats between the local user and the contact in a reference time period.

[0152] Specifically, in some embodiments, the reference time period is divided into K sub-time periods of equal length; wherein K is an integer greater than 1;

[0153] The determining, based on the number of chats between the local user and each contact in the reference time period, a time series vector between the central node and the contact node corresponding to the contact, includes:

[0154] Determine the number of chats between the local user and each contact in each sub-time period;

[0155] Determining a time sparsity factor for each sub-time period based on the length of the interval between each sub-time period and a reference sub-time period, wherein the reference sub-time period is the time period in which the system time displayed on the electronic device is located;

[0156] Based on the number of chats and the time sparsity factor, a time series vector between the central node and the contact node corresponding to the contact is determined.

[0157] In this embodiment, a reference time period can be divided into K consecutive sub-time periods of uniform duration. The number of chats between the local user and the contact within each sub-time period is then counted, and a time sparsity factor for each sub-time period is determined. A time series vector between the central node corresponding to the local user and the contact node corresponding to each contact can then be determined based on the number of chats within each sub-time period, the time sparsity factor for each sub-time period, and the number of sub-time periods K. The time series vector is used to characterize the rhythmic characteristics of the interaction between the local user and each contact. The time sparsity factor is a regulating factor for the interaction weights between the local user and the contact within different reference time periods. It can attenuate historical interactions based on the time of occurrence of the reference time period and the temporal distance between the reference sub-time period. This causes interactions farther from the reference sub-time period to have a smaller weight in the overall time series vector, while interactions closer to the reference sub-time period to have a larger weight. This emphasizes the representativeness of recent interactions on the strength of the relationship and suppresses interference caused by frequent interactions in the past despite a long period of no interaction.

[0158] Specifically, the time series vector can be determined according to the first chat times, the time sparse factors, and K using the following formula 3.

[0159] T(i,j)=1 / |K|∑(k=1→K)[d(i,j,k) / (1+η·k 2 )] (3)

[0160] Among them, user i is used to represent the local user, user j can be used to represent the contact, K represents the total number of reference time periods, k represents the kth reference time period among K reference time periods, 1 / (1+ηk 2 ) is used to represent the time sparseness factor corresponding to the k-th reference time period, η represents the amplification factor for the attenuation of long time spans, d(i,j,k) is used to represent the number of chats in the k-th reference time period, and T(i,j) is used to represent the time series vector of the local user and the contact in the reference time period.

[0161] For example, the reference time period can be set to 1 day, or the most recent 90 days. Then, the sub-time period can be divided into 90 reference time periods, with today as the reference sub-time period, and k = 1 representing the most recent reference time period, namely yesterday. d(i, j, k = 1) represents the total number of chats between local user i and contact j yesterday, which can be 20. Similarly, d(i, j, k = 90) represents the number of chats between local user i and contact j 90 days prior to today, which can be 100. The temporal sparsity factor for k = 1 can be set to 0.5, and d(i, j, k = 1) multiplied by 0.5 is 10. The temporal sparsity factor for k = 90 can be set to 0.0001, and d(i, j, k = 90) multiplied by 0.0001 is 0.01. Therefore, although there were 100 interactions 90 days ago, their impact on the overall features is far less than the 20 interactions that occurred yesterday, reflecting that the temporal sparsity factor gives greater attention to recent interactions.

[0162] In this approach, the temporal sparsity factor can attenuate and weight the number of chats in different sub-time periods, effectively highlighting the representativeness of recent interactions and weakening the interference of interactions long ago on the judgment of the relationship between the local user and the contact, thereby constructing a time series vector that is more in line with the actual interaction rhythm.

[0163] After determining the semantic feature vector and time series vector between the local user and the contact, the number of times the local user and the contact overlap within a reference time period can be counted, and the spatial vector between the local user and the contact can be determined based on the number of times the position overlaps.

[0164] Specifically, in some embodiments, the user's activity area includes P sub-areas;

[0165] The determining, based on the number of position overlaps and the common activity density between the local user and each contact within the reference time period, of a space vector between the central node and the contact node corresponding to each contact includes:

[0166] Determining the number of times the positions of the local user and each contact overlap in each sub-area within the reference time period based on the number of times the local user and each contact appear in the same sub-area within the same time period;

[0167] determining, according to the number of times the local user and each contact jointly participate in an activity, a density of common activities between the local user and each contact in each sub-area within the reference time period;

[0168] A space vector between the central node and the contact node corresponding to each contact is determined according to the number of position overlaps and the common activity density of each sub-region.

[0169] In this embodiment, the common activity density refers to the frequency with which the local user and the contact participate in activities together within the corresponding sub-area, reflecting the density of the local user and the contact participating in the same event simultaneously within the sub-area. The number of location overlaps refers to the total number of times the local user and the contact appear in the same sub-area within a certain time range, indicating the frequency of their spatial contact.

[0170] First, the overall activity area of the local user and at least one contact can be obtained, and then the obtained area can be divided into P sub-areas. The number of position overlaps and the number of position overlaps of the local user and the contact in each sub-area can be counted respectively, and the spatial vector can be determined based on the number of position overlaps and the common activity density of each sub-area.

[0171] For example, the activity area can be divided into square or rectangular areas of fixed size to obtain P sub-areas. For example, the activity area is divided into grids with a side length of 100 meters, and each grid represents a sub-area. If the activity area is a city, the city can be divided into P sub-areas of 100 meters × 100 meters, numbered A1, A2, A3...AP. If the local user and the contact appear in area A5 on the same day, it is considered that they have a location overlap in the sub-area; if the local user and the contact participate in an event in area A6 on the same day, it is considered that they have a joint activity in the sub-area, where a joint activity refers to the behavior of the local user and the contact participating in the same event or being in the same activity scene within the same sub-time period.

[0172] Specifically, the space vector can be determined according to the number of position overlaps of each sub-region and the common activity density using the following formula 4:

[0173] P(i,j)=∑(p=1→P)[h(i,j,p)·log(1+R(p))] (4)

[0174] Among them, user i is used to represent the local user, user j can be used to represent the contact, P represents the total number of sub-areas, p represents the p-th sub-area among P sub-areas, R(p) represents the common activity density in the p-th sub-area, h(i,j,p) is used to represent the number of times the position of the p-th sub-area overlaps, and P(i,j) is used to represent the spatial vector of the local user and the contact within the reference time period.

[0175] Through the above method, the user's activity space can be divided into multiple sub-areas, and the number of location overlaps and the density of common activities in each first area can be combined to accurately characterize the contact intensity and behavioral coordination between the local user and the contacts in the geographic space, thereby constructing a more accurate spatial vector and effectively improving the accuracy of subsequent to-do reasoning.

[0176] In this embodiment, after obtaining the semantic feature vector, time series vector and spatial vector between the local user and the contact, the semantic feature vector, time series vector and spatial vector can be spliced to obtain a spliced vector; and the product of the spliced vector and the interaction intensity is multiplied to determine the comprehensive vector.

[0177] Specifically, the comprehensive vector Z(i,j)=[S(i,j),T(i,j),P(i,j)]·W(i,j,t)

[0178] Among them, user i is used to represent the local user, user j can be used to represent the contact, P(i,j) is used to represent the spatial vector of the local user and the contact in the reference time period, T(i,j) is used to represent the time series vector of the local user and the contact in the reference time period, S(i,j) is used to represent the semantic feature vector of the local user and the contact in the reference time period, and W(i,j,t) is used to represent the interaction intensity between the local user and the contact in the reference time period.

[0179] In some embodiments, determining the relationship strength between the local user and each contact according to the comprehensive vector includes:

[0180] determining a characteristic component of each vector dimension in the comprehensive vector;

[0181] Determining a feature interaction enhancement value of the comprehensive vector according to a collaborative weight coefficient of each feature component;

[0182] The feature interaction enhancement value is mapped to a reference value interval to obtain the relationship strength between the local user and each contact.

[0183] In this embodiment, relationship strength refers to a numerical indicator used by the system to comprehensively assess the closeness of the relationship between the user and the contact based on their interactions across multiple dimensions, including semantics, time, and space. This indicator represents the strength of the social connection between the two. The higher the relationship strength, the stronger the social relationship between the user and the contact.

[0184] Specifically, the characteristic components of each dimension in the comprehensive vector can be determined. By combining the characteristic components between different dimensions in pairs and introducing cross-dimensional weights to perform weighted fusion on the characteristic components of the pairwise combinations, we can finally obtain a multi-perspective feature interaction enhancement value that reflects the synergy of multi-perspective features. This feature interaction enhancement value can be used to quantify the synergistic enhancement effect of different perspective features in relationship modeling.

[0185] After obtaining the feature interaction enhancement value, the feature interaction enhancement value can be mapped to the final relationship strength through the sigmoid function, and the output relationship strength value is between the reference value range of 0 and 1.

[0186] Specifically, the following formula 5 can be used to determine the feature interaction enhancement value according to the collaborative weight coefficient between each feature component; and the relationship strength can be determined according to the feature interaction enhancement value.

[0187] A(i,j)=∑(p=1→m)∑(q=1→m)[z p (i,j)·w a (p,q)·zq(i,j)]

[0188] R s (i,j)=1 / [1+exp(-[α·A(i,j)+b])] (5)

[0189] Among them, user i is used to represent the local user, user j can be used to represent the contact, and Z(i,j) is used to represent the comprehensive vector between the local user and the contact. Then, z p (i, j) can represent the characteristic component of the comprehensive vector in the pth dimension, zq(i, j) represents the characteristic component of the comprehensive vector in the qth dimension, w a (p,q) represents the collaborative weight coefficient between the p-th dimension and the q-th dimension, A(i,j) represents the feature interaction enhancement value, m is the total number of vector dimensions of the feature components of Z(i,j), α and b are learnable feature parameters, R s (i,j) is the relationship strength.

[0190] In the above method, by extracting the characteristic components of each vector dimension in the comprehensive vector and modeling its collaborative weight coefficient, a feature interaction enhancement value that can better reflect the multi-perspective feature linkage effect can be obtained, and the relationship strength between the local user and the contact can be obtained based on the feature interaction enhancement value, thereby ensuring the accuracy of the relationship strength in depicting the real social relationship between the local user and the contact and the robustness of the judgment.

[0191] In some embodiments, determining the commitment sensitivity between the local user and each contact based on the location information, the activity information, and chat messages between the local user and each contact includes:

[0192] Determining, based on chat messages between the local user and each contact, a communication depth between the local user and each contact within a reference time period, wherein the communication depth is used to characterize the communication density and information richness between the local user and each contact in the conversation;

[0193] Determining, based on chat messages between the local user and each contact, a commitment word frequency of the local user and each contact within a reference time period, where the commitment word frequency represents the frequency of words with a commitment tone appearing in conversations between the user and the contact;

[0194] The commitment sensitivity of the local user and each contact is determined according to the commitment word frequency and the communication depth.

[0195] In this embodiment, commitment sensitivity refers to the intensity and certainty of the user's commitment to an item during interaction, and is used to evaluate the user's subjective intention and willingness to complete the item. Therefore, the stronger the user's commitment sensitivity to the contact, the higher the possibility and priority of completing the to-do item promised by the user to the contact.

[0196] Since chat messages include chat messages between the local user and contacts, the communication depth between the local user and contacts can be determined based on characteristics such as chat message length, message frequency, and reply speed. The communication depth is used to characterize the communication density and information richness between the local user and each contact in the conversation.

[0197] Specifically, in some embodiments, determining the communication depth between the local user and each contact within a reference time period based on chat messages between the local user and each contact includes:

[0198] Determining, based on the chat messages, the average message length, average message frequency, and timely message reply rate of the local user and each contact within the reference time period;

[0199] The communication depth between the local user and each contact within a reference time period is determined based on the average message length, the average message frequency, and the timely message reply rate.

[0200] In this embodiment, the average message length refers to the average number of characters in each chat message between the local user and the contact during a reference time period. The average message length can be calculated by adding the number of characters in all chat messages between the local user and the contact and dividing the sum by the total number of chat messages. For example, the average message length can be 10 or 5. A longer average message length indicates richer content in the communication between the local user and the contact.

[0201] The average message frequency refers to the average frequency of chat messages generated by the local user and the contact per unit time within a reference time period, where the unit time can be 1 day or 1 hour. The average message frequency can be obtained by dividing the total number of all chat messages between the local user and the contact by the number per unit time. For example, if the reference time period is 5 days, the unit time can be 1 day, the number of chat messages generated on the first day is 10, the number of chat messages generated on the second day is 2, the number of chat messages generated on the third day is 3, the number of chat messages generated on the fourth day is 0, and the number of chat messages generated on the fifth day is 5. Then the total number of chat messages for these five days is 20, and the total number of days is 5, so the average message frequency is 4 messages / day. The higher the average message frequency, the more frequent the interaction between the local user and the contact during the reference time period.

[0202] The timely response rate refers to the proportion of all chat messages that receive a reply within a preset time threshold. For example, if the time threshold is set to 1 minute and there are 100 chat messages, and 60 of them receive a reply within 1 minute, the timely response rate is 0.6. A higher timely response rate indicates more timely communication and more proactive responses.

[0203] After determining the average message length, average message frequency, and timely message reply rate, the depth of communication between the local user and the contact can be determined based on the following formula 6:

[0204] D(i,j)=1 / [1+exp(-[α1·M a vg (i,j)+α2·Freq(i,j)+α3·Rpl(i,j)])] (6)

[0205] Among them, user i is used to represent the local user, user j can be used to represent the contact, D(i,j) is used to represent the communication depth between the local user and the contact, M a vg (i, j) represents the average message length between the user and the contact during the reference time period, Freq(i, j) represents the average message frequency between the user and the contact during the reference time period, and Rpl(i, j) represents the timely response rate of messages between the user and the contact during the reference time period. α1, α2, and α3 represent the amplification factors for the average message length, average message frequency, and timely response rate, respectively. Communication depth D(i, j) ranges from 0 to 1. A value closer to 1 indicates more frequent conversations, more lengthy messages, and more timely responses between the user and the contact during the reference time period.

[0206] For example, if the average message length between the local user and the contact is 10, the average message frequency is 8, and the timely message reply rate is 0.6, and the amplified sparsity α1, α2, and α3 are set to 0.01, 0.1, and 1 respectively, then the communication depth between the local user and the contact can be calculated to be 0.818.

[0207] For another example, if the average message length between the local user and the contact is 30, the average message frequency is 0.1, and the timely message reply rate is 0, and the amplified sparsity α1, α2, and α3 are set to 0.01, 0.1, and 1 respectively, then the communication depth between the local user and the contact can be calculated to be 0.577.

[0208] In the above scheme, key indicators such as the average message length, frequency and response time between the local user and the contact can be determined based on chat messages, and the communication depth between the two people can be comprehensively calculated through these key indicators, so that the communication depth can fully reflect the richness, frequency and timeliness of user exchanges, thereby more accurately reflecting the closeness of interaction between users.

[0209] After calculating the depth of communication between the user and the contact, the frequency of commitment keywords can be determined based on the number of times the commitment keyword appears in chat messages. Commitment keyword frequency refers to the frequency with which the user uses commitment keywords with a commitment tone in conversations with the contact during a reference time period. Commitment keyword frequency can be used to measure the clarity and strength of the user's expressed task intent. Commitment keywords can include "must" and "must."

[0210] In some embodiments, determining the commitment word frequency of the local user and each contact within a reference time period based on chat messages between the local user and each contact includes:

[0211] Determining a commitment message in at least one chat message between the local user and each contact; the commitment message is a chat message including an indicator variable of a commitment sentence;

[0212] Determining the commitment strength of each commitment message based on the commitment keywords in each commitment message, wherein the commitment strength is used to represent the degree of expression of the commitment tone in a commitment message;

[0213] The commitment word frequency of the local user and each contact is determined according to the commitment strength of each commitment message, the generation time of each commitment message and the reference generation time, wherein the reference generation time is the system time for determining the commitment word frequency.

[0214] In this embodiment, for each of the at least one chat message, if the chat message contains an indicator variable for a commitment sentence, the chat message can be determined to be a commitment message. The indicator variable for a commitment sentence is a variable used to identify whether a chat message contains a sentence with a commitment tone or structure, and is typically represented by a Boolean value. For example, if a message contains a commitment expression such as "I will definitely" or "I will give it to you tomorrow," the variable is 1, and otherwise it is 0.

[0215] After determining the commitment message in at least one chat message, the commitment keywords in each commitment message can be identified, and then the commitment strength of each commitment message can be determined based on the commitment keywords, wherein the commitment strength is used to characterize the degree of expression of the commitment tone in a commitment message, reflecting how strong the commitment intention expressed by the user in the language is. The higher the commitment strength, the firmer the tone and the clearer the commitment. For example, the commitment strength corresponding to the commitment keywords can be divided into three levels. If a commitment message contains highly certain commitment keywords such as "certain", "guaranteed", etc., the commitment strength of the commitment message can be determined to be 3; if a commitment message only contains generally certain commitment keywords, such as "OK", "received", etc., the commitment strength of the commitment message can be determined to be 2; if a commitment message contains uncertain commitment keywords, such as "try your best", "strive for", "not guaranteed", etc., the commitment strength of the commitment message can be determined to be 2.

[0216] Determined to be 1.

[0217] After determining the commitment strength of each commitment message, the commitment word frequency may be determined based on the following formula 7 according to the commitment strength of each commitment message, the generation time of each commitment message, and the reference generation time:

[0218] R(i,j)=[∑(u=1→U)[k(i,j,u)·p(i,j,u)·exp(-δ·(t max -t u ))]] / Zo

[0219] Zo=∑(u=1→U)[exp(-δ·(t ma xt u ))] (7)

[0220] Among them, user i is used to represent the local user, user j can be used to represent the contact, R(i,j) is used to represent the commitment word frequency between the local user and the contact, U represents the total number of chat messages, u represents the u-th chat message, k(i,j,u) is used to represent whether the u-th chat message is a commitment message. If the u-th chat message is a commitment message, k(i,j,u) is 1; if the u-th chat message is not a commitment message, k(i,j,u) is 0. p(i,j,u) is used to represent the commitment strength of the u-th chat message, t u Used to represent the generation time of the u-th chat message, t max It is used to represent the reference generation time, Zo is the time decay coefficient normalization factor, and δ is the time decay factor. R(i,j) represents the commitment word frequency between nodes i and j. The commitment word frequency range is [0, positive upper limit]. A larger commitment word frequency value indicates more and more high-importance commitment statements in recent conversations.

[0221] For example, the local user sends a chat message to the contact within the reference time period. The content of the chat message is "OK, I will definitely submit the PPT next Wednesday." Since this sentence is a chat sentence, its corresponding k(i, j, u) is 1. At the same time, the chat message contains the highly certain modal particle "definitely", so the commitment strength p(i, j, u) of the chat message is 3. If the chat message was sent 1 minute ago, then exp(-δ·(t max -t u ))=0.549, and the commitment word frequency of this chat message is calculated to be 1.647.

[0222] For another example, if the local user sends 10 chat messages to the contact within the reference time period, but there is no chat message with a commitment sentence in the 10 chat messages, the commitment word frequency between the local user and the contact is 0.

[0223] In the above method, the commitment word frequency indicator can be constructed by combining the tone intensity, time distribution and current time of the commitment message, so that the commitment word frequency can more accurately capture the commitment intention expressed by the user in the recent conversation and its strength changes, thereby improving the timeliness and credibility of the recognition of commitment behavior.

[0224] After obtaining the communication depth and commitment word frequency, the commitment sensitivity between the local user and the contact can be determined based on the communication depth and commitment word frequency.

[0225] In some embodiments, determining the commitment sensitivity of the local user and each contact based on the commitment word frequency and the communication depth includes:

[0226] Determining commitment credibility scores of the local user and each contact based on the commitment word frequency and the communication depth;

[0227] The commitment credibility score is mapped to a reference value interval to obtain the commitment sensitivity of the local user and each contact.

[0228] In this embodiment, commitment word frequency refers to the frequency of commitment keywords with a commitment tone appearing in conversations between the user and the contact during a reference time period. Commitment word frequency can be used to measure the clarity and intensity of the user's expressed task intent. Communication depth, on the other hand, represents the density and information richness of the communication between the user and the contact.

[0229] The commitment credibility score can be obtained through weighted modeling based on pragmatic features such as communication depth and commitment word frequency. Therefore, the commitment credibility score can be used to comprehensively measure the intensity of the local user's tone of commitment and the credibility of the fulfillment intention in the interaction between the local user and the contact. The higher the commitment credibility score, the clearer the commitment, the stronger the willingness to execute, and the more suitable it is to be included in the system management as a formal to-do item.

[0230] Specifically, based on the following formula 8, the credibility score can be determined according to the commitment word frequency and communication depth, and the commitment credibility can be mapped to a preset reference value range to obtain the commitment sensitivity, where the preset reference value range can be 0-1.

[0231] C(i,j)=α1·D(i,j)+α2·R(i,j)+α3·[D(i,j)·R(i,j)]

[0232] σ(i,j)=1 / [1+exp(-C(i,j))] (8)

[0233] Among them, user i is used to represent the local user, user j can be used to represent the contact, D(i,j) is used to represent the communication depth between the local user and the contact, R(i,j) is used to represent the commitment word frequency between the local user and the contact, C(i,j) is used to represent the commitment credibility score between the local user and the contact within the reference time period, σ(i,j) is used to represent the commitment sensitivity between the local user and the contact within the reference time period, and α1, α2, and α3 are learnable amplification coefficients.

[0234] In the above method, the commitment credibility score can be calculated by comprehensively considering the commitment word frequency and communication depth, and then standardized into commitment sensitivity within a unified range. Based on commitment sensitivity, the system can more stably and objectively evaluate the strength of the task intention expressed by the user in the conversation, and accordingly improve the accuracy of to-do item recognition between users.

[0235] In this way, we can extract communication depth and commitment word frequency from chat messages and combine them to calculate commitment sensitivity. Note that commitment sensitivity can more comprehensively reflect the tone intensity and communication intention of users expressing commitment in conversations, thereby improving the authenticity of the recognition of to-do items and the accuracy of the recognition of the local user's willingness to execute to-do items.

[0236] Specifically, in some embodiments, determining the execution order of the N to-do items based on the relationship strength, commitment sensitivity, the degree of association of the N to-do items, and the deadline for completion of each to-do item includes:

[0237] determining a base score for each to-do item based on the relationship strength, the commitment sensitivity, and the deadline for completion of each to-do item;

[0238] Determine the synergy between every two to-do items based on the synergy factor between the N to-do items and the basic score of each to-do item;

[0239] Determine a total score for each to-do item based on the base score and the amount of collaboration for each to-do item;

[0240] The execution time of the N to-do items is sorted in descending order of the total scores to obtain the execution order of the N to-do items.

[0241] In this embodiment, the basic score refers to the priority score calculated for each to-do item individually without considering the relationship between the to-do items. It is based solely on factors such as the relationship strength, commitment sensitivity, and deadline corresponding to the to-do item, and is used to measure the importance and urgency of the to-do item itself.

[0242] The synergy refers to the strength of the synergy weight coefficient between a to-do item and other to-do items in execution. It can reflect factors such as whether the to-do item is close to other to-do items in time, whether they come from the same contact person, and whether they can be combined for execution. It is an auxiliary weight that reflects the composability and overall execution efficiency between to-do items.

[0243] Therefore, for any first to-do item that the local user has promised to the contact, the basic score of the to-do item can be calculated first. The calculation formula of the basic score can be the following formula 9:

[0244] A i =R s (i,j)·exp(-β(d i -τ)·σ(i,j)) (9)

[0245] Among them, user i is used to represent the local user, user j can be used to represent the contact, σ(i, j) is used to represent the commitment sensitivity between users corresponding to the first to-do item, R s (i, j) is used to represent the relationship strength between users corresponding to the first to-do item, τ is used to represent the current moment, and d i It is used to represent the deadline of the first to-do item, β is the time decay coefficient, and A i It is used to represent the basic score of the first to-do item.

[0246] After determining the basic score of the first to-do item, the synergy between the first to-do item and other to-do items can also be determined. For each to-do item other than the first to-do item in at least one to-do item, if the first to-do item and the to-do item have the same relevant contact person, and the time interval between the deadline of the first to-do item and the deadline of the to-do item is less than a preset time interval threshold, then it can be considered that there is an association between the first to-do item and the to-do item, and the synergy factor between the first to-do item and the to-do item is 1; otherwise, the synergy factor between the first to-do item and the to-do item is 0. All the synergy factors of the first to-do item are added together to obtain a first sum value, and then the first sum value is multiplied by the basic score of the first to-do item to obtain the synergy amount of the first to-do item.

[0247] For example, let the time interval threshold be 24 hours. If the contact is the superior of the local user and assigns two to-do items to the local user, the first item is "design an event poster" and the second item is "write an event invitation letter", and both to-do items need to be completed tomorrow, then, since the time interval between the deadlines of these benign to-do items is less than the preset time interval threshold, and the relevant contacts of these two to-do items are both contacts, then the synergy factor g of these two to-do items is ij =1.

[0248] After obtaining the basic score and synergy amount of the first to-do item, if the synergy amount of the first to-do item is 0, it means that there is no synergy weight coefficient between the first to-do item and any other to-do items. In this case, the basic score of the first to-do item can be directly determined as the total score of the first to-do item. If the synergy amount of the first to-do item is not 0, the total score of the first to-do item can be calculated using the following formula 10:

[0249] P i =A i ·[1+S(H i )] (10)

[0250] Among them, user i is used to represent the local user, A iIt is used to represent the basic score of the first to-do item, H i Used to characterize the coordination amount of the first to-do item, S(H i )=1 / [1+exp(-α·H i )] is based on H i The Sigmoid function, α is the amplification factor, which can enhance the priority coupling strength between highly correlated to-do items.

[0251] For example, if the first to-do item is sent by the contact to the local user, and the contact has not sent any other to-do items to the local user, there is no collaborative weight coefficient between the first to-do item and other to-do items. If the deadline of the first to-do item is 240 hours later, the relationship strength between the local user and the contact is 0.7, and the commitment sensitivity is 0.6, the basic score of the first to-do item, that is, the total score, is 0.33 when substituted into the above formula nine.

[0252] If the contact now publishes a second to-do item to the local user, and the deadline for the second to-do item is the same as that of the first to-do item, then there is a collaborative weight coefficient between the first to-do item and the second to-do item, and the collaborative amount H of the first to-do item can be calculated. i =0.33, substituting it into the above formula 10, we can calculate that the total score of the first to-do item is 0.52, which is 1.58 times the original score.

[0253] Through this approach, we can combine factors like relationship strength, commitment sensitivity, and deadlines for each to-do item to calculate a basic score. We then combine the synergy weight coefficients between to-do items to construct a synergy measure. This then generates a unified overall score for each to-do item, which is then sorted by that score. To-do items with higher overall scores are assumed to be more important and more likely to be executed. Therefore, this sorting method better reflects users' true intentions, thereby improving the completion rate of tasks.

[0254] In some embodiments, after creating a to-do reminder, wherein the reminder content of the to-do reminder is the to-do item, the method further includes:

[0255] Inputting a to-do item into a task splitting model, and if the complexity of the to-do item is greater than a preset threshold, the task splitting model splits the to-do item into at least two subtasks and outputs an estimated completion time for each subtask;

[0256] At the task reminder moment of each subtask, the second prompt information corresponding to each subtask is displayed, and the second prompt information is used to prompt the user to process the subtask; the task reminder moment is the second time before the expected completion time of each subtask.

[0257] In this embodiment, the system can treat each to-do item with a complexity greater than a preset threshold as a main task and split the main task into multiple subtasks. Specifically, the system can break down the main task layer by layer, that is, the main task is layered and refined according to the execution logic, and then split into multiple smaller, more specific, and easier-to-execute subtasks, until each subtask meets the operability and time control requirements. The preset threshold can be a duration threshold or a step threshold; for example, the duration threshold can be 2 hours or 24 hours, and the step threshold can be 5 steps or 6 steps.

[0258] For example, during the decomposition of the main task, a task tree diagram corresponding to the main task can be constructed, the root node of the task tree diagram is the main task, and starting from the main task as the root node, each lower-level task is a child node of the upper-level task. Therefore, the task nodes at each level can be checked layer by layer to see if they are too complex. For example, if the execution steps of the subtask of a task node are greater than the preset step threshold, where the step threshold can be 6 steps or 5 steps, or the execution time of the subtask is greater than the preset task time threshold, which can be 2 hours or 24 hours, then the task node can be considered too complex. If a task node is considered too complex, the system will continue to decompose it into multiple smaller subtask nodes until the subtasks corresponding to each task node are simple enough to facilitate actual execution.

[0259] For example, in the process of breaking down specific to-do items, the system can use the large language model to understand the intention of the to-do items, and gradually transform the original to-do items into clearer and more actionable sub-task groups. For example, if a to-do item is "prepare a product report", the system may break it down into the following sub-tasks: "determine the report outline", "organize project data", "write PPT content", "optimize typesetting", and "demonstration rehearsal", etc., and the large language model can output the estimated execution time of each sub-task. If the sub-task "write PPT content" is still too complicated, the system will further break it down into multiple sub-tasks such as "write cover", "write analysis content" and "write conclusion" in the next layer, layer by layer, until all task nodes meet the complexity requirements.

[0260] During each round of subtask decomposition, the generated subtasks are recorded in a task set, which can also include information such as the task description, estimated execution duration, estimated completion time, and required resources for each subtask. If a subtask remains complex, the system will enter the next round of recursive decomposition until the minimum execution unit set by the system is reached. This approach not only ensures the clarity of the task structure but also facilitates the execution scheduling and reminder generation of subsequent subtasks. Ultimately, all decomposed subtasks will form a complete set of executable tasks, forming a refined execution plan for the original to-do item.

[0261] After the system breaks down the to-do list into multiple subtasks and determines the estimated completion time of each subtask, the system can determine the second time before the estimated completion time of each subtask as the task reminder time of the subtask, and display the second prompt information corresponding to the subtask at the task reminder time of each subtask to remind the user that the estimated completion time of the subtask is approaching.

[0262] Through the above method, the to-do list can be broken down into multiple subtasks, and the estimated completion time of each subtask can be set separately to make the task more specific and clear. In addition, an advance reminder can be set before the estimated completion time of each subtask to prevent users from forgetting the subtask, thereby improving the probability of completing the subtask on time and the overall execution efficiency of the to-do list.

[0263] In some embodiments, at the task reminder moment of each subtask, a first control and a second control are displayed, wherein the first control is used to indicate that the subtask corresponding to the second prompt information has been completed, and the second control is used to indicate that the subtask corresponding to the second prompt information has not been completed; the at least two subtasks include the first subtask and the second subtask;

[0264] After displaying the second prompt information corresponding to each subtask, the method further includes:

[0265] When a selection input is received on the first control at or before the estimated completion time of the first subtask, in response to the selection input, hiding the second prompt information corresponding to the first subtask;

[0266] When a selection input to the second control is received after the estimated completion time of the second subtask, the estimated completion time of the second subtask and subtasks whose estimated completion time is later than the estimated completion time of the second subtask are updated in response to the selection input.

[0267] In this embodiment, for each subtask, the system will display the second prompt information corresponding to the subtask starting from the task reminder time of the second time before the expected completion time of the subtask to guide the user to perform the task as planned. Figure 3 In the embodiment shown, if the system time displayed in the electronic device is 12:00, and the estimated completion time of the subtask is also 12:00, then the system will display a second prompt message 301 of the to-do item, and the second prompt message includes a first control 302 indicating that the subtask has been completed, and a second control 303 indicating that the subtask has not been completed.

[0268] Since the first control is used to indicate that the user has completed the subtask corresponding to the second prompt information, if the user selects and inputs the first control before the expected completion time or at the expected completion time, the system will consider that the user has completed the subtask on time, so there is no need to continue to display the second reminder information, thereby hiding the currently displayed second prompt information to avoid interference with the user.

[0269] Since the second control is used to indicate that the user has not completed the subtask corresponding to the second prompt information, if the user selects and inputs the second control after the estimated completion time of the subtask, the system will consider that the subtask has not been completed on time, and the system will automatically readjust the estimated completion time of all unfinished subtasks.

[0270] In the above scheme, an interactive display control can be set to facilitate the user to cancel the display of the second prompt information in time by selecting the first control when the user completes the subtask on time, thereby reducing the interference caused by the display of outdated prompt information to the user. When a subtask is not completed on time, the global adjustment of all subtasks can be automatically triggered by selecting the corresponding second control, thereby improving the flexibility and real-time performance of task management.

[0271] Figure 4 is a structural diagram of a to-do list reminder device 400 provided in another embodiment of the present application, such as Figure 4 As shown, the to-do list reminder device may include:

[0272] Display module 401 is configured to input a chat message between the local user and the contact into a language macromodel, and display a first prompt message when the language macromodel identifies a to-do item in the chat message; the first prompt message includes: a description of the identified to-do item and a deadline for completion of the item;

[0273] The creation module 402 is used to create a to-do reminder, where the reminder content of the to-do reminder is the to-do item.

[0274] In this application, the to-do items in the chat message can be automatically identified through the language model, and the prompt information of each to-do item can be displayed, and then a to-do reminder for the to-do item can be created. Compared with the manual identification of to-do items and manual creation of to-do reminders in the related art, the process of item identification and reminder creation is greatly simplified. In this way, the automatic identification of to-do items and generation of to-do reminders through the language model reduces the user's operation steps and improves the efficiency of to-do reminders.

[0275] In another optional example, the apparatus further includes:

[0276] A display module, used for displaying a cancel control;

[0277] A creation module, configured to create a to-do reminder when the language macro model identifies a to-do item in the chat message;

[0278] The deleting module is configured to delete the to-do reminder when a selection input of the cancel control is received within a first duration of displaying the cancel control.

[0279] In another optional example, the creation module is further configured to:

[0280] In a case where a confirmation input for the to-do reminder creation confirmation information is received, a to-do reminder is created in response to the confirmation input.

[0281] In another optional example, the apparatus further includes:

[0282] An input module, configured to input chat messages between the local user and at least one contact into the language model;

[0283] An acquisition module, configured to acquire location information and activity information of a local user and at least one contact;

[0284] A determination module is used to determine the execution order of the N to-do items based on the location information, the activity information and the chat messages between the local user and the at least one contact.

[0285] In another optional example, the determining module is specifically configured to:

[0286] Determining, based on the location information, the activity information, and chat messages between the local user and each contact, the relationship strength and commitment sensitivity between the local user and each contact, wherein the relationship strength is used to represent the closeness of the social connection between the local user and each contact, and the commitment sensitivity is used to represent the priority of the task committed by the local user to the contact;

[0287] An order of executing the N to-do items is determined according to the relationship strength, commitment sensitivity, the degree of association among the N to-do items, and the deadline for completing each to-do item.

[0288] In another optional example, the determining module is further configured to:

[0289] Determining the interaction intensity between the local user and each contact during the reference time period based on chat messages between the local user and each contact, activity information, and location information of the local user and each contact during the reference time period, wherein the interaction intensity is used to represent the activity of the social connection between the local user and the contact during the reference time period;

[0290] Determining, based on chat messages between the local user and each contact, location information, activity information, and interaction intensity of the local user and each contact during a reference time period, a comprehensive vector of the local user and each contact during the reference time period, the comprehensive vector being used to characterize social relationship characteristics between the local user and each contact from at least two perspectives;

[0291] The relationship strength between the local user and each contact is determined according to the comprehensive vector.

[0292] In another optional example, the determining module is specifically configured to:

[0293] determining, based on the chat messages, a chat rate between the local user and each contact within the reference time period;

[0294] Determining, based on the location information, a location overlap rate between the local user and each contact within the reference time period;

[0295] determining, based on the activity information, a common activity rate between the local user and each contact within the reference time period;

[0296] The interaction intensity between the local user and each contact within a reference time period is determined according to the location overlap rate, the chat rate, and the common activity rate.

[0297] In another optional example, the apparatus further includes:

[0298] A construction module, used to construct a multi-dimensional relationship network model based on the interaction intensity between the local user and each contact within a reference time period;

[0299] The local user corresponds to the central node in the multi-dimensional relationship network model, each contact corresponds to a contact node in the multi-dimensional relationship network model, and each interaction strength corresponds to a connection line between the central node and a contact node.

[0300] In another optional example, the determining module is specifically configured to:

[0301] determining, based on chat messages between the local user and each contact within the reference time period, a semantic feature vector between the central node and a contact node corresponding to each contact;

[0302] Determining a time series vector between the central node and the contact node corresponding to the contact according to the number of chats between the local user and each contact in the reference time period;

[0303] Determining a spatial vector between the central node and the contact node corresponding to each contact based on the number of times the local user and each contact overlap within the reference time period and the density of common activities; the common activity density is used to represent the frequency with which two users participate in the same activity together;

[0304] Performing vector splicing on the semantic feature vector, the time series vector, and the spatial vector to obtain a spliced vector;

[0305] The concatenated vector is multiplied by the product of the interaction strength to obtain a comprehensive vector of the local user and each contact in the reference time period.

[0306] In another optional example, the determining module is specifically configured to:

[0307] Performing vector encoding on each of the at least one chat message to obtain an encoding vector corresponding to each chat message;

[0308] Use the language model to perform sentiment inference on each chat message and obtain the sentiment type of each chat message;

[0309] Determine the emotion vector of each chat message according to the emotion type of each chat message;

[0310] Determine a word segmentation vector for each chat message based on a commitment keyword in each chat message, where the commitment keyword is a word segmentation that expresses a commitment tone;

[0311] Determine a text feature vector for each chat message based on the encoding vector, the word segmentation vector, and the sentiment vector of each chat message;

[0312] The text feature vector corresponding to the at least one chat message is subjected to vector fusion to obtain a semantic feature vector between the central node and the contact node corresponding to each contact.

[0313] In another optional example, the determining module is specifically configured to:

[0314] Determine the number of chats between the local user and each contact in each sub-time period;

[0315] Determining a time sparsity factor for each sub-time period based on the length of the interval between each sub-time period and a reference sub-time period, wherein the reference sub-time period is the time period in which the system time displayed on the electronic device is located;

[0316] Based on the number of chats and the time sparsity factor, a time series vector between the central node and the contact node corresponding to the contact is determined.

[0317] In another optional example, the determining module is specifically configured to:

[0318] Determining the number of times the positions of the local user and each contact overlap in each sub-area within the reference time period based on the number of times the local user and each contact appear in the same sub-area within the same time period;

[0319] determining, according to the number of times the local user and each contact jointly participate in an activity, a density of common activities between the local user and each contact in each sub-area within the reference time period;

[0320] A space vector between the central node and the contact node corresponding to each contact is determined according to the number of position overlaps and the common activity density of each sub-region.

[0321] In another optional example, the determining module is specifically configured to:

[0322] determining a characteristic component of each vector dimension in the comprehensive vector;

[0323] Determining a feature interaction enhancement value of the comprehensive vector according to a collaborative weight coefficient of each feature component;

[0324] The feature interaction enhancement value is mapped to a reference value interval to obtain the relationship strength between the local user and each contact.

[0325] In another optional example, the determining module is specifically configured to:

[0326] Determining, based on chat messages between the local user and each contact, a communication depth between the local user and each contact within a reference time period, wherein the communication depth is used to characterize the communication density and information richness between the local user and each contact in the conversation;

[0327] Determining, based on chat messages between the local user and each contact, a commitment word frequency of the local user and each contact within a reference time period, where the commitment word frequency represents the frequency of words with a commitment tone appearing in conversations between the user and the contact;

[0328] The commitment sensitivity of the local user and each contact is determined according to the commitment word frequency and the communication depth.

[0329] In another optional example, the determining module is specifically configured to:

[0330] Determining, based on the chat messages, the average message length, average message frequency, and timely message reply rate of the local user and each contact within the reference time period;

[0331] The communication depth between the local user and each contact within a reference time period is determined based on the average message length, the average message frequency, and the timely message reply rate.

[0332] In another optional example, the determining module is specifically configured to:

[0333] Determining a commitment message in at least one chat message between the local user and each contact; the commitment message is a chat message including an indicator variable of a commitment sentence;

[0334] Determining the commitment strength of each commitment message based on the commitment keywords in each commitment message, wherein the commitment strength is used to represent the degree of expression of the commitment tone in a commitment message;

[0335] The commitment word frequency of the local user and each contact is determined according to the commitment strength of each commitment message, the generation time of each commitment message and the reference generation time, wherein the reference generation time is the system time for determining the commitment word frequency.

[0336] In another optional example, the determining module is specifically configured to:

[0337] Determining commitment credibility scores of the local user and each contact based on the commitment word frequency and the communication depth;

[0338] The commitment credibility score is mapped to a reference value interval to obtain the commitment sensitivity of the local user and each contact.

[0339] In another optional example, the determining module is specifically configured to:

[0340] determining a base score for each to-do item based on the relationship strength, the commitment sensitivity, and the deadline for completion of each to-do item;

[0341] Determine the synergy between every two to-do items based on the synergy factor between the N to-do items and the basic score of each to-do item;

[0342] Determine a total score for each to-do item based on the base score and the amount of collaboration for each to-do item;

[0343] The execution time of the N to-do items is sorted in descending order of the total scores to obtain the execution order of the N to-do items.

[0344] In another optional example, the apparatus further includes:

[0345] a decomposition module, configured to input a to-do item into a task decomposition model; if the complexity of the to-do item exceeds a preset threshold, the task decomposition model decomposes the to-do item into at least two subtasks and outputs an estimated completion time for each subtask;

[0346] The display module is used to display the second prompt information corresponding to each subtask at the task reminder time of each subtask, and the second prompt information is used to prompt the user to process the subtask; the task reminder time is the second time before the expected completion time of each subtask.

[0347] In another optional example, the display module is further configured to:

[0348] At the task reminder moment of each subtask, a first control and a second control are displayed, wherein the first control is used to indicate that the subtask corresponding to the second prompt information has been completed, and the second control is used to indicate that the subtask corresponding to the second prompt information has not been completed; the at least two subtasks include the first subtask and the second subtask;

[0349] When a selection input is received on the first control at or before the estimated completion time of the first subtask, in response to the selection input, hiding the second prompt information corresponding to the first subtask;

[0350] When a selection input to the second control is received after the estimated completion time of the second subtask, the estimated completion time of the second subtask and subtasks whose estimated completion time is later than the estimated completion time of the second subtask are updated in response to the selection input.

[0351] The to-do list reminder device in the embodiment of the present application can be an electronic device or a component in an electronic device, such as an integrated circuit or chip. The electronic device can be a terminal or other device other than a terminal. For example, the electronic device can be a mobile phone, a tablet computer, a laptop computer, a PDA, an in-vehicle electronic device, a mobile Internet device (MID), an augmented reality (AR) / virtual reality (VR) device, a robot, a wearable device, an ultra-mobile personal computer (UMPC), a netbook or a personal digital assistant (PDA), etc. It can also be a server, a network attached storage (NAS), a personal computer (PC), a television (TV), an ATM or a self-service machine, etc. It can also be a server, a network attached storage (NAS), a personal computer (PC), a television (TV), an ATM or a self-service machine, etc., and the embodiment of the present application does not specifically limit it.

[0352] The to-do list reminder device in the embodiment of the present application can be a device having an operating system. The operating system can be an Android operating system, an iOS operating system, or other possible operating systems, which are not specifically limited in the embodiment of the present application.

[0353] The to-do list reminder device provided in the embodiment of the present application can achieve Figure 1 To avoid repetition, the various processes implemented in the method embodiment are not described here.

[0354] Alternatively, as Figure 5 As shown, an embodiment of the present application also provides an electronic device 100, including a processor 110, a memory 119, and a program or instruction stored in the memory 119 and executable on the processor 110. When the program or instruction is executed by the processor 110, each process of the above-mentioned to-do reminder method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0355] It should be noted that the electronic devices in the embodiments of the present application include the above-mentioned mobile electronic devices and non-mobile electronic devices.

[0356] Please refer to Figure 6 , Figure 6 The following is a schematic diagram of the hardware structure of an electronic device implementing an embodiment of the present application. The electronic device 100 includes, but is not limited to, components such as a radio frequency unit 121, a network module 122, an audio output unit 123, an input unit 124, a sensor 125, a display unit 126, a user input unit 127, an interface unit 128, a memory 129, and a processor 120.

[0357] Those skilled in the art will understand that the electronic device 100 may also include a power source (such as a battery) to power each component, and the power source may be logically connected to the processor 120 through a power management system, thereby implementing functions such as charging, discharging, and power consumption management through the power management system. Figure 6 The electronic device structure shown in the figure does not constitute a limitation on the electronic device. The electronic device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently, which will not be repeated here.

[0358] The display unit 126 is configured to input a chat message between the local user and the contact into a language macro model, and display a first prompt message when the language macro model identifies a to-do item in the chat message; the first prompt message includes: a description of the identified to-do item and a deadline for completion of the item;

[0359] The processor 120 is configured to create a to-do reminder, wherein the reminder content of the to-do reminder is the to-do item.

[0360] In this application, the to-do items in the chat message can be automatically identified through the language model, and the prompt information of each to-do item can be displayed, and then a to-do reminder for the to-do item can be created. Compared with the manual identification of to-do items and manual creation of to-do reminders in the related art, the process of item identification and reminder creation is greatly simplified. In this way, the automatic identification of to-do items and generation of to-do reminders through the language model reduces the user's operation steps and improves the efficiency of to-do reminders.

[0361] In another optional example, the electronic device further includes:

[0362] Display unit 126, used to display a cancel control;

[0363] The processor 120 is configured to create a to-do reminder when the language macro model identifies a to-do item in the chat message;

[0364] The processor 120 is configured to delete the to-do reminder if a selection input of the cancel control is received within a first duration of displaying the cancel control.

[0365] In another optional example, the processor 120 is further configured to:

[0366] In a case where a confirmation input for the to-do reminder creation confirmation information is received, a to-do reminder is created in response to the confirmation input.

[0367] In another optional example, the electronic device further includes:

[0368] The user input unit 127 is used to input chat messages between the local user and at least one contact into the language model;

[0369] Processor 120, configured to obtain location information and activity information of a local user and at least one contact;

[0370] The processor 120 is configured to determine an execution order of the N to-do items based on the location information, the activity information, and a chat message between the local user and the at least one contact.

[0371] In another optional example, the processor 120 is specifically configured to:

[0372] Determining, based on the location information, the activity information, and chat messages between the local user and each contact, the relationship strength and commitment sensitivity between the local user and each contact, wherein the relationship strength is used to represent the closeness of the social connection between the local user and each contact, and the commitment sensitivity is used to represent the priority of the task committed by the local user to the contact;

[0373] An order of executing the N to-do items is determined according to the relationship strength, commitment sensitivity, the degree of association among the N to-do items, and the deadline for completing each to-do item.

[0374] In another optional example, the processor 120 is further configured to:

[0375] Determining the interaction intensity between the local user and each contact during the reference time period based on chat messages between the local user and each contact, activity information, and location information of the local user and each contact during the reference time period, wherein the interaction intensity is used to represent the activity of the social connection between the local user and the contact during the reference time period;

[0376] Determining, based on chat messages between the local user and each contact, location information, activity information, and interaction intensity of the local user and each contact during a reference time period, a comprehensive vector of the local user and each contact during the reference time period, the comprehensive vector being used to characterize social relationship characteristics between the local user and each contact from at least two perspectives;

[0377] The relationship strength between the local user and each contact is determined according to the comprehensive vector.

[0378] In another optional example, the processor 120 is specifically configured to:

[0379] determining, based on the chat messages, a chat rate between the local user and each contact within the reference time period;

[0380] Determining, based on the location information, a location overlap rate between the local user and each contact within the reference time period;

[0381] determining, based on the activity information, a common activity rate between the local user and each contact within the reference time period;

[0382] The interaction intensity between the local user and each contact within a reference time period is determined according to the location overlap rate, the chat rate, and the common activity rate.

[0383] In another optional example, the processor 120 is further configured to:

[0384] Construct a multi-dimensional relationship network model based on the interaction intensity between the local user and each contact within the reference time period;

[0385] The local user corresponds to the central node in the multi-dimensional relationship network model, each contact corresponds to a contact node in the multi-dimensional relationship network model, and each interaction strength corresponds to a connection line between the central node and a contact node.

[0386] In another optional example, the processor 120 is specifically configured to:

[0387] determining, based on chat messages between the local user and each contact within the reference time period, a semantic feature vector between the central node and a contact node corresponding to each contact;

[0388] Determining a time series vector between the central node and the contact node corresponding to the contact according to the number of chats between the local user and each contact in the reference time period;

[0389] Determining a spatial vector between the central node and the contact node corresponding to each contact based on the number of times the local user and each contact overlap within the reference time period and the density of common activities; the common activity density is used to represent the frequency with which two users participate in the same activity together;

[0390] Performing vector splicing on the semantic feature vector, the time series vector, and the spatial vector to obtain a spliced vector;

[0391] The concatenated vector is multiplied by the product of the interaction strength to obtain a comprehensive vector of the local user and each contact in the reference time period.

[0392] In another optional example, the processor 120 is specifically configured to:

[0393] Performing vector encoding on each of the at least one chat message to obtain an encoding vector corresponding to each chat message;

[0394] Use the language model to perform sentiment inference on each chat message and obtain the sentiment type of each chat message;

[0395] Determine the emotion vector of each chat message according to the emotion type of each chat message;

[0396] Determine a word segmentation vector for each chat message based on a commitment keyword in each chat message, where the commitment keyword is a word segmentation that expresses a commitment tone;

[0397] Determine a text feature vector for each chat message based on the encoding vector, the word segmentation vector, and the sentiment vector of each chat message;

[0398] The text feature vector corresponding to the at least one chat message is subjected to vector fusion to obtain a semantic feature vector between the central node and the contact node corresponding to each contact.

[0399] In another optional example, the processor 120 is specifically configured to:

[0400] Determine the number of chats between the local user and each contact in each sub-time period;

[0401] Determining a time sparsity factor for each sub-time period based on the length of the interval between each sub-time period and a reference sub-time period, wherein the reference sub-time period is the time period in which the system time displayed on the electronic device is located;

[0402] Based on the number of chats and the time sparsity factor, a time series vector between the central node and the contact node corresponding to the contact is determined.

[0403] In another optional example, the processor 120 is specifically configured to:

[0404] Determining the number of times the positions of the local user and each contact overlap in each sub-area within the reference time period based on the number of times the local user and each contact appear in the same sub-area within the same time period;

[0405] determining, according to the number of times the local user and each contact jointly participate in an activity, a density of common activities between the local user and each contact in each sub-area within the reference time period;

[0406] A space vector between the central node and the contact node corresponding to each contact is determined according to the number of position overlaps and the common activity density of each sub-region.

[0407] In another optional example, the processor 120 is specifically configured to:

[0408] determining a characteristic component of each vector dimension in the comprehensive vector;

[0409] Determining a feature interaction enhancement value of the comprehensive vector according to a collaborative weight coefficient of each feature component;

[0410] The feature interaction enhancement value is mapped to a reference value interval to obtain the relationship strength between the local user and each contact.

[0411] In another optional example, the processor 120 is specifically configured to:

[0412] Determining, based on chat messages between the local user and each contact, a communication depth between the local user and each contact within a reference time period, wherein the communication depth is used to characterize the communication density and information richness between the local user and each contact in the conversation;

[0413] Determining, based on chat messages between the local user and each contact, a commitment word frequency of the local user and each contact within a reference time period, where the commitment word frequency represents the frequency of words with a commitment tone appearing in conversations between the user and the contact;

[0414] The commitment sensitivity of the local user and each contact is determined according to the commitment word frequency and the communication depth.

[0415] In another optional example, the processor 120 is specifically configured to:

[0416] Determining, based on the chat messages, the average message length, average message frequency, and timely message reply rate of the local user and each contact within the reference time period;

[0417] The communication depth between the local user and each contact within a reference time period is determined based on the average message length, the average message frequency, and the timely message reply rate.

[0418] In another optional example, the processor 120 is specifically configured to:

[0419] Determining a commitment message in at least one chat message between the local user and each contact; the commitment message is a chat message including an indicator variable of a commitment sentence;

[0420] Determining the commitment strength of each commitment message based on the commitment keywords in each commitment message, wherein the commitment strength is used to represent the degree of expression of the commitment tone in a commitment message;

[0421] The commitment word frequency of the local user and each contact is determined according to the commitment strength of each commitment message, the generation time of each commitment message and the reference generation time, wherein the reference generation time is the system time for determining the commitment word frequency.

[0422] In another optional example, the processor 120 is specifically configured to:

[0423] Determining commitment credibility scores of the local user and each contact based on the commitment word frequency and the communication depth;

[0424] The commitment credibility score is mapped to a reference value interval to obtain the commitment sensitivity of the local user and each contact.

[0425] In another optional example, the processor 120 is specifically configured to:

[0426] determining a base score for each to-do item based on the relationship strength, the commitment sensitivity, and the deadline for completion of each to-do item;

[0427] Determine the synergy between every two to-do items based on the synergy factor between the N to-do items and the basic score of each to-do item;

[0428] Determine a total score for each to-do item based on the base score and the amount of collaboration for each to-do item;

[0429] The execution time of the N to-do items is sorted in descending order of the total scores to obtain the execution order of the N to-do items.

[0430] In another optional example, the electronic device further includes:

[0431] The processor 120 is configured to input a to-do item into a task splitting model. If the complexity of the to-do item exceeds a preset threshold, the task splitting model splits the to-do item into at least two subtasks and outputs an estimated completion time for each subtask.

[0432] The display unit 126 is used to display the second prompt information corresponding to each subtask at the task reminder time of each subtask, and the second prompt information is used to prompt the user to process the subtask; the task reminder time is the second time before the expected completion time of each subtask.

[0433] In another optional example, the display unit 126 is further configured to:

[0434] At the task reminder moment of each subtask, a first control and a second control are displayed, wherein the first control is used to indicate that the subtask corresponding to the second prompt information has been completed, and the second control is used to indicate that the subtask corresponding to the second prompt information has not been completed; the at least two subtasks include the first subtask and the second subtask;

[0435] When a selection input is received on the first control at or before the estimated completion time of the first subtask, in response to the selection input, hiding the second prompt information corresponding to the first subtask;

[0436] When a selection input to the second control is received after the estimated completion time of the second subtask, the estimated completion time of the second subtask and subtasks whose estimated completion time is later than the estimated completion time of the second subtask are updated in response to the selection input.

[0437] It should be understood that in an embodiment of the present application, the input unit 124 may include a graphics processing unit (GPU) 1241 and a microphone 1242, and the graphics processor 1241 processes the image data of a static picture or video obtained by an image capture device (such as a camera) in a video capture mode or an image capture mode. The display unit 126 may include a display panel 1261, and the display panel 1261 may be configured in the form of a liquid crystal display, an organic light emitting diode, etc. The user input unit 127 includes a touch panel 1271 and at least one of other input devices 1272. The touch panel 1271 is also called a touch screen. The touch panel 1271 may include two parts: a touch detection device and a touch controller. Other input devices 1272 may include, but are not limited to, a physical keyboard, function keys (such as volume control keys, switch keys, etc.), a trackball, a mouse, and a joystick, which will not be repeated here.

[0438] The memory 129 can be used to store software programs and various data. The memory 129 may mainly include a first storage area for storing programs or instructions and a second storage area for storing data, wherein the first storage area may store an operating system, applications or instructions required for at least one function (such as a sound playback function, an image playback function, etc.). In addition, the memory 129 may include a volatile memory or a non-volatile memory, or the memory 129 may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), a static random access memory (SRAM), a dynamic random access memory (DRAM), a synchronous dynamic random access memory (SDRAM), a double data rate synchronous dynamic random access memory (DDRSDRAM), an enhanced synchronous dynamic random access memory (ESDRAM), a synchronous link dynamic random access memory (SLDRAM), and a direct memory bus random access memory (DRRAM). The memory 129 in the embodiment of the present application includes but is not limited to these and any other suitable types of memory.

[0439] Processor 120 may include one or more processing units. Optionally, processor 120 integrates an application processor and a modem processor. The application processor primarily handles operations related to the operating system, user interface, and application programs, while the modem processor primarily processes wireless communication signals, such as a baseband processor. It is understood that the modem processor may not be integrated into processor 120.

[0440] An embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the various processes of the above-mentioned to-do reminder method embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0441] The processor is the processor in the electronic device in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk.

[0442] An embodiment of the present application further provides a chip, which includes a processor and a communication interface. The communication interface and the processor are coupled, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned to-do reminder method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0443] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0444] An embodiment of the present application provides a computer program product, which is stored in a storage medium. The program product is executed by at least one processor to implement the various processes of the above-mentioned to-do reminder method embodiment and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0445] It should be noted that, in this article, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element. In addition, it should be noted that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in the opposite order according to the functions involved. For example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0446] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), including a number of instructions for enabling a terminal (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0447] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this application, ordinary technicians in this field can also make many forms without departing from the purpose of this application and the scope of protection of the claims, all of which are within the protection of this application.

Claims

1. A to-do reminder method, characterized in that: include: Inputting a chat message between the local user and the contact into a language macro model, and displaying a first prompt message when the language macro model recognizes a to-do item in the chat message; The first prompt information includes: the item description information of the identified to-do item and the deadline for completion of the item; Create a to-do reminder, where the reminder content of the to-do reminder is the to-do item.

2. The method according to claim 1, characterized in that The first prompt information further includes: to-do reminder creation prompt information, wherein the to-do reminder creation prompt information is used to prompt that a to-do reminder with the reminder content being the to-do item has been created; When the language model identifies a to-do item in the chat message, the method further includes: Show cancel control; Creating a to-do reminder includes: When the language model identifies a to-do item in the chat message, creating a to-do reminder; After creating the to-do reminder, it also includes: In a case where a selection input of the cancel control is received within a first duration of displaying the cancel control, the to-do reminder is deleted.

3. The method according to claim 1, characterized in that The first prompt information further includes: to-do reminder creation confirmation information: the to-do reminder creation confirmation information is used to prompt the user whether to create a to-do reminder based on the to-do item; The step of creating a to-do reminder includes: In a case where a confirmation input for the to-do reminder creation confirmation information is received, a to-do reminder is created in response to the confirmation input.

4. The method according to claim 1, wherein The step of inputting the chat messages between the local user and the contact into the language model includes: Inputting chat messages between the local user and at least one contact into the language model; When the language macro model identifies N to-do items in the chat message and N is an integer greater than 1, the method further includes: Obtain location information and activity information of the local user and at least one contact; An order of executing the N to-do items is determined according to the location information, the activity information, and chat messages between the local user and the at least one contact.

5. The method according to claim 4, characterized in that The determining, based on the location information, the activity information, and the chat messages between the local user and the at least one contact, an order of executing the N to-do items includes: Determining, based on the location information, the activity information, and chat messages between the local user and each contact, the relationship strength and commitment sensitivity between the local user and each contact, wherein the relationship strength is used to represent the closeness of the social connection between the local user and each contact, and the commitment sensitivity is used to represent the priority of the task committed by the local user to the contact; An order of executing the N to-do items is determined according to the relationship strength, commitment sensitivity, the degree of association among the N to-do items, and the deadline for completing each to-do item.

6. The method according to claim 5, characterized in that The determining, based on the location information, the activity information, and the chat messages between the local user and each contact, the relationship strength between the local user and each contact includes: Determining the interaction intensity between the local user and each contact during the reference time period based on chat messages between the local user and each contact, activity information, and location information of the local user and each contact during the reference time period, wherein the interaction intensity is used to represent the activity of the social connection between the local user and the contact during the reference time period; Determining, based on chat messages between the local user and each contact, location information, activity information, and interaction intensity of the local user and each contact during a reference time period, a comprehensive vector of the local user and each contact during the reference time period, the comprehensive vector being used to characterize social relationship characteristics between the local user and each contact from at least two perspectives; The relationship strength between the local user and each contact is determined according to the comprehensive vector.

7. The method according to claim 6, characterized in that Determining the interaction intensity between the local user and each contact within the reference time period based on chat messages between the local user and each contact, activity information, and location information of the local user and each contact within the reference time period includes: determining, based on the chat messages, a chat rate between the local user and each contact within the reference time period; Determining, based on the location information, a location overlap rate between the local user and each contact within the reference time period; determining, based on the activity information, a common activity rate between the local user and each contact within the reference time period; The interaction intensity between the local user and each contact within a reference time period is determined according to the location overlap rate, the chat rate, and the common activity rate.

8. The method according to claim 6, characterized in that After determining the interaction intensity between the local user and each contact within the reference time period based on the chat messages between the local user and each contact, and the activity information and location information of the local user and each contact within the reference time period, the method further includes: Construct a multi-dimensional relationship network model based on the interaction intensity between the local user and each contact within the reference time period; The local user corresponds to the central node in the multi-dimensional relationship network model, each contact corresponds to a contact node in the multi-dimensional relationship network model, and each interaction strength corresponds to a connection line between the central node and a contact node.

9. The method according to claim 8, characterized in that The determining, based on chat messages between the local user and each contact, location information, activity information, and interaction intensity of the local user and each contact within the reference time period, a comprehensive vector of the local user and each contact within the reference time period includes: determining, based on chat messages between the local user and each contact within the reference time period, a semantic feature vector between the central node and a contact node corresponding to each contact; Determining a time series vector between the central node and the contact node corresponding to the contact according to the number of chats between the local user and each contact in the reference time period; Determining a spatial vector between the central node and the contact node corresponding to each contact based on the number of times the local user and each contact overlap within the reference time period and the density of common activities; the common activity density is used to represent the frequency with which two users participate in the same activity together; Performing vector splicing on the semantic feature vector, the time series vector, and the spatial vector to obtain a spliced vector; The concatenated vector is multiplied by the product of the interaction strength to obtain a comprehensive vector of the local user and each contact in the reference time period.

10. The method according to claim 9, characterized in that The determining, based on the chat messages between the local user and each contact within the reference time period, a semantic feature vector between the central node and the contact node corresponding to each contact, includes: Performing vector encoding on each of the at least one chat message to obtain an encoding vector corresponding to each chat message; Use the language model to perform sentiment inference on each chat message and obtain the sentiment type of each chat message; Determine the emotion vector of each chat message according to the emotion type of each chat message; Determine a word segmentation vector for each chat message based on a commitment keyword in each chat message, where the commitment keyword is a word segmentation that expresses the tone of commitment; Determine a text feature vector for each chat message based on the encoding vector, the word segmentation vector, and the sentiment vector of each chat message; The text feature vector corresponding to the at least one chat message is subjected to vector fusion to obtain a semantic feature vector between the central node and the contact node corresponding to each contact.

11. The method according to claim 9, characterized in that The reference time period is divided into K sub-time periods of equal length; wherein K is an integer greater than 1; The determining, based on the number of chats between the local user and each contact in the reference time period, a time series vector between the central node and the contact node corresponding to the contact, includes: Determine the number of chats between the local user and each contact in each sub-time period; Determining a time sparsity factor for each sub-time period based on the length of the interval between each sub-time period and a reference sub-time period, wherein the reference sub-time period is the time period in which the system time displayed on the electronic device is located; Based on the number of chats and the time sparsity factor, a time series vector between the central node and the contact node corresponding to the contact is determined.

12. The method according to claim 9, characterized in that The user's activity area includes P sub-areas; The determining, based on the number of position overlaps and the common activity density between the local user and each contact within the reference time period, of a space vector between the central node and the contact node corresponding to each contact includes: Determining the number of times the positions of the local user and each contact overlap in each sub-area within the reference time period based on the number of times the local user and each contact appear in the same sub-area within the same time period; determining, according to the number of times the local user and each contact jointly participate in an activity, a density of common activities between the local user and each contact in each sub-area within the reference time period; A space vector between the central node and the contact node corresponding to each contact is determined according to the number of position overlaps and the common activity density of each sub-region.

13. The method according to claim 6, characterized in that Determining the relationship strength between the local user and each contact according to the comprehensive vector includes: determining a characteristic component of each vector dimension in the comprehensive vector; Determining a feature interaction enhancement value of the comprehensive vector according to a collaborative weight coefficient of each feature component; The feature interaction enhancement value is mapped to a reference value interval to obtain the relationship strength between the local user and each contact.

14. The method according to claim 5, characterized in that The determining, based on the location information, the activity information, and the chat messages between the local user and each contact, a commitment sensitivity between the local user and each contact, includes: Determining, based on chat messages between the local user and each contact, a communication depth between the local user and each contact within a reference time period, wherein the communication depth is used to characterize the communication density and information richness in the conversation between the local user and each contact; Determining, based on chat messages between the local user and each contact, a commitment word frequency of the local user and each contact within a reference time period, where the commitment word frequency represents the frequency of words with a commitment tone appearing in conversations between the user and the contact; The commitment sensitivity of the local user and each contact is determined according to the commitment word frequency and the communication depth.

15. The method according to claim 14, characterized in that The determining, based on the chat messages between the local user and each contact, the communication depth between the local user and each contact within a reference time period includes: Determining, based on the chat messages, the average message length, average message frequency, and timely message reply rate of the local user and each contact within the reference time period; The communication depth between the local user and each contact within a reference time period is determined based on the average message length, the average message frequency, and the timely message reply rate.

16. The method according to claim 14, characterized in that The determining, based on the chat messages between the local user and each contact, the commitment word frequency of the local user and each contact within a reference time period includes: Determining a commitment message in at least one chat message between the local user and each contact; the commitment message is a chat message including an indicator variable of a commitment sentence; Determining the commitment strength of each commitment message based on the commitment keywords in each commitment message, wherein the commitment strength is used to represent the degree of expression of the commitment tone in a commitment message; The commitment word frequency of the local user and each contact is determined according to the commitment strength of each commitment message, the generation time of each commitment message and the reference generation time, wherein the reference generation time is the system time for determining the commitment word frequency.

17. The method according to claim 14, characterized in that The determining, based on the commitment word frequency and the communication depth, the commitment sensitivity of the local user and each contact includes: Determining commitment credibility scores of the local user and each contact based on the commitment word frequency and the communication depth; The commitment credibility score is mapped to a reference value interval to obtain the commitment sensitivity of the local user and each contact.

18. The method according to claim 5, characterized in that The determining the execution order of the N to-do items according to the relationship strength, commitment sensitivity, the relevance of the N to-do items, and the deadline for completion of each to-do item includes: determining a base score for each to-do item based on the relationship strength, the commitment sensitivity, and the deadline for completion of each to-do item; Determine the synergy between every two to-do items based on the synergy factor between the N to-do items and the basic score of each to-do item; Determine a total score for each to-do item based on the base score and the amount of collaboration for each to-do item; The execution time of the N to-do items is sorted in descending order of the total scores to obtain the execution order of the N to-do items.

19. The method according to any one of claims 1 to 18, characterized in that After creating a to-do reminder, wherein the reminder content of the to-do reminder is the to-do item, the method further includes: Inputting a to-do item into a task splitting model, and if the complexity of the to-do item is greater than a preset threshold, the task splitting model splits the to-do item into at least two subtasks and outputs an estimated completion time for each subtask; At the task reminder moment of each subtask, the second prompt information corresponding to each subtask is displayed, and the second prompt information is used to prompt the user to process the subtask; the task reminder moment is the second time before the expected completion time of each subtask.

20. The method according to claim 19, characterized in that The method further comprises: At the task reminder moment of each subtask, a first control and a second control are displayed, wherein the first control is used to indicate that the subtask corresponding to the second prompt information has been completed, and the second control is used to indicate that the subtask corresponding to the second prompt information has not been completed; the at least two subtasks include the first subtask and the second subtask; After displaying the second prompt information corresponding to each subtask, the method further includes: When a selection input is received on the first control at or before the estimated completion time of the first subtask, in response to the selection input, hiding the second prompt information corresponding to the first subtask; When a selection input to the second control is received after the estimated completion time of the second subtask, the estimated completion time of the second subtask and subtasks whose estimated completion time is later than the estimated completion time of the second subtask are updated in response to the selection input.

21. A to-do list reminder device, characterized in that: include: A display module is configured to input a chat message between the local user and the contact into a language macro model, and display a first prompt message when the language macro model recognizes a to-do item in the chat message; The first prompt information includes: the item description information of the identified to-do item and the deadline for completion of the item; The creation module is used to create a to-do reminder, wherein the reminder content of the to-do reminder is the to-do item.

22. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the to-do list reminder method according to any one of claims 1 to 20 are implemented.

Images