Time information processing method and system based on large language model

Abstract

The invention discloses a time information processing method and system based on a large language model, and relates to the field of large language models.The method comprises the steps that dialogue information input by a user on a dialogue interface of a client and the large language model is sent to a server so that the server can process the dialogue information, and meanwhile the dialogue information is forwarded to a large language model end; the dialogue information comprises reminding information, and the reminding information comprises reminding time and reminding content; reminding information sent by the server is received, the reminding information comprises timed reminding information, and the timed reminding information is confirmed and labeled reminding information obtained after the server analyzes the dialogue information and / or response information of the large language model side for the dialogue information; the timed reminding information comprises timed reminding time and a timed reminding event; and if the current time reaches the timed reminding time, triggering reminding logic, and executing a reminding task according to the timed reminding information. According to the invention, the large language model can provide a time reminding service.

Description

Technical Field

[0001] The present application relates to the field of large language models, and in particular to a time information processing method and system based on a large language model. Background Art

[0002] A large language model is a language model built by a deep neural network with more than tens of billions of parameters. It has strong language comprehension capabilities, efficient text generation capabilities, and extensive knowledge reserves and information integration capabilities, so it has been widely used in all walks of life.

[0003] However, the current large language model can only answer questions raised by users or communicate with users, and it cannot provide time-related services including time reminder services. Summary of the invention

[0004] The purpose of this application is to provide a time information processing method and system based on a large language model, which can enable the large language model to provide a time reminder service.

[0005] To achieve the above objectives, this application provides the following solutions:

[0006] In a first aspect, the present application provides a method for processing time information based on a large language model, the method being applied to a client, the method comprising:

[0007] Sending the dialogue information input by the user in the dialogue interface between the client and the large language model to the server, so that the server processes the dialogue information and forwards it to the large language model end; the dialogue information includes at least one of the following information: text, voice, picture, video or document, and the dialogue information includes reminder information, and the reminder information includes reminder time and reminder content;

[0008] Receiving reminder information sent by the server, the reminder information including timed reminder information, the timed reminder information being reminder information confirmed and marked after the server parses the conversation information and / or the response information of the large language model end to the conversation information; the timed reminder information includes: timed reminder time and timed reminder event;

[0009] If the current time has reached the timer reminder time, the reminder logic is triggered and the reminder task is executed according to the timer reminder information.

[0010] In a second aspect, the present application provides a time information processing method based on a large language model, the method is applied to a server, and the method includes:

[0011] Receiving a conversation message sent by a client; the conversation message includes at least one of the following information: text, voice, picture, video and document, and the conversation message includes reminder information, and the reminder information includes reminder time and reminder content;

[0012] Parsing the dialogue information and sending the dialogue information to the large language model end;

[0013] receiving response information to the dialogue information sent by the large language model end;

[0014] Parsing the response information to extract the reminder information contained in the dialogue information and the response information by parsing the dialogue information and / or the response information;

[0015] Marking the obtained reminder information as timed reminder information and storing it in a database;

[0016] A timer is set according to the timed reminder time of the timed reminder information and starts timing; the timed reminder time is the reminder time corresponding to the timed reminder information;

[0017] It is detected whether the current time reaches the timer reminder time. If it reaches the timer reminder time, the reminder logic is triggered, and the timer reminder information including the timer reminder time has been reached is sent to the client, and the client executes the reminder task.

[0018] In a third aspect, the present application provides a time information processing system based on a large language model, the time information processing system based on a large language model comprising: a client, a server and a large language model end;

[0019] The server is connected to the client through an API interface and is connected to the large language model end;

[0020] The client includes: a user interface module, an input and output module, a data storage module, and a notification and reminder module;

[0021] The server includes: a middleware module, a natural language processing module, a logic processing module and a database;

[0022] The user interface module is used to provide a dialogue interface of the large language model to the user;

[0023] The input-output module is used to send the dialogue information input by the user in the dialogue interface of the large language model to the middleware module, so that the middleware module processes the dialogue information and forwards it to the large language model end; the dialogue information includes at least one of the following information: text, voice, picture, video or document, and the dialogue information includes reminder information, and the reminder information includes reminder time and reminder content;

[0024] The middleware module is used to coordinate the communication between the client and the large language model API, process client requests, receive and send dialogue information sent by the client input and output module; the dialogue information includes at least one of the following information: text, voice, picture, video and document, and the dialogue information includes reminder information, which includes reminder time and reminder content;

[0025] The natural language processing module is used to parse the dialogue information and send the dialogue information to the large language model end;

[0026] The middleware module is further used to receive response information to the dialogue information sent by the large language model end;

[0027] The natural language processing module is used to parse the response information of the large language model, so as to extract the reminder information contained in the dialogue information and the response information by parsing the dialogue information and / or the response information;

[0028] The logic processing module is used to mark the obtained reminder information as timed reminder information and store it in the database;

[0029] The logic processing module is further used to set a timer and start timing according to the timed reminder time of the timed reminder information; the timed reminder time is the reminder time corresponding to the timed reminder information;

[0030] The logic processing module is further used to detect whether the current time reaches the timer reminder time, and if so, trigger the reminder logic, and the middleware module is further used to send the timer reminder information including the timer reminder time reaching to the data storage module, and the notification and reminder module executes the reminder task;

[0031] The data storage module is used to receive the reminder information sent by the server, and the reminder information includes timed reminder information;

[0032] The notification and reminder module is used to trigger the reminder logic when detecting that the current time has reached the scheduled reminder time, and execute the reminder task according to the scheduled reminder information.

[0033] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0034] The present application provides a time information processing method and system based on a large language model, wherein the conversation information input by a user in a conversation interface between a client and a large language model is sent to a server, so that the server processes the conversation information and forwards it to the large language model end at the same time; the conversation information includes at least one of the following information: text, voice, picture, video or document, and the conversation information includes reminder information, and the reminder information includes reminder time and reminder content; the reminder information sent by the server is received, and the reminder information includes timed reminder information, and the timed reminder information is the reminder information confirmed and marked after the server parses the conversation information and / or the response information of the large language model end to the conversation information; the timed reminder information includes: timed reminder time and timed reminder event; if the current time has reached the timed reminder time, the reminder logic is triggered, and the reminder task is executed according to the timed reminder information. The method disclosed in the present invention enables a large language model to provide a time reminder service, and by setting a server to confirm the time of the reminder information, when the server detects that the reminder time corresponding to the reminder information has arrived, it will send the reminder information to the client. In this way, the client does not need to frequently execute a large number of time access responses. After receiving the reminder information sent by the server, the reminder logic is directly executed, thereby not distracting the task of processing the dialogue chat information, and not causing waste of computing power and reduced access speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0036] Figure 1 A flow chart of a method for processing time information based on a large language model provided in an embodiment of the present application Figure 1 ;

[0037] Figure 2 A flow chart of a method for processing time information based on a large language model provided in an embodiment of the present application Figure 2 ;

[0038] Figure 3 A schematic diagram of a time information processing system based on a large language model provided in one embodiment of the present application;

[0039] Figure 4 is a block diagram of a device for processing time information based on a large language model according to an exemplary embodiment;

[0040] Figure 5 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only 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 without creative work are within the scope of protection of this application.

[0042] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0043] Figure 1 is a flowchart of a method for processing time information based on a large language model according to an exemplary embodiment. Figure 1 As shown, the method is used in a client and includes the following steps S101-S103:

[0044] In step S101, the dialogue information input by the user in the dialogue interface between the client and the large language model is sent to the server, so that the server processes the dialogue information and forwards it to the large language model end; the dialogue information includes at least one of the following information: text, voice, picture, video or document, and the dialogue information includes reminder information, and the reminder information includes reminder time and reminder content.

[0045] When the user needs the large language model to remind him, he will input the dialogue information containing the reminder intention in the dialogue interface of the large language model. At this time, the dialogue information input can be in any form.

[0046] When the input dialog information is text, the client directly sends the text to the server.

[0047] For example: the user asks the large language model to tell a story and stops after half an hour. At this time, the user enters the following text in the dialogue interface of the large language model: Xiaoyu (the name or nickname of the large language model), please tell me the Romance of the Three Kingdoms for half an hour, continuing from the last time. I want to fall asleep while listening.

[0048] When the input dialogue information is voice, the client can parse the voice into text through the built-in voice recognition software or module, and then send the text to the server, or directly send the voice to the server so that the server can parse the voice and obtain the corresponding text. The voice can also be sent to the large language model through the server, and the voice recognition module of the large language model will perform voice recognition. In particular, when the voice recognition module of the large language model can recognize the user's current speaking voice and thus judge the user's emotions by recognizing the user's voice, it can provide the user with more accurate response information.

[0049] For example: A user asks the big language model to remind him that he has a meeting at 10 o'clock and to be reminded to attend the meeting at 9:50. The user enters the following voice in the dialogue interface of the big language model: Xiaoyu, I have a meeting at 10 o'clock, please remind me to attend it at 9:50.

[0050] When the input dialogue information is a picture, the picture can be sent directly to the server so that the server's image recognition module can parse the picture to obtain the corresponding text. The picture can also be sent to the server, which will forward it to the large language model, and the image recognition module of the large language model will perform image recognition, parse the text in the picture, and understand the content in the picture. The client can also parse the picture into text through the built-in picture parsing algorithm, and then send the text to the server.

[0051] For example, if a user has a concert to attend at 9 o'clock, he can enter the concert ticket into the dialogue interface of the large language model.

[0052] When the input dialogue information is a video, the video stream can be directly sent to the server in real time so that the image recognition module of the server can identify each image frame in the video stream and obtain the corresponding text. The video stream can also be sent to the server in real time, and the server forwards it to the large language model, and the image recognition module of the large language model performs image recognition on the video stream, identifies the text of each image frame in the video stream and the specific content of each image frame, and outputs the recognition result. For the embodiment in which the client has an image recognition software or module, each image frame of the video stream can be image recognized and parsed into text by the built-in image recognition software or module, and then the text is sent to the server.

[0053] When the input conversation information is a document, the document is sent to the server and forwarded to the large language model. The large language model parses the document, extracts the text and images in the document, performs a comprehensive analysis of the text and images, outputs a response to the server, and the server sends it to the client based on the response content.

[0054] The client will send the dialogue information input by the user in the dialogue interface of the large language model to the server. After receiving the dialogue information, the server will process the dialogue information. While processing the dialogue information, the server will also send the dialogue information to the large language model end. After processing the dialogue information, the server can extract whether the reminder information in the dialogue information has at least the semantics of the reminder time and the reminder content. After receiving the dialogue information, the large language model end will also parse the dialogue information to extract the reminder information in the dialogue information, and then generate a response information to the reminder information and send it to the server. The server will also parse the response information to extract whether the reminder information in the response information has at least the semantics of the reminder time and the reminder content. Then, the server will obtain the confirmed timed reminder information based on the reminder information obtained by parsing the dialogue information and the reminder information obtained by parsing the response information. If it has at least the semantics of the reminder time and the reminder content, the server will obtain the confirmed timed reminder information.

[0055] For example, a user (host) asks the large language model (small language) to remind him that he has a meeting at 10 o'clock and to be reminded to attend the meeting at 9:50.

[0056] The user inputs in the dialogue interface of the large language model: Xiaoyu, I have a meeting at 10 o'clock, please remind me to attend it at 9:50.

[0057] At this time, after processing the above conversation information, the server extracts the semantics of the reminder information in the conversation information as follows: reminder time: 10 o'clock, reminder content: attend the meeting.

[0058] Xiaoyu responded with the following message: OK, master, you have a meeting at 10 o'clock. I will remind you to attend the meeting at 9:50.

[0059] At this time, after processing the above response information, the server extracts the semantics of the reminder information in the response information as follows: reminder time: 10 o'clock, reminder content: attend the meeting.

[0060] Then the server will obtain the final timed reminder information according to the reminder information obtained by parsing the above-mentioned conversation information and the reminder information obtained by parsing the response information: reminder time: 10 o'clock, reminder content: attend the meeting.

[0061] In step S102, a reminder message sent by the server is received, where the reminder message includes a timed reminder message, where the timed reminder message is the reminder message that is confirmed and marked after the server parses the conversation message and / or the response message of the large language model to the conversation message; the timed reminder message includes: a timed reminder time and a timed reminder content.

[0062] In step S103, if the current time has reached the scheduled reminder time, the reminder logic is triggered and the reminder task is executed according to the scheduled reminder information.

[0063] In one embodiment, triggering reminder logic includes:

[0064] Output reminder information and timed reminder information by voice; and / or, output timed reminder information by text, and / or, output timed reminder information by local notification; and / or, output timed reminder information by message template; and / or, output timed reminder information by telephone notification; and / or, output timed reminder information by playing ringtones.

[0065] The reminder information sent by the server and received by the client includes timed reminder information and may also include reminder information for which the server has no timer set.

[0066] If the reminder information received by the client is a timed reminder information, the timed reminder information is saved in the cache, because the execution time of the reminder task of the timed reminder information is very close, and it is saved in the storage module or cache. When executing the reminder task, it can be quickly called, which improves the execution efficiency. If the reminder information received is a reminder information for which the server does not set a timer, the client can save it in the storage module or cache. The reminder information may be the response information returned by the large language model to the dialogue information according to the user input, and then it is output to the user on the client. For the embodiment of setting the time acquisition module on the client, a timer can be set for it according to the reminder information and start timing. After detecting that the current time has reached the timed reminder time, the client executes the reminder task according to the reminder information.

[0067] If the client receives multiple timed reminder messages from the server, the client can store all of the reminder messages in a storage module or cache, and then the client sets a timer for each reminder message. When it is detected that the current time has reached the reminder time corresponding to a certain reminder message, the client will remind the reminder message. For the multiple timed reminder messages stored in the client storage module or cache, it is possible that the user sent an instruction to query the reminder task list to the server, and the server sent the reminder task list contents of the multiple timed reminder messages to the client for the user to browse and view.

[0068] In the related art, since the large language model is used for dialogue and chat with the user, if the large language model directly executes the reminder task related to time in the dialogue and chat, it will bring a large number of time access responses, which will disperse the dialogue and chat information tasks that should be processed, resulting in a waste of computing power and a reduction in access speed. In the present disclosure, by setting a server to confirm the time of the reminder information, the server will send the reminder information to the client when it detects that the reminder time corresponding to the reminder information has arrived. In this way, the client does not need to frequently execute a large number of time access responses. After receiving the reminder information sent by the server, the reminder logic is directly executed, thereby not dispersing the dialogue and chat information tasks that should be processed, and not causing a waste of computing power and a reduction in access speed.

[0069] Furthermore, the network communication transmission of the conversation chat information between the large language model and the user client also takes a certain amount of time, especially when the network communication is not smooth and the transmission speed is reduced, there will be a significant delay, resulting in time delay and inaccurate time. In another embodiment, since the server sends the timing reminder information to the client some time before the timing reminder time, instead of sending it to the client when the timing reminder time arrives, even if the network communication is not smooth and the transmission speed is reduced, it will not cause a delay in the reminder time.

[0070] In one embodiment, before receiving the timing reminder information sent by the server, the method further includes the following steps A1-A3:

[0071] A1. receiving an inquiry message sent by a server, wherein the inquiry message has the semantics of whether the user has completed the reminder task according to the reminder content of the reminder message, or the inquiry message has the semantics of asking the user whether to confirm the reminder message;

[0072] A2. Display inquiry information to the user in the dialogue interface of the large language model, prompting the user to confirm.

[0073] A3. Send the reply information entered by the user to the server.

[0074] When the inquiry information has the semantics of whether the user has completed the reminder task according to the reminder content of the reminder information: after detecting that the scheduled reminder time has arrived, the server will send an inquiry about the user's execution status to the client. After receiving the inquiry information, the client will display the inquiry information to the user on the dialogue interface of the large language model, prompting the user to confirm, and then the user will reply on the dialogue interface of the large language model, and the client will send the reply information entered by the user on the dialogue interface of the large language model to the server.

[0075] When the inquiry information has the semantics of asking the user whether to confirm the reminder information: the server detects that the response information requires user confirmation, it will send the inquiry information to the client, and the client will display the inquiry information on the dialogue interface of the large language model, prompting the user to confirm, and then the user will reply on the dialogue interface of the large language model, and the client will send the reply information entered by the user on the dialogue interface of the large language model to the server.

[0076] The form in which the inquiry information is displayed to the user in the dialogue interface of the large language model includes at least one of the following forms: text, voice, picture, video and document. The form in which the user responds in the dialogue interface of the large language model includes at least one of the following forms: text, voice, picture and video.

[0077] In one embodiment, the reminder information also includes geographic location information; when it is detected that the current time has reached the scheduled reminder time and the real-time geographic location of the client is located at the position corresponding to the geographic location information in the reminder information, the reminder logic is triggered.

[0078] Specifically, when the client detects that the current time has reached the scheduled reminder time, it starts to detect the client's real-time geographic location. When the client's geographic location reaches the position corresponding to the geographic location information in the reminder information, the reminder logic is triggered.

[0079] When the client receives the scheduled reminder information sent by the server, it not only includes: the scheduled reminder time and the scheduled reminder content, but also the geographical location information corresponding to the scheduled reminder. Therefore, when the client triggers the reminder logic, it is not only necessary that the current time has reached the scheduled reminder time, but also that the client's real-time geographical location information is the geographical location corresponding to the geographical location information included in the scheduled reminder.

[0080] In one embodiment, the timing geographic location information may also be directly used as a trigger condition for the reminder logic.

[0081] For example: A user asks the large language model to remind him to pick up a package downstairs when he returns from a walk in the park.

[0082] The user inputs in the dialogue interface of the large language model: Xiaoyu, I am going to take a walk in the park now. When I come back, remind me to pick up the express downstairs.

[0083] Xiaoyu: Okay, master, you go for a walk in the park now. When you come back, I will remind you to pick up the express downstairs.

[0084] In this example, when the client detects that the real-time geographic location is near a residential area, it triggers the reminder logic to remind the user to pick up the express delivery.

[0085] In one embodiment, when the timed reminder information is confirmed, the timer is started, and the timed reminder time and reminder content or the timed reminder information summary is displayed in the dialogue interface of the large language model in the form of a floating window or a floating button or a following page display box; clicking the floating window or the floating button or the display box can open the details of the timed reminder information; or, clicking the floating window or the floating button or the display box can edit and modify the timed reminder information; or, setting a switch button in the floating window or the floating button or the display box, clicking the switch button in the floating window or the floating button or the display box can turn off or on the timed reminder information.

[0086] In one embodiment, the method further comprises the following steps B1-B4:

[0087] B1. Push a setting button for personalized reminder information on the dialogue interface of the large language model.

[0088] B2. In response to pressing the setting button, a selection menu of personalized reminder information is output, and the selection menu of personalized reminder information includes at least one of the following personalized information: image, audio, video, 2D animation, 3D animation, virtual image and digital human.

[0089] B3. Receive the target personalized information selected by the user in the selection menu.

[0090] B4. When the current time reaches the scheduled reminder time, the reminder logic is triggered, the reminder task is executed according to the scheduled reminder information, and the target personalized information is output.

[0091] In order to improve the user experience, the present disclosure can also push a setting button for personalized reminder information in the dialogue interface of the large language model, so that the personalized information selected by the user is finally output to the user when reminding. For example, the personalized information may include at least one of the following personalized information: image, audio, video, 2D animation, 3D animation, virtual image and digital human.

[0092] For example, when the target personalized information selected by the user is a digital human, when the current time has reached the scheduled reminder time, the digital human will be displayed on the dialogue interface of the large language model, and the reminder information will be broadcasted through the digital human voice.

[0093] In one embodiment, the reminder information also includes at least the reminder time and reminder name among the following information: reminder time, reminder name, reminder number, reminder content, image and document of reminder content; and / or, the response information includes at least one item of the following information: reminder time, reminder name, reminder number, reminder content, image and document of reminder content and common sense, precautions or suggestions associated with the reminder information.

[0094] For example: the user enters dialogue information in the dialogue interface of the client's large language model. The dialogue information includes reminder information such as the time to take medicine, the name of the medicine, the amount of medicine each time, or the image of the medicine packaging and instructions including pictures or videos. The process of the server executing the reminder based on the dialogue information has been explained in detail on the server side and will not be repeated here.

[0095] Furthermore, the large language model responds based on the drug name, time to take the medicine, and dosage per dose in the reminder information. The response information provides users with reference suggestions including drug knowledge, medical knowledge, precautions, dietary recommendations, etc.

[0096] The user sees the reference suggestions on the dialogue interface of the large language model on the client and replies to the large language model. Further, the two parties can continue the dialogue until they stop entering dialogue information or enter a statement to stop or pause the dialogue or the user clicks the stop button to pause or end the dialogue.

[0097] In one embodiment, the method further comprises the following steps C1-C3:

[0098] C1. Save the conversation information in the conversation record.

[0099] C2. Obtain corresponding conversation information from the conversation record according to the keyword, where the keyword is included in the conversation information.

[0100] C3. Load the conversation information through the conversation interface of the large language model.

[0101] For example, when a user is having a concert reminder conversation, he or she can input relevant performance information of the concert into the conversation interface of the large language model, such as an introduction to the orchestra, the conductor, and the repertoire. The user can have a conversation and discuss with the large language model, and at the same time, the client will save the conversation information in the conversation record.

[0102] When the user exits the conversation interface of the large language model and then reopens it, the user can enter the keywords in the reminder content of the conversation information, and then the client searches for the conversation information including the keywords in the conversation list and loads (activates, wakes up) the conversation information dialogue. Furthermore, the client can also send the conversation information to the large language model end, so that the large language model end activates the reminder dialogue according to the context of the conversation information, so that the user and the large language model can continue the conversation.

[0103] In one embodiment, the method further comprises the following steps D1-D3:

[0104] D1. Save the reminder information in the reminder information list.

[0105] D2. Obtain corresponding reminder information from the reminder list according to the keyword, and the keyword is included in the reminder information.

[0106] D3. Load reminder information through the dialogue interface of the large language model.

[0107] For example: after receiving the reminder information sent by the server, the client can also save the reminder information in the reminder information list. When the user exits the dialogue interface of the large language model and reopens it, the user can enter the keywords in the reminder content of the reminder information. Then the client searches for the reminder information in the reminder information list and loads the reminder information through the dialogue interface of the large language model.

[0108] In one embodiment, the method further comprises the following steps E1-E3:

[0109] E1. Receive inquiry information sent by the server, where the inquiry information has the semantics of whether the user has completed the reminder task according to the reminder content of the reminder information.

[0110] E2. Output inquiry information through the dialogue interface of the large language model.

[0111] E3. Send the reply information to the query information input by the user in the dialogue interface of the large language model according to the query information to the server.

[0112] In one embodiment, if the user wants the large language model to execute a reminder, but the dialogue interface of the large language model of the client is not open, before sending the dialogue information input by the user in the dialogue interface of the large language model of the client to the server, it also includes: receiving a voice input by the user, the voice has the semantics of opening the dialogue interface of the large language model; and then the client opens the dialogue interface of the large language model according to the voice.

[0113] In one embodiment, when the reminder information input by the user in the dialogue interface of the large language model of the client is an image or a document, the client sends the image or document to the server for analysis and comprehensive summary, and extracts the reminder information and returns it to the client.

[0114] For example: the user sends a conference promotional picture or poster to the large language model through the client. The large language model recognizes and analyzes the text content and image description content in the picture, summarizes the content, extracts the reminder time and reminder content in the content, and returns it to the client through the server, telling the user a summary of the meeting, including the meeting name, time and location.

[0115] In an embodiment similar to the above-mentioned meeting reminder, the user sends a picture of a concert ticket to the large language model through the client. The large language model recognizes and parses the text content and image description content in the picture, summarizes the content, extracts the reminder time and reminder content in the content, and returns it to the client through the server, telling the user that this is a ticket for a concert and a summary of the concert, including the name of the concert, performance time and venue.

[0116] Furthermore, the large language model recognizes and parses the text content and image description content in the picture, and summarizes the content. If the content includes the semantics of the destination the user wants to go to, a query is sent to the user client on how to get to the destination or what kind of transportation to take to the destination.

[0117] The server sends a query to the user or the large language model gives the query and then the server sends it to the user client to ask the user: "How long in advance do you want to go to the concert? Do you want to drive or take the bus or subway? I will help you calculate the time." The user replies: "Take the bus or subway."

[0118] The server receives and parses the user's reply, looking for specific semantics on how the user goes to the destination or what means of transportation to take to the destination in the reply. If so, the server calls the map API, plans the transportation route based on the starting user's location and the end destination, calculates the time to reach the destination, and sends it to the user client.

[0119] When the user arrives at the entrance ticket inspection channel of the concert, the user inputs the instruction of opening the ticket to the large language model by voice or text, and the server and the large language model receive and respond to the instruction, and send the ticket with the ticket inspection QR code to the client for display, and the user shows the ticket QR code to the ticket inspection scanning device to complete the ticket inspection. This embodiment can also be used for reminders in scenes such as watching movies, taking trains, meetings, exhibitions, etc., and can conveniently and quickly open the entrance ticket inspection QR codes of movie tickets, train tickets, etc. for scanning by the ticket inspection device.

[0120] The above-mentioned embodiment of concert reminder can also be used for reminders of various conferences and exhibitions as well as travel reminders by train or airplane.

[0121] Figure 2 is a flowchart of a method for processing time information based on a large language model according to an exemplary embodiment. Figure 2 As shown, the method is used for a server, and the method includes the following steps S201-S207:

[0122] In step S201, a conversation message sent by a client is received; the conversation message includes at least one of the following information: text, voice, picture, video and document, and the conversation message includes reminder information, and the reminder information includes reminder time and reminder content.

[0123] In step S202, the dialogue information is parsed and sent to the large language model end;

[0124] When the dialog information received by the server is text, the reminder information included in the text can be directly parsed and the text can be sent to the large language model end.

[0125] When the conversation information received by the server is voice, the server can parse the voice into text through the built-in voice analysis software, and then parse the reminder information included in the text, and send the parsed text to the large language model end. The server can also directly send the language to the large language model end, and the large language model end parses the voice into text and sends the parsed text to the server. At this time, the server parses the reminder information included in the text, and the large language model end also parses the reminder information included in the text.

[0126] When the conversation information received by the server is a picture, the server can parse the picture into text through the built-in picture parsing software, and then parse the reminder information included in the text, and send the parsed text to the large language model end. The server can also directly send the language to the large language model end, and the large language model end parses the picture into text and sends the parsed text to the server. At this time, the server parses the reminder information included in the text, and the large language model end also parses the reminder information included in the text.

[0127] When the conversation information received by the server is a video, the server can parse the video into text through the built-in video parsing software, and then parse the reminder information included in the text, and send the parsed text to the large language model end. The server can also directly send the language to the large language model end, and the large language model end parses the video into text and sends the parsed text to the server. At this time, the server parses the reminder information included in the text, and the large language model end also parses the reminder information included in the text.

[0128] In step S203, response information to the dialogue information sent by the large language model is received.

[0129] After receiving the conversation information, the large language model will obtain the response information corresponding to the conversation information in order to interact with the user.

[0130] When the type of dialogue information received by the large language model is text, the large language model directly obtains the response information corresponding to the text. When the type of dialogue information received by the large language model is other types (such as voice, picture or video), the large language model will first convert other types of dialogue information into text, or perform image recognition on pictures and video streams, parse the image content in the pictures and each video frame, and then output the response information for the text or image content.

[0131] For example: A user asks a large language model to tell a story and stops after half an hour.

[0132] The user enters the following in the dialogue interface of the large language model: Xiaoyu (name or nickname of the large language model), please tell me the Romance of the Three Kingdoms for half an hour, continuing from what you told me last time. I want to fall asleep while listening.

[0133] Xiaoyu responded with the following message: OK, master, I will continue to talk about the Romance of the Three Kingdoms from what I did last time. I will stop after half an hour so that you can fall asleep while listening. Let’s start now.

[0134] Another example: A user asks the large language model to remind him that he has a meeting at 10 o'clock and to be reminded to attend the meeting at 9:50.

[0135] The user inputs in the dialogue interface of the large language model: Xiaoyu, I have a meeting at 10 o'clock, please remind me to attend it at 9:50.

[0136] Xiaoyu responded with the following message: OK, master, you have a meeting at 10 o'clock. I will remind you to attend the meeting at 9:50.

[0137] In step S204, the response information is parsed to extract the reminder information contained in the dialogue information and / or the response information by parsing the dialogue information and / or the response information.

[0138] In the present disclosure, the reminder information contained in the conversation information can be extracted by parsing the conversation information, the reminder information contained in the response information can be extracted by parsing the response information, or the reminder information contained in the conversation information and the response information can be extracted by parsing the conversation information and the response information.

[0139] The two-dimensional acquisition methods of obtaining reminder information by parsing the conversation information by the server itself and obtaining reminder information by parsing the response information of the large language model by the server can make the obtained reminder information more accurate. Because if the reminder information obtained by the server itself parsing the conversation information is different from the reminder information obtained by the server by parsing the response information, it means that the reminder information input by the user may be ambiguous, or there is a problem with the response parsed by the server or the large language model. At this time, the server can send "Please confirm which reminder information you want, and give two options for reminder information" to the client, and the client outputs "Please confirm which reminder information you want, and give two options for reminder information" through the dialogue interface of the large language model, so that the user can confirm it on the client, which improves the accuracy of the reminder information. If the reminder information obtained by the server itself parsing the conversation information is the same as the reminder information obtained by the server by parsing the response information, it means that the reminder information is correct, and the server can continue to execute other steps.

[0140] The server can improve the processing efficiency of the entire system by sending the conversation information to the large language model while parsing the conversation information.

[0141] For example, a user inputs in the dialogue interface of the large language model: Xiaoyu, I have a meeting at 10 o'clock, please remind me to attend at 9:50. Xiaoyu responds with the following message: OK, master, you have a meeting at 10 o'clock, I will remind you to attend the meeting at 9:50. The server obtains the reminder message "9:50, attend the meeting" by parsing "Xiaoyu, I have a meeting at 10 o'clock, please remind me to attend at 9:50" and "OK, master, you have a meeting at 10 o'clock, I will remind you to attend the meeting at 9:50".

[0142] In step S205, the obtained reminder information is marked as timed reminder information and stored in a database.

[0143] After obtaining the reminder information, the server will mark the reminder information as scheduled reminder information and store it in the database.

[0144] In step S206, a timer is set according to the timed reminder time of the timed reminder information and starts timing; the timed reminder time is the reminder time corresponding to the timed reminder information.

[0145] After the server stores the timed reminder information in the database, the server sets a timer according to the timed reminder time of the timed reminder information and starts timing.

[0146] In step S207, it is detected whether the current time reaches the timer reminder time. If it reaches the timer reminder time, the timer reminder information including the timer reminder time that has been reached is sent to the client, and the client executes the reminder task.

[0147] When the server detects that the current time reaches the timed reminder time corresponding to the timed reminder information, it sends the timed reminder information including the timed reminder time that has been reached to the client, and the client executes the reminder task.

[0148] In one embodiment, the server stores multiple different timed reminder information in a database, and sets a timer for each timed reminder information, and regularly checks to see which timed reminder time corresponding to which timed reminder information the current time has reached. When it is detected that the current time has reached the timed reminder time corresponding to a certain timed reminder information, the server will send the timed reminder information corresponding to the timed reminder time to the client, so that the client can be reminded of the timed reminder information.

[0149] Of course, in one embodiment, after parsing the reminder information, the server can also send the reminder information directly to the client, and then the client sets the timer and starts timing. After the client detects that the current time has reached the scheduled reminder time, it executes the reminder task according to the reminder information; the client can also save multiple reminder information, and then the client sets a timer for each reminder information. When it detects that the current time has reached the reminder time corresponding to a certain reminder information, the client will remind the reminder for the reminder information.

[0150] In the present disclosure, the client further includes a logic processing module, which is used to mark the timed reminder information sent by the server to the client as client timed reminder information, and store it in a data storage module or cache; it is also used to set a timer and start timing according to the timed reminder time of the client timed reminder information; the client timed reminder time is the same reminder time as the timed reminder information sent by the server; it is also used to detect whether the current time reaches the timed reminder time, and if so, trigger the reminder logic.

[0151] The client further includes a time acquisition module for acquiring the current time of the client system.

[0152] In the present disclosure, the server detecting whether the current time reaches the timed reminder time includes: the server detecting whether the current time has accurately reached the timed reminder time, or the server detecting whether the current time reaches a preset time period before the timed reminder time.

[0153] Among them, the server detects whether the current time has accurately reached the timed reminder time, for example: the timed reminder time is 8:00, then the server detects whether the current time has reached 8:00, and when it reaches 8:00, the server sends the timed reminder information to the client, and the client immediately executes the reminder logic after receiving the timed reminder information. The server detects whether the current time has reached the preset time before the timed reminder time, for example: the timed reminder time is 8:00, then the server detects whether the current time has reached the preset time 7:59, and when it reaches 7:59, the server sends the timed reminder information to the client, and the client's logic processing module and time acquisition module perform preprocessing and execute the reminder task, so that network delay problems caused by factors such as poor network signals can be avoided.

[0154] The method steps of the above embodiment are:

[0155] When the client receives the timed reminder information sent by the server, the timed reminder information is marked as client timed reminder information and stored in the data storage module or cache;

[0156] Obtain the current time of the client system, and set a timer according to the timed reminder time of the client timed reminder information and start timing;

[0157] Detect whether the current time of the client system reaches the scheduled reminder time, and if so, trigger the reminder logic. In one embodiment, parsing the dialogue information and parsing the response information to extract the reminder information contained in the dialogue information and the response information by parsing the dialogue information and the response information includes the following sub-steps E1-E2:

[0158] E1. Parse the conversation information and the response information to detect whether there is a semantic description of the reminder information confirmed by the user and the large language model.

[0159] E2. If it is detected that there is a semantic description of the reminder information that is jointly confirmed, the reminder information contained in the dialogue information and the response information is extracted.

[0160] When the large language model gives suggestions or replies on reminder time and reminder events, it is usually safe for the user to say in the conversation with the large language model: "I have a meeting at 10:30, remind me at 10:20, and the large language model replies: OK, meeting at 10:30, I'll remind you at 10:20." If it is detected that the reply of the large language model has been read by the user, and the user does not reply with a message like "OK" within the set time, the system will assume that the user has confirmed it, and the reminder time setting is complete. If the user has set in advance that the final reminder information must be confirmed by consensus between the user and the large model, the public detection module will perform detection and confirmation.

[0161] In one embodiment, parsing the response information to extract the reminder information contained in the dialogue information and the response information by parsing the dialogue information and the response information includes the following sub-steps F1-F3:

[0162] F1. Parse the response information;

[0163] F2. When detecting that the parsed response information includes inquiry information asking the user whether to confirm the reminder information, sending the inquiry information to the client;

[0164] F3. After receiving the confirmation information input by the user, extract the reminder information contained in the dialogue information and the response information.

[0165] For example: 1. The user asks the large language model to tell a story and stops after half an hour:

[0166] User: Xiaoyu (name or nickname of the large language model), please tell me the Romance of the Three Kingdoms for half an hour, continuing from what you told me last time. I want to fall asleep while listening.

[0167] Xiaoyu: Okay, master, I will continue talking about the Romance of the Three Kingdoms from last time. I will stop after half an hour so that you can fall asleep while listening. Let’s start now.

[0168] User: OK.

[0169] 2. The user asks the large language model to remind him that he has a meeting at 10 o'clock and to be reminded to attend the meeting at 9:50.

[0170] User: Xiaoyu, I have a meeting at 10 o'clock. Please remind me to attend it at 9:50.

[0171] Xiaoyu: OK, master, you have a meeting at 10 o'clock. I will remind you to attend the meeting at 9:50.

[0172] User: OK.

[0173] The confirmation message in both examples above is "OK".

[0174] In one embodiment, the obtained reminder information is marked as timed reminder information and stored in a database, including:

[0175] When the reminder information parsed from the conversation information includes a reminder instruction, the reminder information is directly marked as a timed reminder information and stored in the database;

[0176] or,

[0177] When the reminder information parsed from the response information includes a reminder instruction, the reminder information is directly marked as timed reminder information and stored in the database.

[0178] For example, the user's dialogue message is: "I am boiling eggs, remind me in five minutes." The dialogue message contains the reminder time (after 5 minutes) and the reminder instruction (boil eggs). At this time, the reminder message is directly marked as a timed reminder message, stored in the database, and a 5-minute timer is set and the timing is started according to the timed reminder time of the timed reminder message. Alternatively, the response message replied by the large language model is: "Okay, you are boiling eggs, I will remind you in 5 minutes." The response message contains the reminder time (after 5 minutes) and the reminder instruction (boil eggs). At this time, the reminder message is directly marked as a timed reminder message, stored in the database, and a 5-minute timer is set and the timing is started according to the timed reminder time of the timed reminder message.

[0179] In one embodiment, the method further comprises the following steps G1-G5:

[0180] G1. After the scheduled reminder time arrives, a query message about the user's execution status is sent to the client, and the query message has the semantics of whether the user has completed the reminder task according to the reminder content of the reminder message;

[0181] G2, receiving reply information to the query information input by the user in the dialogue interface of the large language model according to the query information, sent by the client;

[0182] G3. Analyze the reply message to see if it has completed the reminder task;

[0183] G4. If not, start checking and asking for the timing time;

[0184] G5. If yes, terminate the inquiry and set a timer to remind you that the task is complete.

[0185] After detecting that the scheduled reminder time has arrived, the server will send an inquiry about the user's execution status to the client, asking whether the user has completed the reminder task according to the reminder content of the reminder information. After receiving the client's reply information, the server will parse the reply information to determine whether the reminder information contains the semantics of completing the reminder task. When the reply information contains the semantics of completing the reminder task, the inquiry is terminated and the scheduled reminder task is completed. When the reply information does not contain the semantics of completing the reminder task, indicating that the client has not successfully executed the scheduled reminder task, the check query timing time is started. When each check query timing time arrives, a check query about the user's execution status is sent to the client until the received reply information contains the semantics of completing the reminder task.

[0186] The parsing of the reply information may be performed by the server itself, or the server may send the reply information to the large language model end, which will parse the reply information and then return it to the server.

[0187] It is worth noting that the inquiry about the user's execution status has the semantics of whether the user has completed the reminded task according to the reminder content of the reminder information, and the inquiry content will vary depending on the reminder content in different reminder information, but it must include the semantics of whether the user has completed the reminded task according to the reminder content of the reminder information.

[0188] For example:

[0189] Inquiries about medication reminders:

[0190] Two minutes after the client executes the reminder task to remind the user, the user is asked: "Have you taken the medicine?" If the user replies: "I have taken it", when the server parses the semantics of "I have taken it", it determines that the user has completed the reminder task, then the inquiry is terminated and the scheduled reminder task is completed.

[0191] If the user replies: "The water is too hot, wait a moment.", the timed inquiry time is started. When the timed inquiry time is up, the user is asked again: "Have you taken the medicine?". If the user replies in the dialogue interface of the large language model: "I have taken it", when the server parses the semantics of "I have taken it", it is determined that the user has completed the reminder task, so the inquiry is terminated and the timed reminder task is completed.

[0192] The present disclosure also provides a time information processing system based on a large language model, such as Figure 3 As shown, the time information processing system based on the large language model includes: a client 1, a server 2 and a large language model terminal 3;

[0193] The server communicates with the client and connects to the large language model through the API interface;

[0194] The client includes: user interface module, input and output module, data storage module, notification and reminder module;

[0195] The server includes: a middleware module, a natural language processing module, a logic processing module and a database;

[0196] A user interface module, used to provide a conversational interface of a large language model to the user;

[0197] The input / output module is used to send the dialogue information input by the user in the dialogue interface of the large language model to the middleware module, so that the middleware module processes the dialogue information and forwards it to the large language model end; the dialogue information includes at least one of the following information: text, voice, picture, video or document, and the dialogue information includes reminder information, and the reminder information includes reminder time and reminder content;

[0198] The middleware module is used to coordinate the communication between the client and the large language model API, process the client request, and receive and send the dialogue information sent by the client input and output module; the dialogue information includes at least one of the following information: text, voice, picture, video and document, and the dialogue information includes reminder information, which includes reminder time and reminder content;

[0199] The natural language processing module is used to parse the conversation information and send it to the large language model end;

[0200] The middleware module is also used to receive response information to the dialogue information sent by the large language model end;

[0201] A natural language processing module, used to parse the response information of the large language model, so as to extract the reminder information contained in the dialogue information and / or the response information by parsing the dialogue information and the response information;

[0202] A logic processing module, used for marking the obtained reminder information as timed reminder information and storing it in a database;

[0203] The logic processing module is further used to set a timer and start timing according to the timed reminder time of the timed reminder information; the timed reminder time is the reminder time corresponding to the timed reminder information;

[0204] The logic processing module is also used to detect whether the current time reaches the timer reminder time. If it reaches the timer reminder time, the reminder logic is triggered. The middleware module is also used to send the timer reminder information including the timer reminder time that has been reached to the data storage module, and the notification and reminder module executes the reminder task;

[0205] A data storage module is used to receive reminder information sent by the server, and the reminder information includes timed reminder information;

[0206] The notification and reminder module is used to trigger the reminder logic when it detects that the current time has reached the scheduled reminder time, and execute the reminder task according to the scheduled reminder information.

[0207] In one embodiment, the natural language processing module is specifically used to: parse the response information; when detecting that the parsed response information includes inquiry information asking the user whether to confirm the reminder information, send the inquiry information to the input and output module;

[0208] The input and output module is also used to receive the query information sent by the natural language processing module, display the query information to the user in the dialogue interface of the large language model, and prompt the user to confirm; and send the confirmation information entered by the user to the natural language processing module;

[0209] The natural language processing module is also used to extract the reminder information contained in the dialogue information and / or the response information after receiving the confirmation information input by the user.

[0210] In one embodiment, the user interface module is further used to provide a setting button for pushing personalized reminder information in the dialogue interface of the large language model;

[0211] The input / output module is further configured to output a selection menu of personalized reminder information in response to a press of the setting button, wherein the selection menu of personalized reminder information includes at least one of the following personalized information: image, audio, video, 2D animation, 3D animation, virtual image and digital human; and receive target personalized information selected by the user in the selection menu;

[0212] The notification and reminder module is specifically used to trigger the reminder logic when the current time has reached the scheduled reminder time, execute the reminder task according to the scheduled reminder information, and output the target personalized information.

[0213] In one embodiment, the logic processing module is further used to save the conversation information in a conversation record; obtain corresponding conversation information from the conversation record according to a keyword, wherein the keyword is included in the conversation information;

[0214] The input-output module is also used to load conversation information through the conversation interface of the large language model.

[0215] In one embodiment, the logic processing module is further used to save the reminder information in a reminder information list; obtain corresponding reminder information from the reminder list according to a keyword, wherein the keyword is included in the reminder information;

[0216] The input-output module is also used to load the reminder information through the dialogue interface of the large language model.

[0217] In one embodiment, the middleware module is further used to send a query message about the user's execution status to the client after the scheduled reminder time arrives, and the query message has the semantics of whether the user has completed the reminder task according to the reminder content of the reminder message;

[0218] The input and output module is also used to receive the query information sent by the server, output the query information through the dialogue interface of the large language model; and send the confirmation information of the query information input by the user in the dialogue interface of the large language model according to the query information to the server;

[0219] The middleware module is also used to analyze whether the confirmation information contains the semantics of whether the reminder task has been completed; if not, start checking and inquiring about the timing time; if yes, terminate the inquiry and the timing reminder task is completed.

[0220] In one embodiment, the notification and reminder module is further used to: output reminder information or timed reminder information by voice; and / or output timed reminder information by text.

[0221] In one embodiment, the natural language processing module is also used to parse the dialogue information and the response information to detect whether there is a semantic description of the reminder information jointly confirmed by the user and the large language model end; if it is detected that there is a semantic description of the reminder information jointly confirmed, the reminder information contained in the dialogue information and the response information is extracted.

[0222] In one embodiment, the middleware module is further configured to directly mark the reminder information as timed reminder information and store it in a database when the reminder information parsed from the dialog information includes a reminder instruction.

[0223] In one embodiment, it also includes a time module, which is deployed in the client and / or the server;

[0224] When the time module is deployed in the client, the time module is used to obtain the system time of the client, display it to the user through the user interface module, and send it to the notification and reminder module;

[0225] When the time module is deployed in the server, the time module is used to obtain the system time of the server and send it to the logic processing module.

[0226] In one embodiment, a map API is also included for providing business services for geographic location information included in the reminder information, and the middleware module coordinates the connection;

[0227] The notification and reminder module is further configured to trigger the reminder logic when it is detected that the current time has reached the scheduled reminder time and the geographical location information of the user client is located at a preset geographical location.

[0228] The client of the present disclosure also includes an information display module, which is used to display the timed reminder information including the time and content of the timed reminder; when the timed reminder information is confirmed, the timer is started, and the timed reminder time and the content of the reminder or the display summary is displayed in the dialogue interface of the large language model in a floating window or floating button manner or in a following page display box manner; clicking the floating window or floating button or display box can open the details of the timed reminder information; or, clicking the floating window or floating button or display box can edit and modify the timed reminder information; or, setting a switch button in the floating window or floating button or display box, clicking the switch button in the floating window or floating button or display box can turn off or on the timed reminder information; or, the information display module is also used to display the current time and the timed reminder information. When the information display module is used to display the current time and the timed reminder information, it is displayed in at least one of the full screen, half screen or less than half screen modes, and clicking the screen or the exit button in the screen can exit the display mode.

[0229] The scheduled reminder time and reminder content or a display summary is displayed in the dialogue interface in a floating window or floating button manner, and the display summary is: a summary of the scheduled reminder time and reminder content; for example, for a meeting reminder information with a scheduled reminder time of 10 o'clock, it is displayed: meeting at 10:00.

[0230] The information display module is also used to display the current time; the display of the current time is: displaying the scheduled reminder time and reminder content in a floating window or floating button in the dialogue interface or displaying the current time in the display summary, so that the user can understand the time difference between the current time and the scheduled reminder time at any time; for example, for a meeting reminder information with a scheduled reminder time of 10 o'clock, if the current time is 9:25, it will be displayed: 10:00 meeting / 9:25, the display of the scheduled reminder time 10:00 does not change, and the display of the current time 9:25 changes. There is no limitation on the specific display format and arrangement method, as long as it includes content such as 10:00 meeting / 9:25.

[0231] A further embodiment is: calculating the time difference between the current time and the timed reminder time, and displaying them synchronously, so that the user can more easily understand the time difference between the current time and the timed reminder time; for example, displaying: 1 hour and 20 minutes left until 10:00 / meeting. The client of the present disclosure also includes a lock screen display module, which is used for lock screen state detection, and displays the timed reminder information and reminder notification including the time and reminder content of the timed reminder in the lock screen state.

[0232] The method steps are: detecting whether the client has entered the lock screen state, and if so, displaying a reminder notification of the arrival of the scheduled reminder time and / or displaying the set scheduled reminder time and reminder content in the lock screen state; or, when the current time reaches the scheduled reminder time in the lock screen state, triggering the reminder logic, executing the reminder task according to the scheduled reminder information, displaying the reminder information and notification in the lock screen state, and / or exiting the lock screen to open the client to execute the reminder task.

[0233] One embodiment is: the lock screen state detection is to use a key lock manager (KeyguardManager) to monitor and detect screen closing events, and display timed reminder information and reminder notifications in the lock screen state.

[0234] Further embodiments are:

[0235] Check the lock screen status. When the current time reaches the scheduled reminder time or the timer with scheduled reminder information is running, use KeyguardManager to check whether the current screen is in the lock screen state;

[0236] If the screen is in lock screen mode, reminder information and notifications with the lock screen interface are displayed.

[0237] The method of using a keyguard manager (KeyguardManager) to monitor and detect screen closing events and displaying timed reminder information and reminder notifications in a locked screen state also includes:

[0238] 1. Create a lock screen interface activity: Create a LockScreenActivity to display reminder information in the lock screen state; when the device is in the lock screen state and the user receives a notification, the reminder information can be displayed through this activity. The user can interact with the reminder through this activity, such as reading the reminder content, confirming the reminder, etc.

[0239] The specific implementation steps are:

[0240] Create a lock screen interface activity: define a new activity class LockScreenActivity.

[0241] Set the layout file: Create a layout file activity_lock_screen.xml for this activity to display reminder information.

[0242] Handle user interactions: Handle user interactions such as click events in the activity.

[0243] 2. Register LockScreenActivity to display detailed reminder content, including user interaction: after receiving a notification in the lock screen state, the user can enter LockScreenActivity by clicking on the notification to view the detailed reminder content; Enhance user experience: provide a clear and intuitive interface so that users can quickly understand the specific content of the reminder. It is also used to handle user operations, including confirmation reminders: users can confirm reminders in LockScreenActivity, such as clicking the "OK" button to close the reminder; other operations: according to the needs of the application, more user operations can be added in LockScreenActivity, such as replying to messages, jumping to other pages, etc.

[0244] 3. Define the lock screen theme: The purpose is to define a special theme so that the activity can be displayed correctly in the lock screen state; it is used to control the display style of the activity to adapt it to the lock screen interface; ensure that the activity can be displayed in the lock screen state and will not be blocked by the system default lock screen interface.

[0245] It can be seen from the above embodiments that the currently running scheduled reminder tasks and reminder notifications are displayed using the lock screen state, so that the user can see the running status of the scheduled reminder tasks even when the client is locked, which makes it easier for the user to view the scheduled reminder tasks when it is inconvenient to open the client. It provides the user with an additional window for viewing and interacting with reminder information, and provides a better user experience.

[0246] The present disclosure also includes: a smart device;

[0247] The smart device at least includes a voice input module, an audio output module, a processor module, a storage module and a wireless communication module;

[0248] The voice input module includes a microphone or a microphone array for capturing the user's voice;

[0249] The audio output module includes an audio driving circuit and a speaker or earphones for playing reminder information;

[0250] The processor module includes a processor, which is a core component for running an operating system and application programs;

[0251] The storage module includes a memory for storing an operating system and application data;

[0252] The wireless communication module includes a Bluetooth or Wi-Fi or 4G or 5G communication module, which is used to communicate with a client device or a server.

[0253] In one embodiment, the smart device further includes: a camera module;

[0254] The camera module is used to capture pictures or videos.

[0255] In one embodiment, the smart device further includes a display screen for displaying the dialogue between the user and the large language model and reminder information.

[0256] The smart devices also include smart wearable devices, specifically including smart glasses, AR glasses, smart headphones, and smart watches.

[0257] The smart device includes a vehicle-mounted smart terminal, which can be used as a second client of the user. In one embodiment, the server also includes:

[0258] A speech recognition module, used to convert the user's input speech into text;

[0259] The natural language processing module is used to parse the text converted by the speech recognition module;

[0260] Text-to-speech (TTS) module or access to the text-to-speech API: converts system responses into speech output for the user.

[0261] Figure 4 The present invention is a block diagram of a device for processing time information based on a large language model according to an exemplary embodiment, and the device is applicable to a terminal device. For example, the device may be a mobile phone (cell phone), a computer, a digital broadcast terminal, a message transceiver device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, a vehicle-mounted intelligent terminal, etc.

[0262] The device may include one or more of the following components: a processing component, a memory, a power component, a multimedia component, an audio component, an input / output (I / O) interface, a sensor component, and a communication component.

[0263] The processing component generally controls the overall operation of the device, such as operations associated with display, phone calls, data communications, camera operation, and recording operation. The processing component may include one or more processors to execute instructions to complete all or part of the steps of the above-mentioned method. In addition, the processing component may include one or more modules to facilitate interaction between the processing component and other components. For example, the processing component may include a multimedia module to facilitate interaction between the multimedia component and the processing component.

[0264] The memory is configured to store various types of data to support operations on the device. Examples of such data include instructions for any application or method operating on the device, contact data, phone book data, messages, pictures, videos, etc. The memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0265] The power supply components provide power to the various components of the device. The power supply components may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the device.

[0266] The multimedia component includes a screen that provides an output interface between the device and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component includes a front camera and / or a rear camera. When the device is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera may receive external multimedia data. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

[0267] The audio component is configured to output and / or input audio signals. For example, the audio component includes a microphone (MIC), and when the device is in an operating mode, such as a call mode, a recording mode, and a speech recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in a memory or sent via a communication component. In some embodiments, the audio component also includes a speaker for outputting an audio signal.

[0268] The I / O interface provides an interface between the processing component and the peripheral interface module, which may be a keyboard, a click wheel, buttons, etc. These buttons may include but are not limited to: a home button, a volume button, a start button, and a lock button.

[0269] The sensor assembly includes one or more sensors for providing various aspects of the status assessment of the device. For example, the sensor assembly can detect the open / closed state of the device, the relative positioning of components, such as the display and keypad of the device, and the sensor assembly can also detect changes in the position of the device or a component of the device, the presence or absence of user contact with the device, the orientation or acceleration / deceleration of the device, and the temperature change of the device. The sensor assembly may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly may also include an accelerometer, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0270] The communication component is configured to facilitate wired or wireless communication between the device and other devices. The device can access a wireless network based on a communication standard, such as WiFi, 4G or 5G, or a combination thereof. In an exemplary embodiment, the communication component receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0271] In an exemplary embodiment, the apparatus may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above methods.

[0272] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory including instructions, and the instructions can be executed by a processor of the device to perform the above method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0273] A non-temporary computer-readable storage medium, when the instructions in the storage medium are executed by a processor of a device, enables the device to execute the time information processing method based on a large language model corresponding to the above-mentioned client.

[0274] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a time information processing method based on a large language model corresponding to the server is implemented.

[0275] Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0276] In an exemplary embodiment, a computer device is also provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps in the embodiment of the time information processing method based on a large language model corresponding to the above-mentioned server are implemented.

[0277] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, which, when executed by a processor, implements the steps in the embodiment of the time information processing method based on a large language model corresponding to the above server.

[0278] In an exemplary embodiment, a computer program product is provided, including a computer program, which, when executed by a processor, implements the steps in the embodiment of the time information processing method based on a large language model corresponding to the above server.

[0279] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0280] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0281] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0282] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0283] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for processing time information based on a large language model, characterized in that: The method is applied to a client, and the method comprises: Sending the dialogue information input by the user in the dialogue interface between the client and the large language model to the server, so that the server processes the dialogue information and forwards it to the large language model end; the dialogue information includes at least one of the following information: text, voice, picture, video or document, and the dialogue information includes reminder information, and the reminder information includes reminder time and reminder content; Receiving reminder information sent by the server, the reminder information including timed reminder information, the timed reminder information being reminder information confirmed after the server parses the conversation information and / or the response information of the large language model end to the conversation information; the timed reminder information includes: timed reminder time and timed reminder event; If the current time has reached the timer reminder time, the reminder logic is triggered and the reminder task is executed according to the timer reminder information.

2. The time information processing method based on a large language model according to claim 1, characterized in that: Before receiving the timing reminder information sent by the server, the method further includes: receiving the inquiry information sent by the server; the inquiry information has the semantics of whether the user has completed the reminder task according to the reminder content of the reminder information, or the inquiry information has the semantics of asking the user whether to confirm the reminder information; Displaying the inquiry information to the user on the dialogue interface of the large language model, prompting the user to confirm; The confirmation information input by the user is sent to the server.

3. The time information processing method based on a large language model according to claim 1, characterized in that: The reminder information also includes geographic location information; when the current time reaches the scheduled reminder time, the reminder logic is triggered, including: When it is detected that the current time has reached the scheduled reminder time and the real-time geographical location of the client is located at the position corresponding to the geographical location information, the reminder logic is triggered.

4. The time information processing method based on a large language model according to claim 1, characterized in that: When the timed reminder information is confirmed, a timer is started, and the timed reminder time and the reminder content or a summary of the timed reminder information is displayed in a floating window or a floating button or a following page display frame on the dialogue interface of the large language model; Clicking the floating window or the floating button or the display box can open the details of the timed reminder information; Or, click the floating window or the floating button or the display box to edit and modify the timed reminder information; Alternatively, a switch button may be set in the floating window, the floating button or the display frame, and the timed reminder information may be turned off or on by clicking the switch button in the floating window, the floating button or the display frame.

5. The time information processing method based on a large language model according to claim 1, characterized in that: Pushing a setting button for personalized reminder information on the dialogue interface of the large language model; In response to pressing the setting button, outputting a selection menu of personalized reminder information, wherein the selection menu of personalized reminder information includes at least one of the following personalized information: image, audio, video, 2D animation, 3D animation, avatar, and digital human; Receive the target personalized information selected by the user in the selection menu; When the current time reaches the timer reminder time, the reminder logic is triggered, and the reminder task is executed according to the timer reminder information, including: When the current time has reached the timed reminder time, the reminder logic is triggered, the reminder task is executed according to the timed reminder information, and the target personalized information is output.

6. The time information processing method based on a large language model according to claim 1, characterized in that: The reminder information also includes at least the reminder time and reminder name among the following information: reminder time, reminder name, reminder quantity, reminder content, image and document of reminder content; And / or, the response information includes at least one of the following information: reminder time, reminder name, reminder quantity, reminder content, image and document of reminder content, and common sense, precautions or suggestions associated with the reminder information.

7. The time information processing method based on a large language model according to claim 1, characterized in that: The method further comprises: Saving the conversation information in a conversation record; Acquire corresponding conversation information from the conversation record according to a keyword, wherein the keyword is included in the conversation information; Loading the dialogue information through the dialogue interface of the large language model; and / or, saving the reminder information in a reminder information list; Acquire corresponding reminder information from the reminder list according to a keyword, wherein the keyword is included in the reminder information; The reminder information is loaded through the dialogue interface of the large language model.

8. The time information processing method based on a large language model according to claim 1, characterized in that: The method further comprises: Detect whether the screen is locked; If the lock screen state is entered, the reminder notification of the arrival of the scheduled reminder time and / or the set scheduled reminder time and the reminder content are displayed in the lock screen state; or, when the current time in the lock screen state reaches the scheduled reminder time, the reminder logic is triggered, the reminder task is executed according to the scheduled reminder information, the reminder information and notification are displayed in the lock screen state, and / or the lock screen is exited to open the client to execute the reminder task.

9. The time information processing method based on a large language model according to any one of claims 1 to 8, characterized in that: The trigger reminder logic includes: Outputting the reminder information and the timed reminder information by voice; and / or, Output the timing reminder information in the form of text; and / or, Output the timed reminder information by means of local notification; and / or, Output the timing reminder information by means of a message template; and / or, Outputting the timed reminder information by way of telephone notification; and / or, The timing reminder information is outputted by playing a ring tone.

10. A method for processing time information based on a large language model, characterized in that: The method is applied to a server, and the method comprises: Receiving a conversation message sent by a client; the conversation message includes at least one of the following information: text, voice, picture, video and document, and the conversation message includes reminder information, and the reminder information includes reminder time and reminder content; Parsing the dialogue information and sending the dialogue information to the large language model end; receiving response information to the dialogue information sent by the large language model end; Parsing the response information to extract the reminder information contained in the dialogue information and / or the response information by parsing the dialogue information and / or the response information; Marking the obtained reminder information as timed reminder information and storing it in a database; A timer is set according to the timed reminder time of the timed reminder information and starts timing; the timed reminder time is the reminder time corresponding to the timed reminder information; It is detected whether the current time reaches the timer reminder time. If it reaches the timer reminder time, the timer reminder information including the timer reminder time reaching is sent to the client, and the client performs the reminder task.

11. The time information processing method based on a large language model according to claim 10, characterized in that: The parsing of the dialogue information and the parsing of the response information, so as to extract the reminder information contained in the dialogue information and the response information by parsing the dialogue information and the response information, comprises: Parsing the dialogue information and the response information to detect whether there is a semantic description of the reminder information confirmed by the user and the large language model end; If it is detected that there is a semantic description of the reminder information that is commonly confirmed, the reminder information included in the dialog information and the response information is extracted.

12. The time information processing method based on a large language model according to claim 10, characterized in that: The parsing of the response information to extract the reminder information contained in the dialogue information and the response information by parsing the dialogue information and the response information includes: Parsing the response information; When detecting that the parsed response information includes inquiry information asking the user whether to confirm the reminder information, sending the inquiry information to the client; After receiving the confirmation information input by the user, the reminder information included in the dialogue information and the response information is extracted.

13. The time information processing method based on a large language model according to claim 10, characterized in that: The step of marking the obtained reminder information as timed reminder information and storing it in a database includes: When the reminder information parsed from the dialog information includes a reminder instruction, the reminder information is directly marked as the scheduled reminder information and stored in the database.

14. The time information processing method based on a large language model according to claim 10, characterized in that: The method further comprises: After the scheduled reminder time arrives, a query message about the user's execution status is sent to the client, wherein the query message has the semantics of whether the user has completed the reminder task according to the reminder content of the reminder message; receiving reply information to the query information input by the user in the dialogue interface of the large language model according to the query information, sent by the client; Analyze whether the reply message contains the semantics of completing the reminder task; If not, start checking and querying the timing time; If yes, the inquiry is terminated and the timed reminder task is completed.

15. A time information processing system based on a large language model, characterized in that: The time information processing system based on the large language model includes: a client, a server and a large language model end; The server is connected to the client through an API interface and is connected to the large language model end; The client includes: a user interface module, an input and output module, a data storage module, and a notification and reminder module; The server includes: a middleware module, a natural language processing module, a logic processing module and a database; The user interface module is used to provide a dialogue interface of the large language model to the user; The input-output module is used to send the dialogue information input by the user in the dialogue interface of the large language model to the middleware module, so that the middleware module processes the dialogue information and forwards it to the large language model end; the dialogue information includes at least one of the following information: text, voice, picture, video or document, and the dialogue information includes reminder information, and the reminder information includes reminder time and reminder content; The middleware module is used to coordinate the communication between the client and the large language model API, process client requests, receive and send dialogue information sent by the client input and output module; the dialogue information includes at least one of the following information: text, voice, picture, video and document, and the dialogue information includes reminder information, which includes reminder time and reminder content; The natural language processing module is used to parse the dialogue information and send the dialogue information to the large language model end; The middleware module is further used to receive response information to the dialogue information sent by the large language model end; The natural language processing module is used to parse the response information of the large language model, so as to extract the reminder information contained in the dialogue information and the response information by parsing the dialogue information and / or the response information; The logic processing module is used to mark the obtained reminder information as timed reminder information and store it in the database; The logic processing module is further used to set a timer and start timing according to the timed reminder time of the timed reminder information; the timed reminder time is the reminder time corresponding to the timed reminder information; The logic processing module is further used to detect whether the current time reaches the timer reminder time, and if so, trigger the reminder logic, and the middleware module is further used to send the timer reminder information including the timer reminder time reaching to the data storage module, and the notification and reminder module executes the reminder task; The data storage module is used to receive the reminder information sent by the server, and the reminder information includes timed reminder information; The notification and reminder module is used to trigger the reminder logic when detecting that the current time has reached the scheduled reminder time, and execute the reminder task according to the scheduled reminder information.

16. The time information processing system based on a large language model according to claim 15, characterized in that: The natural language processing module is specifically used to: parse the response information; when detecting that the parsed response information includes inquiry information asking the user whether to confirm the reminder information, send the inquiry information to the input and output module; The input-output module is further configured to, after receiving the inquiry information sent by the natural language processing module, display the inquiry information to the user on the dialogue interface of the large language model to prompt the user to confirm; and send the confirmation information input by the user to the natural language processing module; The natural language processing module is further used to extract the reminder information contained in the dialogue information and the response information after receiving the confirmation information input by the user.

17. The time information processing system based on a large language model according to claim 15, characterized in that: It also includes a map API for providing business services for the geographical location information included in the reminder information, and the middleware module coordinates the connection; The notification and reminder module is further configured to trigger the reminder logic when it is detected that the current time has reached the scheduled reminder time and the geographical location information of the user client is located at a preset geographical location.

18. The time information processing system based on a large language model according to claim 15, characterized in that: The system further comprises: a smart device; The smart device includes a voice input module, an audio output module, a processor module, a storage module, and a wireless communication module; The voice input module includes a microphone array for capturing the user's voice; The audio output module includes an audio driving circuit and a speaker, which is used to play the reminder information; The processor module includes a processor, which is a core component for running an operating system and application programs; The storage module includes a memory for storing an operating system and application data; The wireless communication module includes a Wi-Fi or Bluetooth module, which is used to communicate with the middleware module.

19. The time information processing system based on a large language model according to claim 18, characterized in that: The smart device also includes: a camera module; The camera module is used to capture the pictures or videos.

20. The time information processing system based on a large language model according to claim 15, characterized in that: The client also includes an information display module; The information display module is used to display the timed reminder information including the timed reminder time and the reminder content; when the timed reminder information is confirmed, the timer is started, and the timed reminder time and the reminder content or display summary are displayed in the dialogue interface of the large language model in a floating window or floating button manner or in a following page display box manner; click the floating window or the floating button or display box to open the details of the timed reminder information; Or, by clicking the floating window or the floating button or the display frame, the timed reminder information can be edited and modified; or, by setting a switch button in the floating window or the floating button or the display frame, the timed reminder information can be turned off or on by clicking the switch button in the floating window or the floating button or the display frame; Alternatively, the information display module is also used to display the current time and timing reminder information. When the information display module is used to display the current time and timing reminder information, it is displayed in at least one of the full screen, half screen, or less than half screen modes. Clicking the screen or an exit button in the screen can exit the display mode.

21. The time information processing system based on a large language model according to claim 15, characterized in that: The client also includes a lock screen display module; The lock screen display module is used for lock screen status detection, and displays the scheduled reminder information and reminder notification including the scheduled reminder time and the reminder content in the lock screen state.

Images