Intelligent reminding task creation method and device, electronic device, and storage medium

Abstract

The application provides a smart reminding task creation method and device, an electronic device and a storage medium, and belongs to the technical field of communication. The method comprises the following steps: extracting first event information and first time information from a user conversation, and determining a pre-created first reminding task; determining a corresponding first event type and a first preset dynamic time threshold according to the first event information; determining a conflict detection time range according to the first time information and the first preset dynamic time threshold; if there is no second reminding task created in the conflict detection time range, creating the pre-created first reminding task; if there is a second reminding task created, determining that the pre-created first reminding task conflicts with the second reminding task; and determining target prompt suggestion information based on the determination result of the conflict between the pre-created first reminding task and the second reminding task. The smart reminding task creation method and device, the electronic device and the storage medium provided by the application can improve the rationality of task arrangement.

Description

Technical Field

[0001] This application belongs to the field of communication technology, and more specifically, relates to a method and apparatus for creating intelligent reminder tasks, an electronic device, and a storage medium. Background Technology

[0002] In today's fast-paced life and work environment, task creation and reminder functions play a crucial role. They help users effectively manage time and schedule tasks, ensuring important matters are not overlooked, thereby improving efficiency and quality of life and work. Whether it's daily shopping list reminders or appointment reminders, or work-related meeting reminders or project deadline reminders, reminder task creation functions have become indispensable tools.

[0003] Existing reminder task creation methods typically involve users manually inputting event information and the desired reminder time. The system then creates a reminder task based on this information and sends a reminder to the user at the set time. While this traditional method meets basic user needs to some extent, it lacks intelligent detection of task time conflicts. When users manually input multiple reminder tasks, the inability to automatically detect conflicts leads to overlapping reminder times, causing scheduling chaos and wasted resources. Summary of the Invention

[0004] The purpose of this application is to provide a method, device, electronic device, and storage medium for creating intelligent reminder tasks, so as to perform conflict detection on reminder tasks according to event type and preset dynamic time threshold, thereby improving the rationality of task scheduling.

[0005] A first aspect of this application provides a method for creating intelligent reminder tasks, including:

[0006] Extract structured information from user conversations, and determine the first reminder task to be pre-created based on the structured information. The structured information includes first time information and first event information.

[0007] Determine the corresponding first event type and the first preset dynamic time threshold of the first event type based on the first event information;

[0008] The conflict detection time range is determined based on the first real-time information and the first preset dynamic time threshold.

[0009] If no second reminder task has been created within the conflict detection time frame, then a pre-created first reminder task will be created.

[0010] If a second reminder task has already been created within the conflict detection time range, then the first reminder task to be created is determined to be in conflict with the second reminder task.

[0011] Based on the determination of the conflict between the pre-created first reminder task and the second reminder task, the target prompt suggestion information is determined according to the event information corresponding to the pre-created first reminder task and the second reminder task.

[0012] A second aspect of this application provides an intelligent reminder task creation device, comprising:

[0013] The information acquisition module is used to extract structured information from user dialogues and determine the first reminder task to be pre-created based on the structured information. The structured information includes first time information and first event information.

[0014] The preset dynamic time threshold determination module is used to determine the corresponding first event type and the first preset dynamic time threshold of the first event type based on the first event information.

[0015] The conflict detection time range determination module is used to determine the conflict detection time range based on the first time information and the first preset dynamic time threshold.

[0016] The reminder task creation module is used to create a pre-created first reminder task if there is no existing second reminder task within the conflict detection time range; if there is an existing second reminder task within the conflict detection time range, the pre-created first reminder task is determined to conflict with the second reminder task.

[0017] The suggestion determination module is used to determine the target suggestion information based on the result of the conflict judgment between the pre-created first reminder task and the second reminder task, and according to the event information corresponding to the pre-created first reminder task and the second reminder task.

[0018] A third aspect of this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the above-described intelligent reminder task creation method.

[0019] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described intelligent reminder task creation method.

[0020] The beneficial effects of the intelligent reminder task creation method, apparatus, electronic device, and storage medium provided in this application embodiment are as follows: This embodiment can accurately extract first time information and first event information from user dialogue to determine the pre-created first reminder task, avoiding errors or information omissions that may occur when the user manually inputs information; and based on the first event information, it determines the corresponding first event type and the first preset dynamic time threshold of the event type. The dynamic time threshold determined in this way can more accurately define the conflict range between tasks, improving the rationality of reminder task creation. This embodiment determines the conflict detection time range based on the first time information and the first preset dynamic time threshold. If a second reminder task that has already been created exists within the conflict detection time range, it will quickly determine that the pre-created first reminder task conflicts with the second reminder task. This timely conflict determination can avoid discovering the conflict only when the event is close to occurring, thus providing sufficient time to adjust the arrangement and reducing the losses and inconvenience caused by the conflict. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 A flowchart illustrating a smart reminder task creation method provided in an embodiment of this application;

[0023] Figure 2 A structural block diagram of an intelligent reminder task creation device provided in an embodiment of this application;

[0024] Figure 3 This is a schematic block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0025] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0026] It is understood that in the embodiments of this application, data such as user information are involved. When the embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of related data must comply with relevant laws, regulations and standards.

[0027] It should be noted that the terms "first," "second," etc., used in the specification, claims, and drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in sequences other than those illustrated or described herein.

[0028] To make the objectives, technical solutions, and advantages of this application clearer, the following description will be provided in conjunction with the accompanying drawings and specific embodiments.

[0029] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a smart reminder task creation method according to an embodiment of this application. The smart reminder task creation method provided in this embodiment can be executed by an electronic device, and the method may include:

[0030] S101: Extract structured information from user dialogue and determine the first reminder task to be pre-created based on the structured information.

[0031] In this embodiment, user dialogue includes dialogue between a user and an electronic device, or group dialogue between multiple users (at least two users). Structured information includes first-time information and first-event information. This embodiment extracts structured information from the user's original voice or text dialogue using an electronic device. For example, to extract structured information from the user's voice input, speech recognition technology is first used to convert the acquired speech into corresponding text content for subsequent analysis. Then, basic preprocessing is performed on the converted text content, which may include word segmentation and removal of unnecessary interjections, aiding in the extraction of key information.

[0032] This embodiment scans and identifies all time-related words in the text content, such as phrases like "tomorrow afternoon at 3 PM" or "five minutes later." Upon successful identification, these vague or relative time phrases are uniformly converted and normalized into precise, absolute timestamps that can be calculated and processed by the device. This absolute timestamp is then used as the first time information. For example, the first time information could be 2025-7-28 15:00:00. Simultaneously, the device identifies core event content from the text content to determine the specific matter the user wants to be reminded of. Descriptive and modifying words are then filtered out, accurately identifying words or phrases expressing the core action and goal. These identified words or phrases are then used as the first event information. For example, the first event information could be a meeting or returning a customer's call.

[0033] In this embodiment, after identifying the first time information and the first event information, it determines whether the first time information and the first event information are semantically related and whether they jointly describe the same task. If the association is successful, the corresponding reminder task title is automatically generated based on the first event information, and the first time information is used as the trigger time for the reminder task. A pre-created first reminder task is generated based on the reminder task title and trigger time.

[0034] S102: Determine the corresponding first event type and the first preset dynamic time threshold of the first event type based on the first event information.

[0035] In this embodiment, the identified first event information is matched with a preset event type library to determine its first event type. For example, if the first event information is "meeting," its first event type is "conference." Based on the determined first event type, the corresponding first preset dynamic time threshold is retrieved from the threshold attribute mapping table. This preset dynamic time threshold is not a fixed value, but rather a lead time amount dynamically calculated based on the characteristics of the task itself (e.g., deadline).

[0036] In this embodiment, determining the corresponding first event type and the first preset dynamic time threshold of the first event type based on the first event information includes:

[0037] Extract keywords from the information of the first event;

[0038] Match the keyword with a preset event type library to determine the corresponding first event type;

[0039] Based on the mapping table between the first event type and the threshold attribute, determine the set of threshold attributes related to the first event type. The threshold attribute mapping table is used to characterize the relationship between the event type and the time occupied by the event type. The set of threshold attributes includes the event duration, time flexibility level, and preparation time requirement. The set of threshold attributes is used to characterize the time occupied by an event type.

[0040] Based on the threshold attribute set and through threshold calculation rules, the first preset dynamic time threshold is determined.

[0041] The preset event type library in this embodiment is pre-built, and it contains a list or database of various general event categories and their keywords. For example, if the event type is "meeting," the corresponding keywords could be "meeting," "discussion," "weekly meeting," "group meeting," and "bumping into a colleague," etc.; if the event type is "travel," the corresponding keywords could be "going to the airport," "taking the train," "departing," "taking a taxi," and "subway," etc.

[0042] This embodiment performs a random match between keywords of the first event information extracted from the user's dialogue and a preset event type database. It iterates through the preset event type database to find keywords that match the words or phrases of the first event information. If a keyword in the preset event type database highly overlaps with the words or phrases of the first event information, the match is considered successful. The first event type is then determined based on the matching result.

[0043] Based on the first event type, a query is performed in the threshold attribute mapping table to obtain the threshold attribute set corresponding to the first event type. The threshold attribute set may include event duration, time flexibility level, and preparation time requirement. The threshold attribute set is used to characterize the time occupied by an event type. For example, if the first event type is a meeting type, the threshold attribute set corresponding to the meeting type may be [duration is medium, flexibility level is rigid, and preparation time requirement is medium].

[0044] In this embodiment, the threshold calculation rule can be: determine the baseline lead time based on the time elasticity level.

[0045] For time flexibility levels, if the flexibility level is rigid, the baseline lead time is 30 minutes; if the flexibility level is flexible, the baseline lead time is 2 days.

[0046] For the duration of the event, if the duration is long (greater than 2 hours), add 30 minutes to the baseline lead time to obtain the corresponding cumulative value; if the duration is medium (greater than 1 hour and less than 2 hours), add 15 minutes to the baseline lead time to obtain the corresponding cumulative value; if the duration is short (less than 1 hour), do not add the baseline lead time, and use the baseline lead time as the cumulative value.

[0047] Regarding the preparation time requirement, if the preparation time requirement is high, 3 hours are added to the cumulative value to generate the corresponding final first preset dynamic time threshold; if the preparation time requirement is medium, 1 hour is added to the cumulative value to generate the corresponding final first preset dynamic time threshold; if the preparation time requirement is low, the cumulative value is used as the final first preset dynamic time threshold.

[0048] The threshold attribute mapping table in this embodiment is pre-built. The mapping table uses the event type as the primary key, and each event type is associated with a set of predefined threshold attributes. The threshold attributes may include the event duration, time flexibility level, and preparation time requirement.

[0049] The event duration is used to estimate how long each event type typically takes. This duration is determined through reasonable planning or analysis based on the actual duration of similar events in historical data. For example, meetings are labeled as medium (lasting 1-2 hours), and phone calls are labeled as short (lasting less than 30 minutes).

[0050] The time flexibility level characterizes the time-related strictness of each event type. This level includes rigid events and flexible events, determined based on the nature of the event type. In this embodiment, a rigid event refers to an event whose start or end time is strictly limited by external factors that the user cannot unilaterally decide. Examples include other people's schedules, fixed departure times, and scheduled interview times. A flexible event refers to an event whose execution time is primarily determined by the user's personal will, preferences, or internal plans, without strict external time constraints, and can be completed within a relatively broad time window. Examples include when to go shopping or when to start reading.

[0051] Preparation time requirements are used to assess how much preparation time users need before an event begins, and are based on the preparation time set by the user for different tasks. For example, giving a report requires "high" preparation time, holding a meeting requires "medium" preparation time, and drinking water requires no preparation time.

[0052] Before determining the set of threshold attributes related to the first event type according to the mapping table between the first event type and the threshold attribute, this embodiment includes:

[0053] Acquire users’ historical data, which includes historical average duration and historical average preparation time.

[0054] Based on the mapping table between the first event type and the threshold attribute, determine the set of threshold attributes related to the first event type, including:

[0055] Determine the initial set of threshold attributes based on the mapping table between the first event type and the threshold attribute;

[0056] The first feature vector is determined based on keywords, the first event type, the first time information, and the historical average duration.

[0057] Based on the first feature vector, and through the duration difference prediction model, the duration difference of the first event type is predicted;

[0058] Based on the frequency of user modification or cancellation of historical event types that are the same as the first event type in historical data, predict the time elasticity coefficient corresponding to the first event type, and determine the time elasticity level of the first event type based on the time elasticity coefficient and the preset elasticity level range.

[0059] The second feature vector is determined based on keywords, the first event type, the first time information, and the historical average preparation time.

[0060] Based on the second feature vector, and through the preparation time difference prediction model, the preparation time difference of the first event type is predicted;

[0061] Based on the duration difference, time elasticity level, and preparation time difference, the initial threshold attribute set is modified to obtain a threshold attribute set related to the first event type.

[0062] This embodiment determines the historical average duration (the time a user typically plans to spend processing a certain type of event) and the historical average preparation time (the time from when a user creates a certain type of event to when they begin executing that event) based on historical data.

[0063] In this embodiment, the initial threshold attribute set can be determined based on the mapping table between the first event type and the threshold attribute.

[0064] In this embodiment, the keywords, type, time information, and historical average duration of the first event information are combined to form a first feature vector. The first feature vector is input into a pre-trained duration difference prediction model. Based on the learned complex patterns, the model intelligently calculates the duration of the current first feature vector and outputs a predicted value, which is used as the duration difference of the first event type.

[0065] In this embodiment, the keywords, type, and time information of the first event are combined with the historical average preparation time to form a second feature vector. This second feature vector is then input into a pre-trained preparation time difference prediction model. This model specifically analyzes user behavior habits to predict how much lead time the user needs to prepare for the task, outputting a predicted value. This predicted value is used as the preparation time difference for the first event type.

[0066] This embodiment predicts the time elasticity coefficient corresponding to the first event type based on the frequency with which users modify or cancel historical event types that are the same as the first event type, and determines the time elasticity level of the first event type based on the time elasticity coefficient and the preset elasticity level range.

[0067] In this embodiment, the initial threshold attribute set is modified based on the duration difference, time elasticity level, and preparation time difference obtained above, and the threshold attribute set for the first event type is determined.

[0068] The threshold attribute set obtained in this embodiment helps to provide a solid data foundation that fits the user's actual habits for the subsequent calculation of the first preset dynamic time threshold, making the final generated reminder task more intelligent and practical.

[0069] For example, if the structured information extracted from the user's conversation includes: the first event information is a team weekly meeting, and the first time information is 9:00 AM on Monday. Based on the first event information, the first event type is determined to be a meeting type. Based on the meeting type and the threshold attribute mapping table, the initial threshold attribute set can be determined to be: [event duration is 1 hour, time flexibility level is rigid, preparation time requirement is 15 minutes].

[0070] The duration difference for the first event type obtained using the above method is +15 minutes, the preparation time difference is -5 minutes, and the time flexibility level is rigid. Therefore, based on the duration difference, preparation time difference, and time flexibility level of the first event type, the corresponding attributes in the initial threshold attribute set are corrected, resulting in a corrected threshold attribute set [event duration 75 minutes, time flexibility level rigid, preparation time requirement 10 minutes]. This corrected threshold attribute set is used as the threshold attribute set related to the first event type.

[0071] The training process for the duration prediction model in this embodiment is as follows:

[0072] For each historical event type, keywords corresponding to that event type are extracted, and a keyword list is formed based on these keywords. These keywords are then converted into fixed-length numerical vectors using word embedding techniques. One-hot encoding is used to convert the historical event type into a binary vector. The historical time information (planned start time) corresponding to that event type is decomposed into multiple time-series features. The historical average duration of the same event type is determined based on the same event types in the historical data. All the obtained numerical vectors (keyword vector, event type vector, time-series features, and historical average duration) are concatenated to form a complete first feature vector representing each event type. The first feature vectors corresponding to each historical event type are determined using the above method. The planned duration corresponding to each first feature vector is determined based on historical data.

[0073] The dataset is composed of each first feature vector and its corresponding planned duration. This dataset is then divided into a training set and a test set according to a preset ratio (e.g., 8:2). The duration prediction model is trained using the training set, and its generalization ability is tested using the test set. This process yields a well-trained duration prediction model.

[0074] The training process for the preparation time prediction model in this embodiment is as follows:

[0075] For each historical event type, keywords corresponding to that event type are extracted, and a keyword list is formed based on these keywords. These keywords are then converted into fixed-length numerical vectors using word embedding techniques. One-hot encoding is then used to convert the historical event type into a binary vector. The historical time information (planned start time) corresponding to this event type is decomposed into multiple time-series features. The historical average preparation time for the same historical event type is determined based on similar event types found in historical data.

[0076] All the numerical vectors obtained above (keyword vector, event type vector, time series features, and historical average preparation time) are concatenated to form a complete second feature vector representing each event type. The second feature vectors corresponding to each historical event type are then determined using the method described above. Finally, the planned preparation time corresponding to each second feature vector is determined based on historical data.

[0077] The dataset is constructed using each second feature vector and its corresponding planned preparation time. This dataset is then divided into a training set and a test set according to a preset ratio (e.g., 8:2). The preparation time prediction model is trained using the training set, and its generalization ability is tested using the test set. This process yields a well-trained preparation time prediction model.

[0078] S103: Determine the conflict detection time range based on the first time information and the first preset dynamic time threshold.

[0079] In this embodiment, the time point obtained by subtracting a first preset dynamic time threshold from the first time information is used as the start time of the conflict detection time range. For example, if the first time information is 2025-08-04 15:00:00, subtracting the lead time of 1 hour and 30 minutes (the first preset dynamic time threshold) yields the start time of the conflict detection time range as 2025-08-04 13:30:00.

[0080] In this embodiment, the time point obtained by adding the first time information to the estimated event duration is used as the end time of the conflict detection time range. For example, if the first time information is 2025-08-04 15:00:00, adding the estimated time of 1 hour (estimated event duration) yields the end time of the conflict detection time range as 2025-08-04 16:00:00.

[0081] The conflict detection time range is determined based on the start and end times of the conflict detection time range. In this embodiment, the conflict detection time range can be [2025-08-04 13:30:00, 2025-08-04 16:00:00].

[0082] S104: If there is no created second reminder task within the conflict detection time range, then create a pre-created first reminder task; if there is a created second reminder task within the conflict detection time range, then determine that the pre-created first reminder task conflicts with the second reminder task.

[0083] In this embodiment, the conflict detection time range is used as a condition to query the constructed reminder task database to determine whether there are any created second reminder tasks whose planned execution time is within or partially within the conflict detection time range.

[0084] If no second reminder task has been created within the conflict detection time range, it is determined that the pre-created first reminder task does not conflict with the user's existing schedule. The status of the pre-created first reminder task is changed from pre-created to activated, and it is officially stored in the database. The first reminder task is successfully created.

[0085] If at least one second reminder task exists within the conflict detection timeframe, the pre-created first reminder task is determined to conflict with the second reminder task. In this case, the first reminder task will not be automatically created; instead, a conflict resolution mechanism will be triggered. In this embodiment, the conflict resolution mechanism may generate and send a conflict notification message to the user. This message will inform the user of the conflict and may provide options such as still creating, canceling, or rescheduling the task.

[0086] S105: Based on the determination result of the conflict between the pre-created first reminder task and the second reminder task, determine the target prompt suggestion information according to the event information corresponding to the pre-created first reminder task and the second reminder task respectively.

[0087] This embodiment determines the target prompt suggestion information based on the event information corresponding to the pre-created first reminder task and the second reminder task, including:

[0088] Based on the event information corresponding to the pre-created first and second reminder tasks, determine whether the pre-created first and second reminder tasks can be executed in parallel.

[0089] If it is determined that the pre-created first reminder task and the second reminder task cannot be executed in parallel, the preset end time of the second reminder task is determined based on the second time information of the second reminder task, and a first prompt suggestion information is generated according to the preset end time. The first prompt suggestion information is used to prompt the user to modify the first time of the pre-created first reminder task to a first target time, which is later than the preset end time of the second reminder task.

[0090] If it is determined that the pre-created first reminder task and the second reminder task can be executed in parallel, then based on the first event type and the second event type, and through the parallel execution impact library, the impact level of parallel execution is determined, and a second prompt suggestion information is generated according to the impact level. The second prompt suggestion information is used to prompt the user to confirm that the pre-created first reminder task and the second reminder task are executed in parallel, or to adjust the time of the pre-created first reminder task. The parallel execution impact library is used to characterize the impact level and impact description between the two event types.

[0091] Determine the target prompt or suggestion information based on the first or second prompt or suggestion information.

[0092] In this embodiment, based on the first event information (including the first event type, keywords, etc.) corresponding to the pre-created first reminder task and the second event information (including the second event type, keywords, etc.) corresponding to the already created second reminder task, it is determined whether the pre-created first reminder task and the second reminder task can be executed in parallel. For example, important meetings that require the user's full attention and fitness exercises that require the user's manual operation are usually determined not to be executed in parallel; similarly, data backup that can run in the background and document writing that requires the user's focus may also be determined not to be executed in parallel. The user's walk and listening to music are determined to be executed in parallel, and device charging that does not require the user's active participation and most other tasks can also be determined to be executed in parallel.

[0093] If it is determined that the pre-created first reminder task and the already created second reminder task cannot be executed in parallel, a sequential execution suggestion process is initiated. Based on the second time information of the second reminder task (i.e., the planned start time of the second reminder task) and its predicted event duration, the preset end time of the second reminder task is calculated. Using this preset end time as a benchmark, a first target time is calculated (for example, a buffer time is reserved after the preset end time, which can be 5 or 10 minutes). In this embodiment, a first prompt suggestion message is generated based on the first target time. This first prompt suggestion message clearly informs the user of the conflict situation and solution in natural language. For example, if it is detected that the pre-created first reminder task conflicts with the team weekly meeting and cannot be executed in parallel, the user can be prompted: the meeting is expected to end at 16:00, and it is recommended to change the current task to start at 16:10.

[0094] If it is determined that the pre-created first reminder task and the already created second reminder task can be executed in parallel, the parallel execution suggestion process is initiated. The first event type and the second event type are used as composite query conditions to access the parallel execution impact database. This database is a predefined knowledge base that stores the mutual influence relationships between different event type combinations, and labels each pair of combinations with an impact level (e.g., high interference, medium impact, low interference / parallelizable) and a corresponding impact description. Second reminder suggestion information is generated based on the impact level.

[0095] If the impact level is high interference, the generated second suggestion message might be: "Although phone communication and coding can be done simultaneously, doing so may significantly reduce efficiency. We suggest you reschedule your phone communication." If the impact level is moderate or low interference, the generated second suggestion message might be: "Listening to music and reading can be done in parallel, but please be mindful of the volume to avoid distraction. Please confirm whether you wish to do them simultaneously?" In this embodiment, the second suggestion message can provide the user with options such as "Confirm Parallelism" and "Modify Time".

[0096] This embodiment determines the target prompt or suggestion information to be presented to the user based on the generated first or second prompt or suggestion information.

[0097] Specifically, based on the first event type and the second event type, and through the parallel execution impact library, the impact level of parallel execution is determined, and second suggestion information is generated according to the impact level, including:

[0098] Combine the first event type with the second event type to generate a composite primary key;

[0099] The query is performed in parallel based on the composite primary key to obtain the query results, which include the influence level corresponding to the composite primary key.

[0100] The content of the suggested tips should be determined based on the level of impact.

[0101] The suggested content is encapsulated to obtain a second suggested information.

[0102] This embodiment combines the first event type and the second event type to generate a composite primary key. This composite primary key can follow a fixed rule, for example, concatenating the event types in alphabetical order. Assuming the first event type is telephone communication and the second event type is document writing, the generated composite primary key would be "Telephone Communication - Document Writing".

[0103] Using the generated composite primary key as the query condition, perform a retrieval operation on the parallel execution impact database to obtain the query results, and extract the impact level that uniquely corresponds to the composite primary key from the query results.

[0104] This embodiment sets up a prompt content mapping table, which is used to bind different impact levels with pre-written suggestion templates.

[0105] If the impact level corresponding to the unique composite primary key is high interference, the corresponding suggestion template could be: "Warning: Executing [Event Type A] and [Event Type B] in parallel will severely reduce efficiency and user experience. We recommend that you reschedule the current task." If the impact level corresponding to the unique composite primary key is moderate interference, the corresponding suggestion template could be: "Please note that performing [Event Type A] and [Event Type B] simultaneously may interfere with each other. We recommend that you prioritize handling one of them, or confirm that they are executed in parallel." If the impact level corresponding to the unique composite primary key is low interference / can be parallelized, the corresponding suggestion template could be: "[Event Type A] and [Event Type B] can be executed in parallel. Are you sure you want to do it simultaneously?".

[0106] Based on the impact level between the first event type and the second event type, determine the corresponding suggestion template, and fill the placeholders in the selected suggestion template with the first event type and the second event type (e.g., [Event Type A] and [Event Type B]) to form the final, specific suggestion content.

[0107] In this embodiment, the obtained suggested prompts, their corresponding impact levels, and user-selectable action buttons (e.g., confirm parallel processing, modify time, or cancel creation) are encapsulated into a standard data structure (e.g., a JSON object). This encapsulated data structure serves as the final second suggestion message. This message is sent to the system's user interaction module, ready to be displayed or broadcast to the user.

[0108] As can be seen from the above, this embodiment determines the corresponding first event type and the first preset dynamic time threshold of the event type based on the first event information, fully considering the characteristics of different types of events in terms of time arrangement, and improving the accuracy of task creation. This embodiment determines the conflict detection time range based on the first time information and the first preset dynamic time threshold, and checks whether there is a second reminder task already created within this range. This helps to discover potential task time overlap problems in advance, avoids users being assigned multiple conflicting tasks in the same time period, thereby saving users' time and energy, preventing the chaos of plans caused by task conflicts, and improving the efficiency and success rate of task execution. Furthermore, when a conflict is detected, this embodiment does not simply prompt the conflict, but determines whether they can be executed in parallel based on the event information of the pre-created first reminder task and the created task, and generates corresponding prompt suggestions based on the judgment result. This helps to provide users with practical solutions, help users efficiently handle task conflicts, and save users' time for thinking and adjusting on their own.

[0109] In one embodiment of this application, the method further includes:

[0110] If the first-time information is extracted from the user's dialogue but the first event information is not extracted, then the first event completion operation is performed to obtain the first event completion information;

[0111] If the first event information is extracted from the user's dialogue but the first time information is not extracted, then the first time completion operation is performed to obtain the first time completion information;

[0112] The first reminder task to be created is determined based on the information completed in the first event or the information completed in the first moment.

[0113] The process of performing the first event completion operation yields the first event completion information, including:

[0114] Based on the context of the user's conversation and the first-time historical information, query the user's first historical behavior data;

[0115] Based on the first historical behavior data, generate at least one candidate event item;

[0116] Based on at least one candidate event item, and according to the user's selection, the first event completion information is determined;

[0117] The process of performing immediate completion operation yields immediate completion information, including:

[0118] Based on the context of the user's conversation and the information of the first historical event, query the user's second historical behavior data;

[0119] Based on the second historical behavior data, generate at least one candidate time event;

[0120] Based on at least one candidate time item, and according to the user's selection, the information is completed as soon as possible.

[0121] In this embodiment, after extracting the user's structured information, the integrity of the extracted structured information is verified, and different completion processes are triggered according to the type of missing information.

[0122] If the first-time information is successfully extracted from the user's conversation, but no specific first event information is extracted (for example, the user only says: "Remind me at 3 pm tomorrow"), then the possible event is inferred based on the context of the user's conversation and the user's habits.

[0123] This embodiment can combine the context of the current user conversation (e.g., conversation content from the previous few minutes) and extracted first-time information as query conditions to retrieve the user's first historical behavior data. This first historical behavior data includes tasks frequently performed by the user at similar times (e.g., both "Wednesday afternoons"), recently unfinished to-do items, and related tasks mentioned in the conversation context. Based on the first historical behavior data, and sorted according to relevance, time proximity, and execution frequency, a list containing at least one candidate event is generated. For example, the candidate list for "Remind me at 3 PM tomorrow" might include: [Attend a project review meeting, talk to a fitness coach, submit a weekly report]. This embodiment can present the candidate list to the user through a graphical interface or voice interaction for the user to choose from. Based on the user's selection instruction (e.g., the user clicks or selects the first option via voice), this embodiment sets the user's final selected option as the first event completion information.

[0124] In this embodiment, the first time information and the first event completion information are combined to determine the first reminder task to be created.

[0125] If the first event information is successfully extracted from the user's conversation, but the specific first time information is not extracted (for example, the user only says: remind me to call Xiao Wang back), then an appropriate time is recommended based on the context of the user's conversation and the user's habits.

[0126] This embodiment can combine the context of the current user dialogue and the extracted first event information as query conditions to query the user's second historical behavior data. This second historical behavior data includes the user's habitual timing for performing this type of event (e.g., exercising in the evening), the user's current schedule availability, and the default or urgency level of the event.

[0127] This embodiment determines a list containing at least one candidate time item based on second historical behavior data and combined with schedule gaps. These time points are typically specific, absolute timestamps. For example, the candidate list for "Remind me to call Xiao Wang back" might be: [4:00 PM today, 10:30 AM tomorrow, 2:00 PM tomorrow]. This embodiment can present this candidate list to the user through a graphical interface or voice interaction for the user to select. Based on the user's selection command, this embodiment sets the user's final selected option as the first option to complete the information.

[0128] In this embodiment, the first event information and the first time completion information are combined to determine the pre-created first reminder task.

[0129] This embodiment can effectively handle user commands with incomplete information. By actively inferring and interactively confirming, it transforms vague user intentions into precise and executable reminder tasks, significantly improving the robustness of the system and the user experience.

[0130] In one embodiment of this application, the method further includes:

[0131] The first event information is segmented according to the preset analysis rules to obtain the keywords of the first event information;

[0132] Each keyword is matched with a preset sensitive word library. If at least one keyword in each keyword matches the preset sensitive word library, the risk classification of the first event information is performed based on the first event information and its context, and through a natural language understanding model, to obtain the risk level corresponding to the first event information.

[0133] Determine the corresponding response strategy based on the risk level.

[0134] This embodiment can perform preliminary, keyword-based rapid matching and screening of event information. Based on preset analysis rules (e.g., word segmentation algorithms based on dictionaries or statistical models), the first event information extracted from the user's dialogue can be segmented into independent lexical units to obtain multiple keywords, thereby generating a keyword list.

[0135] All the keywords obtained from the segmentation are compared one by one with a preset sensitive word library, which contains a set of words related to security, privacy, fraud, illegal or inappropriate content.

[0136] If none of the keywords match the sensitive word database, the first event information is deemed risk-free, and the subsequent task creation process continues.

[0137] If at least one keyword in the keyword list matches the sensitive word database (for example, the keyword "transfer" is a sensitive word), then the first event information is determined to be risky, and a deeper risk classification is performed on the first event information to determine the risk level corresponding to the first event information.

[0138] This embodiment outputs the risk level corresponding to the first event information based on the first event information and its context, using a pre-trained natural language understanding model. This natural language understanding model is used for risk intent identification and can understand the true semantics and context of language.

[0139] In this embodiment, the risk levels corresponding to the first event information include high risk, medium risk, low risk, and no risk. High-risk information explicitly involves illegal or irregular activities, fraud, and serious personal attacks; medium-risk information involves sensitive privacy, potential financial risks, or mildly inappropriate remarks; low-risk information contains only sensitive words but the context is harmless; and no-risk information is determined by the model to be a false alarm.

[0140] If the risk level corresponding to the first event information is high risk, the creation of the alert task is rejected, and a general security alert message is generated and sent to the user. If the risk level corresponding to the first event information is medium risk, the creation process is paused, and a verification message is sent to the user. For example, "The alert you created involves financial transactions. Please confirm whether you initiated it yourself?" [Confirm] / [Cancel]. This embodiment determines whether to continue creation based on the user's subsequent confirmation. If the risk level corresponding to the first event information is low risk or no risk, the first event information is determined to be safe or a false alarm, and the normal creation process of the first alert task is allowed and continues.

[0141] To further accurately complete the missing parts of the structured information, this embodiment can also implement the following steps when determining at least one candidate event:

[0142] Based on the user's first historical behavior data, obtain the context of the conversation with the user and / or one or more historical event types associated with the first time information, and determine the first candidate event set based on the frequency of each historical event type;

[0143] Based on a preset event relationship mapping table, one or more related events that have a preset logical relationship with the events represented by the context of the user's dialogue and / or the first-time information are identified, and a second candidate event set is generated; wherein, the event relationship mapping table stores the degree of association between different event types;

[0144] The first candidate event set and the second candidate event set are merged, and the comprehensive score of each candidate event in the merged set is calculated according to the preset weighted scoring rules.

[0145] Sort the comprehensive scores in descending order and select the top 10 candidate events as candidate event items.

[0146] This embodiment acquires the user's historical notification events and archives them according to event type, time, and keywords to obtain a historical behavior database. This database is used to query high-frequency historical events. The user's primary historical behavior data is determined based on the context of the user's dialogue and initial historical information.

[0147] If a user's dialogue is "Help me arrange things for next Monday afternoon," parsing this text reveals the first time information as Monday afternoon, but the first event information is missing. In this case, using "Monday afternoon" as the key, the historical behavior database is queried to find all event types that occurred on historical Monday afternoons. The frequency of each historical event type is calculated, and the three most frequent events are selected to form the first candidate event set. For example, the first candidate event set could be [team weekly meeting, project debriefing, client visit].

[0148] This embodiment uses Monday afternoon as a general scenario, queries the event relationship mapping table for the general event with the highest relevance to this scenario, filters out events with a relevance higher than a preset relevance threshold (which can be 0.5) from the queried general events, and generates a second candidate event set based on the filtered events. For example, the second candidate event set could be [planning work schedule, processing emails].

[0149] This embodiment merges the first candidate event set and the second candidate event set to obtain a new candidate event set, which can be [team weekly meeting, project review, client visit, work plan planning, email processing].

[0150] In this embodiment, the frequency of each event in the new candidate event set is converted into a standardized frequency value of 0-1, and the correlation degree corresponding to each event is determined according to the preset event relationship mapping table.

[0151] Based on the standardized frequency values ​​of each event, the correlation between each event, and the comprehensive score calculation formula, a comprehensive score is calculated for each candidate event. The comprehensive score calculation formula can be: ,in, S The overall score for each candidate event, p The normalized frequency value for each candidate event. q The relevance of each candidate event, The weights corresponding to the standardized frequency values, The weights corresponding to the degree of relevance. In this embodiment and The values ​​of each weight can be adjusted according to the actual scenario.

[0152] The comprehensive scores of each candidate event are sorted in descending order, and the top 3 candidate events are selected as candidate event items for users to choose from.

[0153] Based on the same inventive concept, this application also provides an intelligent reminder task creation device for implementing the intelligent reminder task creation method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more intelligent reminder task creation device embodiments provided below can be found in the limitations of the intelligent reminder task creation method described above, and will not be repeated here.

[0154] This application provides an intelligent reminder task creation device, such as... Figure 2 As shown, the intelligent reminder task creation device 20 includes: an information acquisition module 21, a preset dynamic time threshold determination module 22, a conflict detection time range determination module 23, a reminder task creation module 24, and a prompt suggestion determination module 25.

[0155] In one embodiment of this application, the information acquisition module 21 is used to extract structured information from user dialogue and determine a pre-created first reminder task based on the structured information. The structured information includes first time information and first event information.

[0156] The preset dynamic time threshold determination module 22 is used to determine the corresponding first event type and the first preset dynamic time threshold of the first event type based on the first event information.

[0157] The conflict detection time range determination module 23 is used to determine the conflict detection time range based on the first time information and the first preset dynamic time threshold.

[0158] The reminder task creation module 24 is used to create a pre-created first reminder task if there is no existing second reminder task within the conflict detection time range; if there is an existing second reminder task within the conflict detection time range, it is determined that the pre-created first reminder task conflicts with the second reminder task.

[0159] The suggestion determination module 25 is used to determine the target suggestion information based on the result of the conflict between the pre-created first reminder task and the second reminder task, and according to the event information corresponding to the pre-created first reminder task and the second reminder task.

[0160] In one embodiment of this application, when determining the corresponding first event type and the first preset dynamic time threshold of the first event type based on the first event information, the preset dynamic time threshold determining module 22 is specifically used for:

[0161] Extract keywords from the information of the first event;

[0162] Match the keywords with the preset event type library to determine the corresponding first event type;

[0163] Based on the mapping table between the first event type and the threshold attribute, determine the set of threshold attributes related to the first event type. The threshold attribute mapping table is used to characterize the relationship between the event type and the time occupied by the event type. The set of threshold attributes includes the event duration, time flexibility level, and preparation time requirement. The set of threshold attributes is used to characterize the time occupied by an event type.

[0164] Based on the threshold attribute set and through threshold calculation rules, the first preset dynamic time threshold is determined.

[0165] In one embodiment of this application, before determining the set of threshold attributes related to the first event type according to the mapping table between the first event type and the threshold attribute, a preset dynamic time threshold determination module 22 is specifically used for:

[0166] Acquire users’ historical data, which includes historical average duration and historical average preparation time.

[0167] When determining the set of threshold attributes related to the first event type based on the mapping table between the first event type and the threshold attribute, the preset dynamic time threshold determination module 22 is specifically used for:

[0168] Determine the initial set of threshold attributes based on the mapping table between the first event type and the threshold attribute;

[0169] The first feature vector is determined based on keywords, the first event type, the first time information, and the historical average duration.

[0170] Based on the first feature vector, and through the duration difference prediction model, the duration difference of the first event type is predicted;

[0171] Based on the frequency of user modification or cancellation of historical event types that are the same as the first event type in historical data, predict the time elasticity coefficient corresponding to the first event type, and determine the time elasticity level of the first event type based on the time elasticity coefficient and the preset elasticity level range.

[0172] The second feature vector is determined based on keywords, the first event type, the first time information, and the historical average preparation time.

[0173] Based on the second feature vector, and through the preparation time difference prediction model, the preparation time difference of the first event type is predicted;

[0174] Based on the duration difference, time elasticity level, and preparation time difference, the initial threshold attribute set is modified to obtain a threshold attribute set related to the first event type.

[0175] In one embodiment of this application, when determining the target prompt suggestion information based on the event information corresponding to the pre-created first reminder task and the second reminder task, the prompt suggestion determination module 25 is specifically used for:

[0176] Based on the event information corresponding to the pre-created first and second reminder tasks, determine whether the pre-created first and second reminder tasks can be executed in parallel.

[0177] If it is determined that the pre-created first reminder task and the second reminder task cannot be executed in parallel, the preset end time of the second reminder task is determined based on the second time information of the second reminder task, and a first prompt suggestion information is generated according to the preset end time. The first prompt suggestion information is used to prompt the user to modify the first time of the pre-created first reminder task to a first target time, which is later than the preset end time of the second reminder task.

[0178] If it is determined that the pre-created first reminder task and the second reminder task can be executed in parallel, then based on the first event type and the second event type, and through the parallel execution impact library, the impact level of parallel execution is determined, and a second prompt suggestion information is generated according to the impact level. The second prompt suggestion information is used to prompt the user to confirm that the pre-created first reminder task and the second reminder task are executed in parallel, or to adjust the time of the pre-created first reminder task. The parallel execution impact library is used to characterize the impact level and impact description between the two event types.

[0179] Determine the target prompt or suggestion information based on the first or second prompt or suggestion information.

[0180] In one embodiment of this application, when determining the impact level of parallel execution based on a first event type and a second event type, and through a parallel execution impact library, and generating second suggestion information based on the impact level, the suggestion determination module 25 is specifically used for:

[0181] Combine the first event type with the second event type to generate a composite primary key;

[0182] The query is performed in parallel based on the composite primary key to obtain the query results, which include the influence level corresponding to the composite primary key.

[0183] The content of the suggested tips should be determined based on the level of impact.

[0184] The suggested content is encapsulated to obtain a second suggested information.

[0185] In one embodiment of this application, the device further includes an information completion module, which is used for:

[0186] If the first-time information is extracted from the user's dialogue but the first event information is not extracted, then the first event completion operation is performed to obtain the first event completion information;

[0187] If the first event information is extracted from the user's dialogue but the first time information is not extracted, then the first time completion operation is performed to obtain the first time completion information;

[0188] The first reminder task to be created is determined based on the information completed in the first event or the information completed in the first moment.

[0189] The process of performing the first event completion operation yields the first event completion information, including:

[0190] Based on the context of the user's conversation and the first-time historical information, query the user's first historical behavior data;

[0191] Based on the first historical behavior data, generate at least one candidate event item;

[0192] Based on at least one candidate event item, and according to the user's selection, the first event completion information is determined;

[0193] The process of performing immediate completion operation yields immediate completion information, including:

[0194] Based on the context of the user's conversation and the information of the first historical event, query the user's second historical behavior data;

[0195] Based on the second historical behavior data, generate at least one candidate time event;

[0196] Based on at least one candidate time item, and according to the user's selection, the information is completed as soon as possible.

[0197] In one embodiment of this application, the device further includes a risk level determination module, which is used to:

[0198] The first event information is segmented according to the preset analysis rules to obtain the keywords of the first event information;

[0199] Each keyword is matched with a preset sensitive word library. If at least one keyword in each keyword matches the preset sensitive word library, the risk classification of the first event information is performed based on the first event information and its context, and through a natural language understanding model, to obtain the risk level corresponding to the first event information.

[0200] Determine the corresponding response strategy based on the risk level.

[0201] See Figure 3 , Figure 3This is a schematic block diagram of an electronic device provided according to an embodiment of this application. Figure 3 The electronic device 300 in this embodiment may include one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memories 304 store computer programs, including program instructions. The processors 301 execute the program instructions stored in the memories 304. Specifically, the processors 301 are configured to invoke the program instructions to perform the functions of each module / unit in the above-described device embodiments, for example... Figure 2 The functions shown are: information acquisition module 21, preset dynamic time threshold determination module 22, conflict detection time range determination module 23, reminder task creation module 24, and suggestion determination module 25.

[0202] It should be understood that, in the embodiments of this application, the processor 301 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0203] Input device 302 may include a touchpad, a fingerprint sensor (for collecting the user's fingerprint information and fingerprint orientation information), a microphone, etc., and output device 303 may include a display (LCD, etc.), a speaker, etc.

[0204] The memory 304 may include read-only memory and random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include non-volatile random access memory. For example, the memory 304 may also store information such as first event information, first time information, first preset dynamic time threshold, and target prompt suggestions.

[0205] In specific implementations, the processor 301, input device 302, and output device 303 described in the embodiments of this application can execute the implementation method described in the intelligent reminder task creation method provided in the embodiments of this application, or they can execute the implementation method of the electronic device described in the embodiments of this application, which will not be repeated here.

[0206] In another embodiment of this application, a computer-readable storage medium is provided. This computer-readable storage medium stores a computer program, which includes program instructions. When executed by a processor, the program instructions implement all or part of the processes in the methods described above. Alternatively, the computer program can instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0207] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the foregoing embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., provided on the electronic device. Furthermore, the computer-readable storage medium can include both internal and external storage units of the electronic device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0208] Those skilled in the art will recognize that the modules / units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0209] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the electronic devices and units described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0210] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of modules / units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules, units, or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces or modules / units, or it may be an electrical, mechanical, or other form of connection.

[0211] The modules / units described as separate components may or may not be physically separate. Similarly, the components shown as modules / units may or may not be physical modules / units; they may be located in one place or distributed across multiple network modules / units. Some or all of the modules / units can be selected to achieve the purpose of the embodiments of this application, depending on actual needs.

[0212] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing unit, or each module can exist physically separately, or two or more modules can be integrated into one unit. The integrated modules / units described above can be implemented in hardware or in the form of software functional modules / units.

[0213] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for creating intelligent reminder tasks, characterized in that, include: Structured information is extracted from user conversations, and a pre-created first reminder task is determined based on the structured information, wherein the structured information includes first time information and first event information; Based on the first event information, determine the corresponding first event type and the first preset dynamic time threshold of the first event type; The conflict detection time range is determined based on the first time information and the first preset dynamic time threshold. If there is no created second reminder task within the conflict detection time range, then the pre-created first reminder task is created; If a second reminder task has already been created within the conflict detection time range, then the pre-created first reminder task is determined to conflict with the second reminder task; Based on the determination result of the conflict between the pre-created first reminder task and the second reminder task, and according to the event information corresponding to the pre-created first reminder task and the second reminder task respectively, it is determined whether the pre-created first reminder task and the second reminder task can be executed in parallel; If it is determined that the pre-created first reminder task and the second reminder task cannot be executed in parallel, then the preset end time of the second reminder task is determined based on the second time information of the second reminder task, and a first prompt suggestion information is generated according to the preset end time. The first prompt suggestion information is used to prompt the user to modify the first time of the pre-created first reminder task to a first target time, which is later than the preset end time of the second reminder task. If it is determined that the pre-created first reminder task and the second reminder task can be executed in parallel, then based on the first event type and the second event type, and through the parallel execution impact library, the impact level of parallel execution is determined, and a second prompt suggestion information is generated according to the impact level. The second prompt suggestion information is used to prompt the user to confirm that the pre-created first reminder task and the second reminder task are executed in parallel, or to adjust the time of the pre-created first reminder task. The parallel execution impact library is used to characterize the impact level and impact description between the two event types. The target prompt or suggestion information is determined based on the first or second prompt or suggestion information.

2. The intelligent reminder task creation method as described in claim 1, characterized in that, Determining the corresponding first event type and the first preset dynamic time threshold for the first event type based on the first event information includes: Extract keywords from the first event information; The keywords are matched with a preset event type library to determine the corresponding first event type; Based on the first event type and threshold attribute mapping table, a set of threshold attributes related to the first event type is determined. The threshold attribute mapping table is used to characterize the relationship between the event type and the time occupied by the event type. The set of threshold attributes includes event duration, time flexibility level, and preparation time requirement. The set of threshold attributes is used to characterize the time occupied by an event type. Based on the set of threshold attributes and through threshold calculation rules, a first preset dynamic time threshold is determined.

3. The intelligent reminder task creation method as described in claim 2, characterized in that, Before determining the set of threshold attributes related to the first event type based on the first event type and threshold attribute mapping table, the process includes: Obtain the user's historical data, which includes the historical average duration and the historical average preparation time; The step of determining the set of threshold attributes related to the first event type based on the mapping table between the first event type and the threshold attribute includes: Determine the initial set of threshold attributes based on the first event type and threshold attribute mapping table; A first feature vector is determined based on the keywords, the first event type, the first time information, and the historical average duration. Based on the first feature vector, and through the duration difference prediction model, the duration difference of the first event type is predicted; Based on the frequency of user modification or cancellation of historical event types that are the same as the first event type in the historical data, the time elasticity coefficient corresponding to the first event type is predicted, and the time elasticity level of the first event type is determined based on the time elasticity coefficient and the preset elasticity level range. The second feature vector is determined based on the keywords, the first event type, the first time information, and the historical average preparation time. Based on the second feature vector, and through the preparation time difference prediction model, the preparation time difference of the first event type is predicted; Based on the duration difference, time elasticity level, and preparation time difference, the initial threshold attribute set is modified to obtain a threshold attribute set related to the first event type.

4. The intelligent reminder task creation method as described in claim 1, characterized in that, The process involves determining the impact level of parallel execution based on the first event type and the second event type, and using a parallel execution impact library. A second suggestion message is then generated based on the impact level, including: Combine the first event type with the second event type to generate a composite primary key; Based on the composite primary key, the parallel execution impact database is queried to obtain the query results, which include the impact level corresponding to the composite primary key; The suggested content should be determined based on the level of impact. The suggested content is encapsulated to obtain the second suggested information.

5. The intelligent reminder task creation method as described in claim 1, characterized in that, The method further includes: If the first-time information is extracted from the user's dialogue but the first event information is not extracted, then the first event completion operation is performed to obtain the first event completion information; If the first event information is extracted from the user's dialogue but the first time information is not extracted, then the first time completion operation is performed to obtain the first time completion information; The pre-created first reminder task is determined based on the first event completion information or the first time completion information; The step of performing the first event completion operation to obtain the first event completion information includes: Based on the context of the user's dialogue and the first-time historical information, query the user's first historical behavior data; Based on the first historical behavior data, at least one candidate event item is generated; Based on the at least one candidate event item, and according to the user's selection, the first event completion information is determined; The step of performing the first-time information completion to obtain the first-time information completion includes: Based on the context of the user's dialogue and the first historical event information, query the user's second historical behavior data; Based on the second historical behavior data, generate at least one candidate time event; Based on the at least one candidate time item, and according to the user's selection, the first time completion information is determined.

6. The intelligent reminder task creation method as described in claim 1, characterized in that, The method further includes: The first event information is segmented according to preset analysis rules to obtain the keywords of the first event information; Each keyword is matched with a preset sensitive word library. If at least one keyword among the keywords matches the preset sensitive word library, the first event information is classified into risk levels based on the first event information and its context, and through a natural language understanding model, to obtain the risk level corresponding to the first event information. The corresponding response strategy is determined based on the risk level.

7. A smart reminder task creation device, characterized in that, include: The information acquisition module is used to extract structured information from user dialogue and determine a pre-created first reminder task based on the structured information. The structured information includes first time information and first event information. A preset dynamic time threshold determination module is used to determine the corresponding first event type and the first preset dynamic time threshold of the first event type based on the first event information. The conflict detection time range determination module is used to determine the conflict detection time range based on the first time information and the first preset dynamic time threshold. The reminder task creation module is used to create the pre-created first reminder task based on the determination result that there is no second reminder task already created within the conflict detection time range; Based on the determination result that there is a second reminder task already created within the conflict detection time range, it is determined that the pre-created first reminder task conflicts with the second reminder task; The suggestion determination module is used to determine whether the pre-created first reminder task and the second reminder task can be executed in parallel based on the determination result of the conflict between the pre-created first reminder task and the second reminder task, and according to the event information corresponding to the pre-created first reminder task and the second reminder task respectively. If it is determined that the pre-created first reminder task and the second reminder task cannot be executed in parallel, then the preset end time of the second reminder task is determined based on the second time information of the second reminder task, and a first prompt suggestion information is generated according to the preset end time. The first prompt suggestion information is used to prompt the user to modify the first time of the pre-created first reminder task to a first target time, which is later than the preset end time of the second reminder task. If it is determined that the pre-created first reminder task and the second reminder task can be executed in parallel, then based on the first event type and the second event type, and through the parallel execution impact library, the impact level of parallel execution is determined, and a second prompt suggestion information is generated according to the impact level. The second prompt suggestion information is used to prompt the user to confirm that the pre-created first reminder task and the second reminder task are executed in parallel, or to adjust the time of the pre-created first reminder task. The parallel execution impact library is used to characterize the impact level and impact description between the two event types. The target prompt or suggestion information is determined based on either the first or the second prompt or suggestion information.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 6.

Images