Schedule generation method and device, equipment and storage medium

Abstract

The invention discloses a schedule generation method and device, equipment and a storage medium. The schedule generation method comprises the following steps: a target vehicle performs text conversion on target voice data to obtain target text data, wherein the target voice data comprises near-end voice data acquired by audio acquisition equipment in the target vehicle and / or far-end voice data received from other equipment; performing information extraction on the target text data to obtain schedule text information related to a schedule in the target text data; processing the schedule text information based on a preset event identifier and a preset time identifier to obtain schedule event information and schedule time information; and combining the schedule event information and the schedule time information to obtain a target schedule. According to the scheme, the accuracy of the generated target schedule can be improved.

Description

Technical Field

[0001] The present application relates to the field of automobile technology, and in particular to a schedule generation method, device, equipment and storage medium. Background Art

[0002] Some schedule applications in car applications do not support voice input, and require the driver or passenger to manually and multi-step operations to complete the establishment of a single schedule. In actual operation, this often requires taking your eyes off the road, which poses a safety hazard. In addition, the need for manual operation makes the schedule generation process cumbersome. In some application scenarios, although some schedule applications support voice input to improve the cumbersome process of manually establishing schedules, they all require voice input to have a certain paradigm. In actual operation, the voice input often deviates from the requirements, resulting in the inability to generate a schedule or the generated schedule is inaccurate.

[0003] In view of the existing technical defects, how to provide a more accurate solution for schedule generation is a technical problem that needs to be solved urgently by those skilled in the art. Summary of the invention

[0004] The present application at least provides a schedule generation method, apparatus, device and storage medium.

[0005] The present application provides a schedule generation method, comprising: a target vehicle performs text conversion on target voice data to obtain target text data, the target voice data including near-end voice data collected by an audio collection device in the target vehicle and / or far-end voice data received from other devices; extracting information from the target text data to obtain schedule text information related to the schedule in the target text data; processing the schedule text information based on a preset event identifier and a time identifier to obtain schedule event information and schedule time information; and combining the schedule event information and the schedule time information to obtain a target schedule.

[0006] In some embodiments, schedule text information is processed based on pre-set event identifiers and time identifiers to obtain schedule event information and schedule time information, including: splitting the schedule text information using the positions of the event identifiers and time identifiers to obtain initial event information and initial time information; determining target time information and target event information based on the relationship between the initial time information and preset time parameters; determining schedule time information based on the target time information; and using the target event information as schedule event information.

[0007] In some embodiments, target time information and target event information are determined based on the relationship between the initial time information and the preset time parameters, including: in response to the initial time information including the preset time parameters, time information adjacent to the preset time parameters in the initial time information are determined as candidate start time and candidate end time in the candidate time information, respectively; or, in response to the initial time information not including the preset time parameters, the initial time information is determined as the candidate start time in the candidate time information; based on the schedule time information corresponding to the created target schedule, the candidate end time in the candidate time information is determined; based on the value of the time range of the candidate time information and the initial event information, the target time information and target event information are determined.

[0008] In some embodiments, the target time information includes a target start time and a target end time, and the target time information and the target event information are determined based on the value of the time range of the candidate time information and the initial event information, including: in response to the value of the candidate start time in the candidate time information being empty, deleting the candidate time information and the initial event information; or, in response to the value of the candidate start time in the target time information being not empty, determining whether the candidate start time is a moment before the candidate end time; in response to the candidate start time being a moment before the candidate end time, using the initial event information and the candidate time information as the target event information and target time information, respectively.

[0009] In some embodiments, there are multiple target time information, and based on the target time information, the schedule time information is determined, including: for each target time information, the following steps are performed: determine whether there is time overlap between the time range of the target time information and the time range of the reference time information, the reference time information includes other target time information other than the target time information and / or schedule time information corresponding to the created target schedule; in response to the time range of the target time information and the time range of the reference time information not overlapping, determine the target time information as the schedule time information; or, in response to the time range of the target time information and the time range of the reference time information overlapping, determine the overlapping time information corresponding to the target time information, the overlapping time information being at least one of the reference time information; determine the schedule time information based on the target time information and the overlapping time information.

[0010] In some embodiments, based on the target time information and the overlapping time information, the schedule time information is determined, including: merging the time range of the target time information and the time range of the overlapping time information to obtain a merged time range, and using the merged time range as the time range of the schedule time information; or, selecting at least a portion of the time range from the optional time range as the time range of the schedule time information, the optional time range being other time ranges except the time range of the reference time information.

[0011] In some embodiments, information extraction is performed on target text data to obtain schedule text information related to the schedule in the target text data, including: determining a target prompt word of the target text data based on a preset prompt word template and text content of at least one preset target schedule, the preset prompt word template being a template corresponding to partial information of the target prompt word, the partial information of the target prompt word including an event identifier and a time identifier related to each preset target schedule; inputting the target prompt word and the target text data into a large language model for information extraction to obtain the schedule text information.

[0012] In some embodiments, the above-mentioned schedule generation method also includes: based on the target text data, determining the day program number corresponding to the target text data; in response to the value of the day program number corresponding to the target text data being less than or equal to the total number of schedules in the created target schedule, determining the day program number corresponding to the target text data as the target day program number so as to perform a search operation, a deletion operation, and / or a modification operation on the created target schedule corresponding to the target day program number.

[0013] The present application provides a schedule generation device, including: a conversion module, an extraction module, a processing module and a combination module. The conversion module is used for the target vehicle to perform text conversion on the target voice data to obtain the target text data, and the target voice data includes the near-end voice data collected by the audio collection device in the target vehicle and / or the far-end voice data received from other devices; the extraction module is used to extract information from the target text data to obtain the schedule text information related to the schedule in the target text data; the processing module is used to process the schedule text information based on the pre-set event identifier and time identifier to obtain the schedule event information and schedule time information; the combination module is used to combine the schedule event information and the schedule time information to obtain the target schedule.

[0014] The present application provides an electronic device, including a memory and a processor, wherein the processor is used to execute program instructions stored in the memory to implement the above-mentioned schedule generation method.

[0015] The present application provides a computer-readable storage medium on which program instructions are stored. When the program instructions are executed by a processor, the above-mentioned schedule generation method is implemented.

[0016] The above scheme processes the schedule text information related to the schedule obtained through information extraction based on the pre-set event identifier and time identifier, which can improve the accuracy of the processed schedule event information and schedule time information, and then combine the schedule event information and schedule time information to obtain the target schedule, which can improve the accuracy of the target vehicle's voice data processing to obtain the target schedule.

[0017] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings herein are incorporated into the specification and constitute a part of the specification. These drawings illustrate embodiments consistent with the present application and are used together with the specification to illustrate the technical solution of the present application.

[0019] Figure 1 This is a schematic diagram of the process of an embodiment of the schedule generation method of the present application. Figure 1 ;

[0020] Figure 2 This is a schematic diagram of the process of an embodiment of the schedule generation method of the present application. Figure 2 ;

[0021] Figure 3 This is a schematic diagram of the process of an embodiment of the schedule generation method of the present application. Figure 3 ;

[0022] Figure 4 This is a schematic diagram of the process of an embodiment of the schedule generation method of the present application. Figure 4 ;

[0023] Figure 5 It is a structural schematic diagram of an embodiment of a schedule generating device of the present application;

[0024] Figure 6 It is a structural schematic diagram of an embodiment of the electronic device of the present application;

[0025] Figure 7 It is a structural diagram of an embodiment of a computer-readable storage medium of the present application. DETAILED DESCRIPTION

[0026] The scheme of the embodiment of the present application is described in detail below in conjunction with the drawings of the specification.

[0027] In the following description, for the purpose of explanation rather than limitation, specific details such as specific system structures, interfaces, and technologies are provided to facilitate a thorough understanding of the present application.

[0028] The term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the objects associated before and after are in an "or" relationship. In addition, "many" in this article means two or more than two. In addition, the term "at least one" in this article means any combination of at least two of any one or more of a plurality of, for example, including at least one of A, B, and C, can mean including any one or more elements selected from the set consisting of A, B, and C.

[0029] The present application provides some schedule generation methods and schedule generation devices. The application scenarios of the schedule generation method include but are not limited to the generation of a target schedule by a vehicle-mounted system. The executor of the schedule generation method may be a schedule generation device. For example, the schedule generation device may be arranged in a terminal device or a server or other processing device, wherein the terminal device may be a device for schedule generation, a user equipment (UE), a mobile device, a user terminal, a terminal, a cellular phone, a cordless phone, a personal digital assistant (PDA), a handheld device, a computing device, a vehicle-mounted device, etc. In some possible implementations, the schedule generation method may be implemented by a processor calling computer-readable instructions stored in a memory.

[0030] See also Figure 1 , Figure 1 This is a schematic diagram of the process of an embodiment of the schedule generation method of the present application. Figure 1 .

[0031] Specifically, the schedule generation method may include the following steps:

[0032] Step S11: The target vehicle performs text conversion on the target voice data to obtain target text data.

[0033] The target voice data includes near-end voice data collected by the audio collection device in the target vehicle and / or far-end voice data received from other devices. The target vehicle may be a vehicle for which a schedule needs to be generated. The audio collection device may be an on-board audio collection device capable of collecting audio data inside the target vehicle, such as a microphone. The audio collection device may be set at a specific position in the target vehicle. The specific position may be the position of the driver or the position of other passengers. Near-end audio data may refer to voice data obtained by collecting audio data inside the target vehicle by the audio collection device in the target vehicle. Other devices may be peer devices communicating with the target vehicle. The peer devices may be other vehicles, base stations, cloud servers, and mobile terminals. Far-end voice data may refer to audio data sent to the target vehicle by other devices communicating with the target vehicle, and can be played by an audio playback device in the target vehicle. For example, the audio playback device in the target vehicle may be a vehicle-mounted speaker. For example, taking the peer device as a mobile terminal, the near-end voice data collected by the audio collection device of the mobile terminal is directly used as the far-end voice data, or the near-end voice data collected by the audio collection device of the mobile terminal is used as the far-end voice data after being processed by the preset method. The target text data may refer to all text contents of the target voice data after text conversion.

[0034] In some application scenarios, the target voice data includes the near-end voice data. In other application scenarios, the target voice data includes the far-end voice data. Wherein, when the target voice data includes only the near-end voice data or the far-end audio data, the target voice data is audio data related to a single-person single-round conversation. The scenario of a single-person single-round conversation may be that the target vehicle is in a single-person conversation scenario. In other application scenarios, the target voice data includes the near-end voice data and the far-end voice data. When the target voice data includes the near-end voice data and the far-end voice data, the target voice data is audio data related to a multi-person multi-round conversation. Exemplarily, the scenario of a multi-person multi-round conversation may be that the target vehicle is in a call scenario. Before the above step S11, it may also include identifying the scene in which the target vehicle is located to obtain a scene recognition result of the target vehicle. The scene recognition result includes that the target vehicle is in a single-person conversation scenario or a call scenario. The method for identifying the scene in which the target vehicle is located may be to determine whether the vehicle-mounted Bluetooth call function of the target vehicle is in use. In response to the target vehicle's in-vehicle Bluetooth call function not being in use, the scene recognition result of the target vehicle is determined as a single-person conversation scene. In response to the target vehicle's in-vehicle Bluetooth call function being in use, the scene recognition result of the target vehicle is determined as a call scene. In response to the target vehicle's scene recognition result being a call scene, the near-end voice data and the far-end voice data are channel-merged to obtain target voice data. Specifically, the far-end input signal and the near-end input signal during the call are spliced ​​in the channel dimension to obtain the target voice data. In other application scenarios, the near-end voice data after preset processing and the far-end voice data after preset processing are used as target voice data, and the preset processing may be echo cancellation processing or noise reduction processing.

[0035] In some application scenarios, the method of converting the target voice data into text may be to directly perform voice recognition on the target voice data to obtain the target text data. The specific voice recognition method may be to use an acoustic model to convert the input target voice data into a phoneme sequence, and then use a language model to convert the phoneme sequence into the final target text data. In other application scenarios, a pre-trained language model (for example, a BERT model) is used to perform natural language processing (NLP) on the target voice data to obtain the target text data.

[0036] Step S12: extracting information from the target text data to obtain schedule text information related to the schedule in the target text data.

[0037] Information extraction may be to extract text content of interest in the target text data. The text content of interest may be text content related to the schedule. The text content of interest may be set according to the accuracy requirement of the schedule generation. The text content of interest may be time content related to the schedule, or event content related to the schedule. The event content related to the schedule includes at least one entity object, relationship, attribute, etc. At least part of the content in the target text data is schedule text information. The schedule text information related to the schedule in the target text data may be at least part of the content in the target text data, or a combination of at least part of the content in the target text data and a separation mark. The separation mark may be used to divide different text contents of interest. The above step S12 may be to filter the target text data according to the text content of interest to obtain the schedule text information related to the schedule in the target text data. The above step S12 may be to input the target text data into a pre-trained information extraction model for information extraction to obtain the schedule text information related to the schedule.

[0038] Step S13: Processing the schedule text information based on the preset event identifier and time identifier to obtain schedule event information and schedule time information.

[0039] The event identifier is used to separate the event content related to the schedule in the schedule text information. The time identifier is used to separate the time content related to the schedule in the schedule text information. Exemplarily, the event identifier and the time identifier can be expressed as [Event] and [Time] respectively. The event identifier and the time identifier can be the above-mentioned separation marks, or can be related marks corresponding to the above-mentioned separation marks. It can be understood that the schedule text information obtained by information extraction can contain several separation marks, and the several separation marks can have compatible content or the same function as the event identifier and the time identifier. In the case where several separation marks have the same content as the event identifier and the time identifier, the separation marks included in the schedule text information are at least one event identifier and at least one time identifier.

[0040] In the case where the plurality of separation marks have the same function as the event identifier and the time identifier, the plurality of separation marks may include a related mark capable of separating the time field in the schedule text information and a related mark capable of separating the event field in the schedule text information. The related mark is different from the event identifier and the time identifier.

[0041] Schedule event information may refer to the event content used to indicate the final target schedule. The number of schedule event information may be at least one. Each target event information is the event content of the final generated schedule. Exemplarily, the event content corresponding to each schedule event information may include event name, specific time, specific location, participants, notes, event serial number, etc. Participants may be information that requires listing participants if the activity involves other people. Notes may be other related details or matters that require special attention. The event name may be a brief description of the event content, such as event types such as "project meeting" or "appointment". The specific time is at least one specific time point or at least one specific time period included in the schedule time period corresponding to the schedule time information. Schedule time information may refer to the time content used to indicate the final target schedule. The number of schedule time information may be at least one. Each target time information is the time content of the final generated schedule. The time content in the schedule time information is the schedule time period corresponding to the schedule time information. Each schedule time information may match any schedule event information. In some application scenarios, the way to match the schedule time information with the schedule event information may be to pre-determine the field content between two adjacent different separators as a group of schedule related information. Based on the pre-set event identifier and time identifier, each group of schedule-related information is processed to obtain the schedule event information and schedule time information corresponding to each group of schedule-related information, and automatically match them. In other application scenarios, the schedule text information is split according to adjacent event identifiers and time identifiers to obtain any group of matching schedule event information and schedule time information. Specifically, the adjacent event identifier and time identifier can be that the position of the event identifier is before the position of the time identifier, or that the position of the event identifier is after the position of the time identifier.

[0042] The above step S13 can also be that when the schedule text information includes an event identifier and a time identifier, the schedule text information is split according to the positions of the event identifier and the time identifier in the schedule text information to obtain the schedule event information and the schedule time information respectively. Specifically, the content of the field after the event identifier and before the next identifier is directly used as the schedule event information corresponding to the event identifier. The content of the field after the time identifier and before the next identifier is directly used as the schedule time information corresponding to the time identifier.

[0043] Step S14: Combine the schedule event information and the schedule time information to obtain the target schedule.

[0044] The target schedule may be a schedule that needs to be generated in the vehicle-mounted schedule application software or the vehicle-mounted schedule database. The number of target schedules may be one or more. Each target schedule includes a schedule time information and at least one schedule event information.

[0045] The way to combine the schedule event information and the schedule time information may be to combine the schedule event information and the schedule time information according to the schedule generation logic or the schedule generation template to obtain the target schedule. Wherein, the schedule generation logic or the schedule generation template may refer to the schedule generation rules built into the vehicle-mounted schedule application software or the vehicle-mounted schedule database. Exemplarily, the schedule generation logic may be to fill the schedule event information and the schedule time information matching the schedule event information into the event position and time position in the same schedule, generate the target schedule, and obtain the sorting sequence number of the target schedule in each created target schedule. Exemplarily, the schedule event information and the schedule time information matching the schedule event information may be filled into the schedule generation template to obtain the target schedule. And the target schedule is stored in the vehicle-mounted schedule application software or the vehicle-mounted schedule database to obtain the sorting sequence number of the target schedule in each created target schedule. It can be understood that, through step S14, a target schedule can be generated on the vehicle-mounted schedule application software of the target vehicle. In the case where the number of schedule event information and the number of schedule time information are at least one, at least one target schedule can be generated on the vehicle-mounted schedule application software of the target vehicle through step S14.

[0046] The above scheme processes the schedule text information related to the schedule obtained through information extraction based on the pre-set event identifier and time identifier, which can improve the accuracy of the processed schedule event information and schedule time information, and then combine the schedule event information and schedule time information to obtain the target schedule, which can improve the accuracy of the target vehicle's voice data processing to obtain the target schedule.

[0047] See also Figure 2 , Figure 2 This is a schematic diagram of the process of an embodiment of the schedule generation method of the present application. Figure 2 .

[0048] In some embodiments, the above step S13 may include the following steps:

[0049] Step S21: splitting the schedule text information according to the positions of the event identifier and the time identifier to obtain initial event information and initial time information.

[0050] The position of the event identifier may be the text position of the event identifier in the schedule text information. The position of the time identifier may be the text position of the time identifier in the schedule text information. The initial event information may be used to represent the field content obtained by splitting the schedule text information according to the event identifier. In some application scenarios, the field content after the event identifier and before the next identifier is used as the initial event information corresponding to the event identifier. In other application scenarios, the field content before the event identifier and after the previous identifier is used as the initial event information corresponding to the event identifier. The initial time information may be used to represent the field content obtained by splitting the schedule text information according to the time identifier. In some application scenarios, the field content after the time identifier and before the next identifier is used as the initial time information corresponding to the time identifier. In other application scenarios, the field content before the time identifier and after the previous identifier is used as the initial time information corresponding to the time identifier.

[0051] Exemplarily, in the case where the event identifier and the time identifier can be the above-mentioned separation marks, the separation marks included in the schedule text information are at least one event identifier and at least one time identifier. Taking the current separation mark as the time identifier as an example, the initial time information can be the field content between the current separation mark and the next separation mark, and the initial time information can also be the field content between the current separation mark and the previous separation mark. Taking the current separation mark as the event identifier as an example, the initial event information can be the field content between the current separation mark and the next separation mark, and the initial event information can also be the field content between the current separation mark and the previous separation mark. Among them, the previous separation mark or the next separation mark can be the previous separation mark adjacent to the current separation mark or the next separation mark adjacent to the current separation mark. It can be understood that, in the case where the event identifier and the time identifier are not the above-mentioned separation marks, the target separation marks corresponding to the event identifier and the time identifier in the schedule text information are determined. Among them, the target separation mark is the above-mentioned related mark with the same function as the event identifier and the time identifier. Taking the current separator mark as a related mark with the same function as the time identifier as an example, the initial time information can be the field content between the current target separator mark and the next target separator mark, and the initial time information can also be the field content between the current target separator mark and the previous target separator mark.

[0052] Step S22: Determine target time information and target event information according to the relationship between the initial time information and the preset time parameters.

[0053] The preset time parameter may characterize the field content of the time continuity between the start time and the end time in the time period, and the text position of the preset time parameter is the field content between the start time and the end time. It is understandable that the preset time parameter is used to represent the field content of time continuity. The field content of the preset time parameter may be "to", "to", "until", etc., or may be a comma or hyphen. Exemplarily, if T1 represents the start time, T2 represents the end time. In the case where the preset time parameter is "to", the initial time information may be T1 to T2. In the case where the preset time parameter is a hyphen, the initial time information may also be T1-T2. This is only an example, and the specific form of the field content of the preset time parameter is not limited. The relationship between the initial time information and the preset time parameter may be whether the field content in the initial time information includes the field content of the preset time parameter, or whether the field content in the initial time information includes the field content with the same function as the preset time parameter. Among them, having the same function as the preset time parameter may be the field content used to separate the start time and the end time in the time period, and the field content may be a time period separator identifier. The text position of the time period separator identifier may be between the start time and the end time.

[0054] The method of determining the target time information and the target event information may be: when the relationship between the initial time information and the preset time parameter meets the preset condition, the initial time information is determined as the target time information. The initial event information matching the initial time information is determined as the target event information. In some application scenarios, when the relationship between the initial time information and the preset time parameter does not meet the preset condition, the initial time information and the initial event information matching the initial time information are deleted. In other application scenarios, when the relationship between the initial time information and the preset time parameter does not meet the preset condition, the initial time information is subjected to time supplementation processing to obtain the target time information corresponding to the initial time information. The initial event information matching the initial time information is determined as the target event information. Among them, the preset condition may be that the field content in the initial time information includes the field content of the preset time parameter or includes a time period separator identifier. Among them, the time supplementation processing may be supplementing or updating the start time and end time in the initial time information to obtain the candidate start time and the candidate end time. In some application scenarios, the candidate start time and the candidate end time are directly used as the start time and the end time in the target time information. In other application scenarios, when the candidate start time and the candidate end time meet the time reasonableness condition, the candidate start time and the candidate end time are used as the start time and the end time in the target time information. The time reasonableness condition includes that the values ​​of the field contents corresponding to the candidate start time and the candidate end time are not empty, and / or, and the timestamp in the field contents corresponding to the candidate start time is the time before the timestamp in the field contents corresponding to the candidate end time.

[0055] In some embodiments, the above step S22 may include the following steps: in response to the initial time information including a preset time parameter, the time information adjacent to the preset time parameter in the initial time information is determined as the candidate start time and the candidate end time in the candidate time information. Or, in response to the initial time information not including the preset time parameter, the initial time information is determined as the candidate start time in the candidate time information. Next, based on the schedule time information corresponding to the created target schedule, the candidate end time in the candidate time information is determined. Subsequently, based on the value of the time range of the candidate time information and the initial event information, the target time information and the target event information are determined.

[0056] The candidate time information may represent the time information obtained by processing the initial time information. The candidate time information may be the time information having the same content as the initial time information, or may be the time information obtained by supplementing or updating the initial time information. The candidate time information includes the candidate start time and the candidate end time.

[0057] In some application scenarios, when the initial time information includes a preset time parameter, the time information adjacent to the preset time parameter in the initial time information is directly determined as the candidate start time and the candidate end time in the candidate time information. The time information adjacent to the preset time parameter includes the initial start time and the initial end time. Among them, the initial start time is the field content before the preset time parameter in the initial time information. The initial end time is the field content after the preset time parameter in the initial time information. Specifically, the field content before the preset time parameter in the initial time information is determined as the candidate start time, and the field content after the preset time parameter in the initial time information is determined as the candidate start time.

[0058] In other application scenarios, when the initial time information includes a preset time parameter, the initial time information is directly determined as the candidate start time. Specifically, the field content corresponding to the initial time information is directly used as the candidate start time. The initial time information can be supplemented by selecting a timestamp that does not conflict with the schedule time information corresponding to the created target schedule as the candidate end time. Specifically, the schedule time information corresponding to the created target schedule is sorted according to priority. The schedule time information corresponding to the created target schedule that meets the preset priority condition is compared with the initial time information to determine whether a time conflict occurs. The preset priority conditions include higher priority, lower priority, and middle priority. The created target schedule can be a historical target schedule. The schedule time information corresponding to the created target schedule can be the historical schedule time information corresponding to the historical target schedule. The historical time period corresponding to each historical schedule time information is a time period composed of the target start time and the target end time in the historical schedule information.

[0059] The above-mentioned step of determining the candidate end time in the candidate time information based on the schedule time information corresponding to the created target schedule includes: selecting a timestamp different from the target start time and / or target end time in the schedule time information corresponding to the created target schedule as the candidate end time. In other application scenarios, after the field content corresponding to the initial time information is directly used as the candidate start time, if the candidate start time is not in the historical time period corresponding to any historical schedule time information, a timestamp adjacent to the candidate start time and not belonging to the historical time period corresponding to any historical schedule time information is determined as the candidate end time corresponding to the candidate start time. The candidate end time corresponding to the candidate start time is directly used as the candidate end time in the candidate time information. At this time, the time range of the candidate time information is composed of the candidate start time and the candidate end time corresponding to the candidate start time. In other application scenarios, after the field content corresponding to the initial time information is directly used as the candidate start time, if the candidate start time is in the historical time period corresponding to any historical schedule time information, the candidate start time is updated to obtain a new candidate start time. The method for updating the candidate start time includes: determining a timestamp in a historical time period that is adjacent to the candidate start time and does not belong to any historical schedule time information as a new candidate start time corresponding to the candidate start time. Determining a timestamp in a historical time period that is adjacent to the new candidate start time and does not belong to any historical schedule time information as a candidate end time corresponding to the new candidate start time. At this time, the time range of the candidate time information is composed of the new candidate start time and the candidate end time corresponding to the new candidate start time.

[0060] In some application scenarios, the time range of the candidate time information is directly determined as the time range of the target time information. In other application scenarios, it is determined whether the time range of the candidate time information meets the reasonable time condition. If the reasonable time condition is met, the time range of the candidate time information is directly determined as the time range of the target time information. If the reasonable time condition is not met, the candidate time information and the candidate event information corresponding to the candidate time information are discarded. The reasonable time condition includes that the value of the field content corresponding to the candidate start time and the candidate end time in the time range of the candidate time information is not empty, and / or the timestamp in the field content corresponding to the candidate start time is the time before the timestamp in the field content corresponding to the candidate end time. In other application scenarios, the initial event information is directly determined as the target event information. In other application scenarios, the initial event information is preprocessed to obtain the target event information. The way to preprocess the initial event information can be to update the initial event information according to the event template to obtain the target event information. The event template can be a requirement for the expected format of the target event information.

[0061] In some embodiments, the target time information includes a target start time and a target end time. The above step of determining the target time information and the target event information based on the value of the time range of the candidate time information and the initial event information may include the following steps: in response to the value of the candidate start time in the candidate time information being empty, the candidate time information and the initial event information are deleted. Or, in response to the value of the candidate start time in the target time information being not empty, it is determined whether the candidate start time is a moment before the candidate end time. Then, in response to the candidate start time being a moment before the candidate end time, the initial event information and the candidate time information are used as the target event information and the target time information, respectively.

[0062] The target start time is the start time in the time range where the target time information is located. The target end time is the end time in the time range where the target time information is located. The value of the candidate start time is the field content corresponding to the candidate start time in the candidate time information. The value of the candidate start time is empty, which may be that the field content corresponding to the candidate start time does not exist or the field content corresponding to the candidate start time is not a specific timestamp. The value of the candidate end time is the field content corresponding to the candidate end time in the candidate time information. The value of the candidate end time is empty, which may be that the field content corresponding to the candidate end time does not exist or the field content corresponding to the candidate end time is not a specific timestamp.

[0063] In some application scenarios, when the value of the candidate start time in the candidate time information is empty, the candidate time information and the initial event information are deleted. In other application scenarios, when the value of the candidate start time in the candidate time information is not empty, the initial event information and the candidate time information are directly used as the target event information and the target time information, respectively. In other application scenarios, when the value of the candidate start time in the candidate time information is not empty, it is determined whether the candidate start time is a moment before the candidate end time. In some application scenarios, when the candidate start time is a moment before the candidate end time, the initial event information and the candidate time information are used as the target event information and the target time information, respectively. At this time, the initial event information and the candidate time information are valid information. In other application scenarios, when the candidate start time is not a moment before the candidate end time, the candidate time information and the initial event information are deleted.

[0064] Step S23: Determine the schedule time information based on the target time information.

[0065] In some application scenarios, the target time information is directly determined as the schedule time information. In other application scenarios, it is determined whether the target time information has a time conflict with the historical target time information of several historical target schedules. If the target time information does not have a time conflict with the historical target time information of several historical target schedules, the target time information is directly determined as the schedule time information. If the target time information has a time conflict with the historical target time information of several historical target schedules, a time range that does not conflict with the time range of the several historical target time information is randomly selected as the time range of the schedule time information.

[0066] See also Figure 3 , Figure 3 This is a schematic diagram of the process of an embodiment of the schedule generation method of the present application. Figure 3 .

[0067] In some embodiments, there are multiple target time information, and the above step S23 may include the following steps: for each target time information, perform the following steps: Figure 3The following steps are shown: Step S31: Determine whether the time range of the target time information overlaps with the time range of the reference time information. The reference time information includes other target time information other than the target time information and / or schedule time information corresponding to the created target schedule. Step S32: In response to the time range of the target time information not overlapping with the time range of the reference time information, determine the target time information as schedule time information. Step S33: In response to the time range of the target time information overlapping with the time range of the reference time information, determine the overlapping time information corresponding to the target time information. The overlapping time information is at least one of the reference time information. Step S34: Determine the schedule time information based on the target time information and the overlapping time information.

[0068] Among them, the number of other target time information except the target time information can be at least one. It can be understood that the initial time information and the initial time information obtained by splitting in the above step S21 are both multiple, and the schedule text information is split using the positions of the event identifier and the time identifier to obtain multiple initial event information and multiple initial time information. The above step S22 can be: for each initial time information, determine the target time information and the target event information according to the relationship between the initial time information and the preset time parameter. The number of target time information is less than or equal to the number of initial time information. The reference time information includes a number of first reference time information and / or a number of second reference time information. The number of first reference time information includes other target time information except the target time information. The number of second reference time information includes schedule time information corresponding to the created target schedule.

[0069] In some application scenarios, when the reference time information includes several first reference time information, it is determined whether the time range of the target time information overlaps with the time range of each first reference time information. When the time range of the target time information does not overlap with the time range of any first reference time information, the target time information is determined as schedule time information. When the time range of the target time information overlaps with the time range of at least one first reference time information, the overlapping time information corresponding to the target time information is determined. The overlapping time information is at least one of the first reference time information. Among them, the target start time and / or target end time in the time range of the target time information are within the time range of at least one first reference time information.

[0070] In some other application scenarios, when the reference time information includes a plurality of second reference time information, it is determined whether the time range of the target time information overlaps with the time range of each second reference time information. When the time range of the target time information does not overlap with the time range of any second reference time information, the target time information is determined as schedule time information. When the time range of the target time information overlaps with the time range of at least one second reference time information, the overlapping time information corresponding to the target time information is determined. The overlapping time information is at least one of the second reference time information. Among them, the target start time and / or target end time in the time range of the target time information are within the time range of at least one second reference time information.

[0071] In some other application scenarios, when the reference time information includes a plurality of first reference time information and a plurality of second reference time information, it is determined whether the time range of the target time information overlaps with the time range of each reference time information. When the time range of the target time information does not overlap with the time range of any reference time information, the target time information is determined as schedule time information. When the time range of the target time information overlaps with the time range of at least one reference time information, the overlapping time information corresponding to the target time information is determined. The overlapping time information is at least one of the reference time information. Among them, the target start time and / or target end time in the time range of the target time information are within the time range of at least one reference time information.

[0072] The time range of the overlapping time information may be the overlapping start time to the overlapping end time. Exemplarily, when both the target start time and the target end time are within the time range of the same reference time information, the overlapping start time is the target start time, and the overlapping end time is the target end time. Exemplarily, when the target start time is within the time range of the first target reference time information, and the target end time is not within the time range of any reference time information, the overlapping start time is the target start time, and the overlapping end time is the end time corresponding to the time range of the first target reference time information. The time range of the first target reference time information includes the target start time. Exemplarily, when the target start time is not within the time range of any reference time information, and the target end time is within the time range of the second target reference time information, the overlapping start time is the start time corresponding to the time range of the second target reference time information, and the overlapping end time is the target end time. The time range of the second target reference time information includes the target end time. Exemplarily, when the target start time is within the time range of the first target reference time information, and the target end time is within the time range of the second target reference time information, the overlapping start time is the target start time, and the overlapping end time is the target end time. The time range of the first target reference time information includes the target start time. The time range of the second target reference time information includes the target end time.

[0073] In other application scenarios, each reference time information is sorted according to priority. When executing the above step S31, it is only determined whether the time range of the target time information overlaps with the time range of the candidate reference time information. The candidate reference time information is the first reference time information and / or the second reference time information that meets the preset priority conditions. The preset priority conditions include higher priority, lower priority and middle priority. In some application scenarios, determining whether the time range of the target time information overlaps with the time range of the candidate reference time information includes: determining whether the time range of the target time information overlaps with the time range of the candidate reference time information and / or the first candidate time information. Among them, when the candidate reference time information is the first reference time information or the second reference time information that meets the preset priority conditions, the first candidate time information is all the second reference time information or all the first reference time information.

[0074] The above step S34 may be to select the time information corresponding to the longest time period or the shortest time period from the two time periods corresponding to the target time information and the overlapping time information as the final time information. Then the time range of the final time information is directly used as the time range of the schedule time information. The final time information is any one of the target time information and the overlapping time information.

[0075] In some embodiments, the above step S34 may include the following steps: merging the time range of the target time information and the time range of the overlapping time information to obtain a merged time range, and using the merged time range as the time range of the schedule time information. Or, selecting at least part of the time range from the optional time range as the time range of the schedule time information. The optional time range is other time ranges except the time range of the reference time information.

[0076] The merged time range includes the time range of the target time information and the time range of the overlapping time information. After the above step S33, the reference event information corresponding to the first target reference time information and / or the reference event information corresponding to the second target reference time information is used as the event information to be merged. The reference event information is the target event information that matches the first target reference time information in the reference time information and / or the schedule event information that matches the second target reference time information. After obtaining the merged time range, the target event information is merged with the event information to be merged to obtain new target event information.

[0077] The optional time range may be other time ranges except the time range in which the reference time information is located. In some application scenarios, the method of selecting at least part of the time range from the optional time range as the time range in which the schedule time information is located may be: at least part of the time range is selected from the optional time range to be before the overlapping start time and closest to the overlapping start time. Alternatively, at least part of the time range is selected from the optional time range to be after the overlapping end time and closest to the overlapping end time. In other application scenarios, the method of selecting at least part of the time range from the optional time range as the time range in which the schedule time information is located may also be: the overlapping time information and the reference time information are input into the large language model to obtain the free time found by the large language model. Compared with searching using methods such as segment trees, the large language model can combine real information to provide users with a more reasonable free time.

[0078] Step S24: Use the target event information as schedule event information.

[0079] In some application scenarios, the target event information is directly used as the schedule time information. In other application scenarios, the target event information is updated to obtain new target event information, and the new target event information is used as the schedule time information.

[0080] After the above step S14, the target schedule may be added as a schedule in the vehicle-mounted schedule application software or the vehicle-mounted schedule database, or the target schedule that has been created in the vehicle-mounted schedule application software or the vehicle-mounted schedule database may be modified or replaced.

[0081] It can be considered that the schedule-related information is segmented according to the time identifier and the event identifier to avoid the influence of other irrelevant information. In addition, the present application automatically supplements the missing information and finally completes the corresponding operation to realize the addition, modification or replacement of the schedule in the vehicle schedule application software or the vehicle schedule database, and can improve the convenience of the schedule-related operations.

[0082] In some embodiments, the above step S12 may include the following steps: first, based on a preset prompt word template and the text content of at least one preset target schedule, determine the target prompt word of the target text data. The preset prompt word template is a template corresponding to the partial information of the target prompt word. The partial information of the target prompt word includes an event identifier and a time identifier related to each preset target schedule. Then, the target prompt word and the target text data are input into a large language model for information extraction to obtain the schedule text information.

[0083] The text content of at least one preset target schedule. The text content of any preset target schedule can be used to fill the preset prompt word template to obtain the target prompt word. At least one of the preset target schedules is selected according to a random order, a polling order, or a custom order as the text content of at least one preset target schedule. The text content of at least one preset target schedule is directly filled into the preset prompt word template to obtain the target prompt word of the target text data.

[0084] The large language model may be a large language model (LLM). For example, the large language model may be GPT, Wenxinyiyan, etc., or may be other language models that can extract information from input target prompt words and target text data to obtain schedule text information, and this application does not limit this.

[0085] In some application scenarios, the preset prompt word template includes a first separation mark for dividing the text content of the preset target schedule. The text content of the preset target schedule is equivalent to the input content of the large language model. Event identifier and time identifier. The event identifier is used to separate the event-related field content in the text content of each preset target schedule. The time identifier is used to separate the time-related field content in the text content of each preset target schedule. In other application scenarios, the preset prompt word template may also include a second separation mark for dividing the input content and output content of the large language model. In other application scenarios, the schedule text information related to the schedule includes the time information related to the schedule and the event information related to the schedule. The preset prompt word template includes instruction information for extracting information from the target text data. The instruction information can be used to indicate the extraction of the time information related to the schedule and the event information related to the schedule in the target text data. Among them, the time information related to the schedule includes the start time and the end time that match the event information related to the schedule. It can be understood that the preset prompt word template includes at least a plurality of first separation marks, a plurality of second separation marks and instruction information.

[0086] The above-mentioned method of determining the target prompt word of the target text data based on the preset prompt word template and the text content of at least one preset target schedule includes: filling at least part of the text content of at least one preset target schedule into the preset prompt word template to obtain the target prompt word of the target text data. Specifically, taking the example of filling the target selection information in the text content of any preset target schedule into the preset prompt word template to obtain the target prompt word of the target text data, the target selection information includes mandatory information. The mandatory information includes the time information related to the schedule and the event information related to the schedule in the preset target schedule. It can be understood that the target selection information can only include mandatory information. The target selection information can also include optional information, and the optional information is at least one of the following: the optional information includes background information, the role played by the large language model, the example of extracting the schedule, and the output type of the large language model. Among them, the output type of the large language model is used to indicate the initial time information and initial event information corresponding to the schedule text information output by the large language model with a specific separation mark. Exemplarily, the target prompt word of the target text data obtained may include: instruction information, input data, mandatory information and optional information. The input data is the target text data.

[0087] It is understandable that the above step of filling at least part of the text content of at least one preset target schedule into the preset prompt word template to obtain the target prompt word of the target text data needs to be performed according to the following requirements: the first requirement is that the target prompt word includes instruction information, and the prompt model needs to extract the schedule text information in the schedule initial time information and the schedule initial event information. The second requirement is that the target prompt word includes a separation mark. Considering the richness of the output of the large language model, the output may contain a large amount of redundant information, which is not convenient for subsequent automatic processing. It is necessary to use a separation mark to add a separation identifier to the schedule text information. The third requirement is that the target prompt word needs to remove irrelevant information in the target text data. The event mentioned in the specific target text data may be in the past, or it may be just an assumption, without a clear plan, which needs to be eliminated. The fourth requirement is that the target prompt word needs to prompt the current time, so that the large language model can perform relative time reasoning to obtain time information related to the schedule, such as reasoning "two days later" to obtain the start time and end time matching the event information related to the schedule. The fifth requirement is that examples need to be given in the target prompt word so that the large language model can output in an example format. For example, the target prompt word can refer to the following content: Extract the schedule that "p1" intends to do in the future from the dialogue. The schedule needs to output [Time] (including "start time" and "end time") and [Event] in sequence, and [Event] is described in one sentence. The following is an example: [Date and time during the dialogue] 8:00 on April 12, 2004 (Monday). p2: Hello, at three o'clock in the afternoon, let's go to XX Square for two hours. p1: OK, let's take the subway and see you in half an hour. The extracted event is "Event: [Time] 15:00 on April 12, 2004 to 17:00 on April 12, 2004 [Event] Go shopping at XX Square. [Time] 14:00 on April 12, 2004 [Event] Meet at Changhe Road on Line 6." Now I will give you the following dialogue. Please give the schedule that "p1" intends to do in the future according to the format of the example.

[0088] Among them, [Time] and [Event] are the time identifier and event identifier in the important information separator, respectively, so that each schedule output by the large language model has a [Time] symbol at the beginning of the time, [Event] at the beginning of the event, and ends with a period. "p1, p2" in the dialogue content example are identifiers set by the dialogue management module to distinguish speakers. In the schedule extraction example, the present application also implicitly prompts the preset time parameter "to" between the start time and the end time. It can be considered that the target prompt word of the present application can limit the schedule text information obtained by extracting information from the large language model to the information required for the schedule.

[0089] In some application scenarios, the user operation intention is analyzed from the schedule text information output by the large language model to obtain the intention analysis result. The intention analysis result may include at least one of the following: creating a schedule, retrieving a schedule, and deleting a schedule. The above method of obtaining the intention analysis result can use but is not limited to using the large language model for inference. According to the intention analysis result, the text content is input to respond to the operation.

[0090] See also Figure 4 , Figure 4 This is a schematic diagram of the process of an embodiment of the schedule generation method of the present application. Figure 4 .

[0091] In some embodiments, the schedule generation method further includes: step S41: determining the daily program number corresponding to the target text data based on the target text data. Step S42: in response to the value of the daily program number corresponding to the target text data being less than or equal to the total number of schedules in the created target schedule, determining the daily program number corresponding to the target text data as the target daily program number so as to perform a search operation, a deletion operation, and / or a modification operation on the created target schedule corresponding to the target daily program number.

[0092] The target text data is input into the large language model, and the serial number text information corresponding to the target text data output by the large language model is determined. The serial number text information is segmented by the serial number identifier to obtain the daily program number to be operated, which is the daily program number corresponding to the target text data.

[0093] The key point in the process of the large language model completing the extraction of the day program number is to design a reasonable serial number prompt word, so that the large language model can output the accurate day program number according to the user command information and the existing schedule information. The setting of the serial number prompt word must meet the following conditions: The first condition is to add a serial number identifier. If the model is only required to output a digital serial number, it is easy to cause the day program number to be operated to be mixed with other digital content, so it needs to be distinguished. The second condition specifies the output of the large language model for exceptional cases. When the user command does not have a corresponding day program number, in order to prevent the large language model from outputting too long, the large language model reply content is restricted. The third condition is to prompt the current time to facilitate the large language model to reason about relative time, such as "the second schedule in two days." Exemplarily, a basic sequence number prompt for completing the task is as follows: "The confirmed schedule is "1. Go hiking from 9:00 on August 2, 2004 to 11:00 on August 2, 2024". The current time is August 1, 2004 (Monday). The question is "***", please tell me the daily program number to be deleted. If there is one, the sequence number starts from 1 and returns with "A" at the beginning; if not, return "None"."

[0094] In the example of the sequence number prompt word, "***" is the user input content. In the example of the sequence number prompt word, we use "A" as the sequence number identifier of the sequence number to be operated, and prompt the large language model to only return "none" when no relevant schedule is queried.

[0095] When the value of the day program number corresponding to the target text data is less than or equal to the total number of schedules in the created target schedule, the day program number corresponding to the target text data output by the large language model is a valid serial number, and the day program number corresponding to the target text data is determined as the target day program number.

[0096] In some application scenarios, the operation command corresponding to the operation command of the user on the vehicle display interface of the target vehicle. The user's operation command can represent that the user expects to search, delete, and / or modify the target schedule corresponding to the target day program number. The target vehicle performs target processing on the created target schedule corresponding to the target day program number according to the operation instruction corresponding to the received operation command. The target processing includes at least one of the following: search operation, delete operation, and / or modify operation. In some application scenarios, the target processing may be to perform a search operation first, then a delete operation or a modify operation. In other application scenarios, the target processing may be to perform a search operation first, then a modify operation, and then a delete operation. It can be understood that when the target schedule corresponding to the target day program number is subjected to target processing, the operation response result corresponding to the target processing is fed back to the display interface. The operation response result includes at least one of the following: a successful search result or a failed search result, a successful deletion result or a failed deletion result, a successful modification result or a failed modification result. In other application scenarios, the sequence number text information corresponding to the target text data output by the large language model carries the operation instruction information corresponding to the user's operation command. The target vehicle performs the above target processing on the created target schedule corresponding to the target day program number according to the operation instruction information.

[0097] It can be considered that by determining the day program number from the target text data, when the day program number meets the corresponding conditions, the day program number is determined as the target day program number, and the created target schedule corresponding to the target day program number is subjected to target processing, which can improve the convenience of searching, deleting, and / or modifying the created target schedule.

[0098] The above scheme processes the schedule text information related to the schedule obtained through information extraction based on the pre-set event identifier and time identifier, which can improve the accuracy of the processed schedule event information and schedule time information, and then combine the schedule event information and schedule time information to obtain the target schedule, which can improve the accuracy of the target vehicle's voice data processing to obtain the target schedule.

[0099] See also Figure 5 , Figure 5 It is a structural diagram of an embodiment of a schedule generation device of the present application. The schedule generation device 50 includes a conversion module 51, an extraction module 52, a processing module 53 and a combination module 54. The conversion module 51 is used for the target vehicle to perform text conversion on the target voice data to obtain the target text data, and the target voice data includes the near-end voice data collected by the audio collection device in the target vehicle and / or the far-end voice data received from other devices; the extraction module 52 is used to extract information from the target text data to obtain the schedule text information related to the schedule in the target text data; the processing module 53 is used to process the schedule text information based on the pre-set event identifier and time identifier to obtain the schedule event information and schedule time information; the combination module 54 is used to combine the schedule event information and the schedule time information to obtain the target schedule.

[0100] The above scheme processes the schedule text information related to the schedule obtained through information extraction based on the pre-set event identifier and time identifier, which can improve the accuracy of the processed schedule event information and schedule time information, and then combine the schedule event information and schedule time information to obtain the target schedule, which can improve the accuracy of the target vehicle's voice data processing to obtain the target schedule.

[0101] Please refer to the schedule generation method for the functions performed by each module, which will not be repeated here.

[0102] See also Figure 6 , Figure 6 6 is a schematic diagram of the structure of an embodiment of a vehicle of the present application. The electronic device 60 includes a memory 61 and a processor 62, and the processor 62 is used to execute program instructions stored in the memory 61 to implement the steps in the above-mentioned schedule generation method embodiment. In a specific implementation scenario, the electronic device 60 may include but is not limited to: a microcomputer, a server, and in addition, the electronic device 60 may also include a mobile device such as a laptop computer and a tablet computer, which is not limited here.

[0103] Specifically, the processor 62 is used to control itself and the memory 61 to implement the steps in the above-mentioned schedule generation method embodiment. The processor 62 can also be called a CPU (Central Processing Unit). The processor 62 may be an integrated circuit chip with signal processing capabilities. The processor 62 can also be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field-programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. In addition, the processor 62 can be implemented by an integrated circuit chip.

[0104] The above scheme processes the schedule text information related to the schedule obtained through information extraction based on the pre-set event identifier and time identifier, which can improve the accuracy of the processed schedule event information and schedule time information, and then combine the schedule event information and schedule time information to obtain the target schedule, which can improve the accuracy of the target vehicle's voice data processing to obtain the target schedule.

[0105] See also Figure 7 , Figure 7 The computer-readable storage medium 70 stores program instructions 701, which implement the steps in any of the above-mentioned schedule generation method embodiments when executed by a processor.

[0106] The above scheme processes the schedule text information related to the schedule obtained through information extraction based on the pre-set event identifier and time identifier, which can improve the accuracy of the processed schedule event information and schedule time information, and then combine the schedule event information and schedule time information to obtain the target schedule, which can improve the accuracy of the target vehicle's voice data processing to obtain the target schedule.

[0107] In some embodiments, the functions or modules included in the system provided by the embodiments of the present disclosure can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments, and for the sake of brevity, it will not be repeated here.

[0108] The above description of various embodiments tends to emphasize the differences between the various embodiments. The same or similar aspects can be referenced to each other, and for the sake of brevity, they will not be repeated herein.

[0109] In the several embodiments provided in the present application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device implementation described above is only schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation, such as units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0110] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0111] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) or a processor (processor) to perform all or part of the steps of each implementation method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code.

Claims

1. A schedule generation method, characterized in that: The method comprises: The target vehicle performs text conversion on the target voice data to obtain target text data, wherein the target voice data includes near-end voice data collected by an audio collection device in the target vehicle and / or far-end voice data received from other devices; Extracting information from the target text data to obtain schedule text information related to the schedule in the target text data; Processing the schedule text information based on a preset event identifier and a time identifier to obtain schedule event information and schedule time information; The schedule event information and the schedule time information are combined to obtain a target schedule.

2. The method according to claim 1, characterized in that The processing of the schedule text information based on the preset event identifier and time identifier to obtain schedule event information and schedule time information includes: Splitting the schedule text information using the positions of the event identifier and the time identifier to obtain initial event information and initial time information; Determining the target time information and target event information according to the relationship between the initial time information and the preset time parameter; Determining the schedule time information based on the target time information; The target event information is used as the schedule event information.

3. The method according to claim 2, characterized in that The determining the target time information and the target event information according to the relationship between the initial time information and the preset time parameter includes: In response to the initial time information including the preset time parameter, determining the time information adjacent to the preset time parameter in the initial time information as the candidate start time and the candidate end time in the candidate time information respectively; or, In response to the initial time information not including the preset time parameter, determining the initial time information as the candidate start time in the candidate time information; and determining the candidate end time in the candidate time information based on the schedule time information corresponding to the created target schedule; The target time information and the target event information are determined based on the value of the time range in which the candidate time information is located and the initial event information.

4. The method according to claim 3, characterized in that The target time information includes a target start time and a target end time, and the determining of the target time information and the target event information based on the value of the time range of the candidate time information and the initial event information includes: In response to the value of the candidate start time in the candidate time information being empty, deleting the candidate time information and the initial event information; or, In response to the value of the candidate start time in the target time information being not empty, determining whether the candidate start time is a moment before the candidate end time; In response to the candidate start time being a time before the candidate end time, the initial event information and the candidate time information are used as the target event information and the target time information, respectively.

5. The method according to claim 2, characterized in that: The target time information is multiple, and the determining the schedule time information based on the target time information includes: For each target time information, perform the following steps: Determine whether the time range of the target time information overlaps with the time range of the reference time information, wherein the reference time information includes other target time information other than the target time information and / or schedule time information corresponding to the created target schedule; In response to the time range of the target time information not overlapping with the time range of the reference time information, determining the target time information as the schedule time information; or, In response to a time range where the target time information is located overlapping with a time range where the reference time information is located, determining overlapping time information corresponding to the target time information, wherein the overlapping time information is at least one of the reference time information; The schedule time information is determined based on the target time information and the overlapping time information.

6. The method according to claim 5, characterized in that The determining the schedule time information based on the target time information and the overlapping time information includes: Merging the time range of the target time information and the time range of the overlapping time information to obtain a merged time range, and using the merged time range as the time range of the schedule time information; or, At least a part of the time range is selected from the optional time range as the time range where the schedule time information is located, and the optional time range is other time ranges except the time range where the reference time information is located.

7. The method according to any one of claims 1 to 6, characterized in that: The step of extracting information from the target text data to obtain schedule text information related to the schedule in the target text data includes: Determining a target prompt word of the target text data based on a preset prompt word template and text content of at least one preset target schedule, wherein the preset prompt word template is a template corresponding to partial information of the target prompt word, and the partial information of the target prompt word includes an event identifier and a time identifier related to each of the preset target schedules; The target prompt word and the target text data are input into a large language model for information extraction to obtain the schedule text information.

8. The method according to any one of claims 1 to 6, characterized in that: The method further comprises: Based on the target text data, determining a daily program number corresponding to the target text data; In response to the value of the day program number corresponding to the target text data being less than or equal to the total number of schedules in the created target schedule, the day program number corresponding to the target text data is determined as the target day program number so as to perform a search operation, a deletion operation, and / or a modification operation on the created target schedule corresponding to the target day program number.

9. A schedule generating device, characterized in that: include: A conversion module, used for the target vehicle to perform text conversion on the target voice data to obtain target text data, wherein the target voice data includes near-end voice data collected by the audio collection device in the target vehicle and / or far-end voice data received from other devices; An extraction module, used to extract information from the target text data to obtain schedule text information related to the schedule in the target text data; A processing module, used for processing the schedule text information based on a preset event identifier and a time identifier to obtain schedule event information and schedule time information; The combination module is used to combine the schedule event information and the schedule time information to obtain a target schedule.

10. An electronic device, characterized in that: include: A memory and a processor, wherein the memory stores program instructions, and the processor retrieves the program instructions from the memory to execute the method according to any one of claims 1 to 8.

11. A computer-readable storage medium having program instructions stored thereon, characterized in that: When the program instructions are executed by a processor, they are used to implement the method according to any one of claims 1 to 8.

Images