Multimodal event processing method, device, equipment and medium

Abstract

The embodiments of the present disclosure disclose event processing methods, devices, equipment and media based on multimodality. A specific implementation of the method includes: in response to receiving event query information, querying whether there is event processing information corresponding to the first target event; in response to determining that it does not exist, determining at least one event; determining at least one event description location; extracting event specification information from at least one specification file; generating first event processing information, first event processing image and first event processing audio; publishing the first event processing information, first event processing image and first event processing audio to the first event processing platform; popping up a text display window to display image controls and audio controls. This implementation can publish accurately generated event processing information, event processing images and event processing audio to the event processing platform in a timely manner to provide an event processing method under multimodality.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of computer technology, and in particular to a multimodal event processing method, apparatus, device, and medium. Background Art

[0002] At present, in various fields, efficient handling of events is crucial to effectively and promptly solve corresponding problems. For the handling of events, the usual approach is to make subjective judgments based on the historical work experience of relevant staff to determine the handling method of the event.

[0003] However, when the above method is adopted, the following technical problems often occur:

[0004] Differences in personal experience and subjective judgment may lead to inaccurate and inefficient judgments, and there is also a lack of clear definition of whether an event meets the conditions for handling. In addition, in terms of event handling interaction, the interaction mode is relatively single, which does not allow users to clearly understand the content of event handling information under the required mode, resulting in a poor interaction experience.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the inventive concept and therefore it may contain information that does not form the prior art that is already known in this country to a person of ordinary skill in the art. Summary of the invention

[0006] The content of this disclosure is used to introduce concepts in a brief form, which will be described in detail in the detailed implementation section below. The content of this disclosure is not intended to identify the key features or essential features of the technical solution claimed for protection, nor is it intended to limit the scope of the technical solution claimed for protection.

[0007] Some embodiments of the present disclosure propose a multimodal-based event processing method, apparatus, device, and medium to solve one or more of the technical problems mentioned in the above background technology section.

[0008] In a first aspect, some embodiments of the present disclosure provide a multimodal event processing method, comprising: in response to a first event processing platform corresponding to a first region receiving event query information from a target user for a first target event, querying from an event processing database whether there is event processing information corresponding to the first target event, wherein the event processing information includes: an event processing path and event processing evaluation information; in response to determining that there is no event, determining from the event processing database at least one event whose corresponding event content similarity with the first target event is greater than a target similarity; based on the at least one event, determining at least one event description position of the first target event in at least one specification file; based on the at least one event, determining at least one event description position of the first target event in at least one specification file; an event description location, extracting event specification information corresponding to the first target event from the at least one specification file; generating first event processing information, a first event processing image and a first event processing audio using a pre-trained event processing information generation model according to the event specification information; publishing the first event processing information, the first event processing image and the first event processing audio to the first event processing platform; in response to determining that the target user clicks on an event processing control in an event processing page corresponding to the first event processing platform, popping up a text display window corresponding to the first event processing information, as well as displaying an image control corresponding to the first event processing image and an audio control corresponding to the first event processing audio.

[0009] In a second aspect, some embodiments of the present disclosure provide an event processing device based on multimodality, comprising: a query unit, configured to, in response to a first event processing platform corresponding to a first region receiving event query information from a target user for a first target event, query whether there is event processing information corresponding to the above-mentioned first target event from an event processing database, wherein the above-mentioned event processing information includes: an event processing path and event processing evaluation information; a first determination unit, configured to, in response to determining that there is no event, determine from the above-mentioned event processing database at least one event whose corresponding event content similarity with the above-mentioned first target event is greater than the target similarity; a second determination unit, configured to determine, based on the above-mentioned at least one event, at least one event description position of the above-mentioned first target event in at least one specification file; an extraction unit, configured to According to the above-mentioned at least one event description location, event specification information corresponding to the above-mentioned first target event is extracted from the above-mentioned at least one specification file; a generation unit is configured to generate first event processing information, a first event processing image and a first event processing audio according to the above-mentioned event specification information using a pre-trained event processing information generation model; a publishing unit is configured to publish the above-mentioned first event processing information, the above-mentioned first event processing image and the above-mentioned first event processing audio to the above-mentioned first event processing platform; a display unit is configured to pop up a text display pop-up window corresponding to the above-mentioned first event processing information in response to determining that the above-mentioned target user clicks on the event processing control in the event processing page corresponding to the above-mentioned first event processing platform, and display an image control corresponding to the above-mentioned first event processing image and an audio control corresponding to the first event processing audio.

[0010] In a third aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device on which one or more programs are stored, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation manner in the first aspect.

[0011] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method described in any implementation manner in the first aspect is implemented.

[0012] The above-mentioned various embodiments of the present disclosure have the following beneficial effects: through the event processing method based on multimodality of some embodiments of the present disclosure, the accurately generated event processing information, event processing images and event processing audio can be timely published to the event processing platform to provide an event processing method under multimodality. Specifically, the reason why the relevant event processing is not efficient and accurate is that the judgment may be inaccurate and inefficient due to differences in personal experience and subjective judgment, and there is also a lack of clear definition of whether the event meets the conditions for handling. In addition, in terms of event processing interaction, the interaction mode is relatively single, and users cannot clearly understand the content of the event processing information under the required mode, resulting in a poor interactive experience. Based on this, the event processing method based on multimodality of some embodiments of the present disclosure, first, in response to the first event processing platform corresponding to the first region receiving the event query information of the target user for the first target event, query whether there is event processing information corresponding to the above-mentioned first target event from the event processing database, wherein the above-mentioned event processing information includes: event processing path and event processing evaluation information. Here, after receiving the event query information of the first target event, it is first determined by querying the event processing database whether there is corresponding event processing information, so as to improve the processing efficiency of the first target event. Then, in response to determining that there is no corresponding event, at least one event whose corresponding event content similarity with the first target event is greater than the target similarity is determined from the event processing database. Here, the specification content related to the first target event is subsequently determined by determining at least one similar event, so as to determine how the first target event should be processed. Then, according to the at least one event, at least one event description position of the first target event in at least one specification file can be accurately determined. Then, according to the at least one event description position, the event specification information corresponding to the first target event is extracted from the at least one specification file to determine the processing specification requirements related to the first target event. Furthermore, according to the event specification information, the first event processing information, the first event processing image and the first event processing audio under multimodality can be accurately generated by using the pre-trained event processing information generation model, so as to facilitate the subsequent multimodal display of the processing results of the first target event and increase the diversity of the display. Furthermore, the first event processing information, the first event processing image and the first event processing audio are published to the first event processing platform, so as to facilitate feedback to the target user to view the corresponding event processing information. Finally, in response to determining that the target user clicks on the event processing control in the event processing page corresponding to the first event processing platform, a text display pop-up window corresponding to the first event processing information is popped up, as well as an image control corresponding to the first event processing image and an audio control corresponding to the first event processing audio are displayed, so as to facilitate the target user to view the processing status of the first target event in a multi-modal manner.In summary, through the method of querying the event processing database and determining at least one similar event, as well as the use of the event processing information generation model, accurately generated event processing information, event processing images and event processing audio can be published to the event processing platform in a timely manner to provide a multi-modal event processing method. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that components and elements are not necessarily drawn to scale.

[0014] Figure 1 is a flowchart of some embodiments of a multimodal event processing method according to the present disclosure;

[0015] Figure 2 is a schematic diagram of the structure of some embodiments of a multi-modal event processing device according to the present disclosure;

[0016] Figure 3 It is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0017] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0018] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present disclosure can be combined with each other.

[0019] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0020] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0021] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0022] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0023] refer to Figure 1 , shows a process 100 of some embodiments of the multimodal event processing method according to the present disclosure. The multimodal event processing method comprises the following steps:

[0024] Step 101, in response to a first event processing platform corresponding to a first region receiving event query information from a target user for a first target event, querying from an event processing database whether there is event processing information corresponding to the first target event.

[0025] In some embodiments, in response to the first event processing platform corresponding to the first region receiving the event query information of the target user for the first target event, the execution subject (for example, the electronic device) of the above-mentioned multimodal event processing method can query whether there is event processing information corresponding to the above-mentioned first target event from the event processing database. Among them, the first region can be the region that receives the event query request. The first event processing platform can be a platform that supports various event processing. In practice, the event can be various types of events. For example, the event can be a consultation type event, a complaint type event, a suggestion type event, or a city transportation event. Various processing can include but is not limited to: storage processing, analysis processing, and deletion processing. The first event processing platform can be a platform that supports various users to handle events. For example, the first event processing platform can be a platform that supports querying information related to event handling. The target user can be a user who inputs an event query on the first event processing platform. The first target event can be an event object for the target user to query event processing information. The event query information can be a query request for querying event processing information corresponding to the first target event. The event processing database can be a database storing event data related to various events. In practice, the event processing database stores various events and corresponding event processing information. The event processing information may be the relevant processing information for executing event processing. Among them, the above-mentioned event processing information includes: event processing path and event processing evaluation information. The event processing path may be the processing path for event processing (that is, representing the processing flow corresponding to solving the first target event). The event processing evaluation information may be the user's evaluation information on the event processing information. In practice, the event processing evaluation information may be obtained by users filling in comments on the event processing information. The event processing evaluation information may characterize the quality of various event processing corresponding to the first target event. For example, the event processing evaluation information may be: the event processing information corresponding to the first target event is very accurate. The first event processing platform is a platform that is pre-deployed on the target server and has multiple event processing interfaces.

[0026] In some optional implementations of some embodiments, the execution subject may query from an event processing database whether there is event processing information corresponding to the first target event, including the following steps:

[0027] The first step is to determine the event category and event scope corresponding to the first target event. The event category may be the business category corresponding to the event corresponding to the first target event. For example, the event category may be, but is not limited to, at least one of the following: government event category, civil event category, bank event category, real estate event category. The event scope may be the event occurrence scope corresponding to the first target event. For example, the event occurrence scope may be the event occurrence location scope.

[0028] As an example, the execution subject may extract the event category and event range from the event query information.

[0029] The second step is to generate a first database query statement for the event category and the event range, wherein the first database query statement may be a SQL query statement.

[0030] As an example, the execution subject may use the event category and the event range as query elements and generate a first database query statement through a conventional SQL statement generation method.

[0031] The third step is to use the first database query statement to determine whether there is an event identical to the first target event in the event processing database.

[0032] As an example, the first database query statement is executed to determine whether there is an event identical to the target event in the event processing database.

[0033] In the fourth step, in response to determining that there is at least one corresponding candidate event processing information, determine the event keyword set corresponding to the first target event. The event keywords in the event keyword set may be keywords corresponding to the key content of the event extracted from the event query information. For example, if the first target event is a civil affairs event, the corresponding event keyword set may include: responsibility attribution, processing object, execution time, and civil affairs event identification.

[0034] The fifth step is to generate a second database query statement for the event keyword set, wherein the second database query statement may be a SQL query statement generated based on the event keyword set.

[0035] As an example, the execution entity may use each event keyword in the event keyword set as a query element corresponding to a query statement to generate a second database query statement.

[0036] The sixth step is to use the second database query statement to determine whether there is candidate event processing information corresponding to the first target event in the at least one candidate event processing information.

[0037] In the seventh step, in response to determining that the event does not exist, generate event processing information indicating that the event processing database does not contain the event corresponding to the first target event.

[0038] In some optional implementations of some embodiments, the event processing evaluation information is generated by the following steps:

[0039] The first step is to obtain a set of event processing comments on the first target event by each user on the first event processing platform within a predetermined time period. The predetermined time period may be a previous historical time period under the current time. For example, if the current time is September 1, the corresponding predetermined time period may be August 1-August 31. That is, the event processing comments are comments made by users on the first event processing platform on the event processing information corresponding to the first target event.

[0040] In the second step, at least one event processing comment is selected from the above-mentioned event processing comment set using the obtained comment screening regularization information. The comment screening regularization information may be a regularization method for selecting high-quality comments. In practice, the comment screening regularization information may be a method manually set by relevant technical personnel. The event processing comments in at least one event processing comment may be high-quality comment data. The comment screening regularization information may be written based on the Python language.

[0041] In the third step, for each event processing comment in the at least one event processing comment, the following generation steps are performed:

[0042] Sub-step 1: Determine a comment keyword set corresponding to the event handling comment, wherein the comment keyword set may be each keyword corresponding to the key comment content in the event handling comment.

[0043] As an example, the execution subject may determine the comment keyword set corresponding to the event processing comment through a keyword extraction model. In practice, the keyword extraction model may be a pre-trained neural network model for extracting keywords. In practice, the keyword extraction model may be a time series neural network model. The keyword extraction model may be trained based on a conventional model training method.

[0044] Sub-step 2, obtaining at least one evaluation direction information corresponding to the previous historical time period. The previous historical time period may be a historical time period before the predetermined time period and having a time period continuity relationship. The evaluation direction information may be a direction category under the evaluation direction of the event processing information by the user in the previous historical time period. The evaluation direction may represent the characteristic comments on the event processing information. For example, if the evaluation direction is the efficiency direction of the event processing information, the corresponding evaluation direction information may be that the efficiency of the processing flow corresponding to the event processing information is low, or that the efficiency of the processing flow corresponding to the event processing information is high.

[0045] Sub-step 3, determining the matching similarity between the evaluation keyword set and each evaluation direction information in the at least one evaluation direction information, and obtaining at least one matching similarity. The matching similarity can represent the content similarity between the evaluation content reflected by the evaluation keyword set and the evaluation content under the evaluation direction. The higher the matching similarity, the more similar the evaluation content under the evaluation direction is to the evaluation content reflected by the evaluation keyword set.

[0046] Sub-step 4: selecting at most two evaluation direction information with the highest corresponding matching similarity from the at least one evaluation direction information as an evaluation direction information group.

[0047] Sub-step 5, input the above-mentioned event processing comments into the evaluation question information generation model group corresponding to the above-mentioned evaluation direction information group to generate an evaluation text information group. Among them, there is a one-to-one correspondence between the evaluation direction information in the evaluation direction information group and the evaluation question information generation model in the evaluation question information generation model group. The evaluation question information generation model can be a neural network model for generating evaluation question information. There is a one-to-one correspondence between the evaluation text information in the evaluation text information group and the evaluation question information generation model in the evaluation question information generation model group. The evaluation text information can be evaluation question information in text form. For different evaluation directions, there are corresponding evaluation question information generation models. The evaluation question information can be the problems existing under the corresponding evaluation direction. In practice, the evaluation question information generation model can have different model structures for different evaluation directions. In practice, the corresponding model structure of the evaluation question information generation model can adopt an attention mechanism model based on text feature extraction. The corresponding training method of the evaluation question information generation model can be trained by a conventional model training method.

[0048] In the fourth step, each evaluation text information in the obtained at least one evaluation text information group is grouped according to the evaluation direction information to obtain an evaluation text information group set, wherein each evaluation text information group in the evaluation text information group set has the same evaluation direction.

[0049] The fifth step is to select a key evaluation text information group from each evaluation text information group in the evaluation text information group set according to the corresponding matching similarity group.

[0050] As an example, for each evaluation text information group, the execution subject may select evaluation text information with a matching similarity in the front target number from the evaluation text information group to obtain an evaluation text information candidate group. Then, semantic content duplication is removed for each evaluation text information in the evaluation text information candidate group to generate a key evaluation text information group.

[0051] The sixth step is to generate the above-mentioned event processing evaluation information based on the obtained key evaluation text information set.

[0052] As an example, the execution subject may directly determine the key evaluation text information group set as the event processing evaluation information.

[0053] As another example, the execution subject may select the key evaluation text information group with the highest number of corresponding text information from the key evaluation text information group set as the target key evaluation text information group, and determine the target key evaluation text information group as the event processing evaluation information.

[0054] Optionally, the execution subject may determine the matching similarity between the evaluation keyword set and each evaluation direction information in the at least one evaluation direction information, comprising the following steps:

[0055] The first step is to generate a keyword vector corresponding to each evaluation keyword in the evaluation keyword set to obtain a keyword vector set, wherein the keyword vector can represent the semantic content of the word corresponding to the evaluation keyword.

[0056] As an example, the above execution entity may input each evaluation keyword into a word vector generation model to generate a keyword vector and obtain a keyword vector set.

[0057] The second step is to determine at least one direction keyword corresponding to the above evaluation direction information.

[0058] As an example, the execution subject may determine at least one direction keyword corresponding to the evaluation direction information by querying a keyword database.

[0059] The third step is to generate a directional keyword vector corresponding to each directional keyword in the at least one directional keyword to obtain at least one directional keyword vector. The directional keyword vector can represent the word semantic content of the word semantic content corresponding to the directional keyword. The implementation method can refer to the generation of keyword vectors.

[0060] The fourth step is to concatenate the keyword vectors in the keyword vector set to generate a first concatenated vector.

[0061] In the fifth step, each keyword vector in the above keyword vector is input into the attention feature information generation model to generate the first keyword attention feature information. The attention feature information generation model may be an attention mechanism model of the keyword attention feature information. In practice, the attention feature information generation model may be an attention mechanism model of multiple key semantic feature information in advance. The key semantic feature information may represent the feature information of the key semantic content corresponding to each vector in the vector set. The key semantic feature information may be feature information in the form of vectors.

[0062] In the sixth step, each directional keyword vector in the at least one directional keyword vector is concatenated to generate a second concatenated vector.

[0063] The seventh step is to input each directional keyword vector in the at least one directional keyword vector into the attention feature information generation model to generate second keyword attention feature information.

[0064] In step 8, the first splicing vector and the first keyword attention feature information are spliced ​​in a predetermined order to obtain a third splicing vector. The predetermined order may be a predetermined splicing order. For example, the predetermined order may be a splicing order in which the first splicing vector is on the left and the first keyword attention feature information is on the right.

[0065] The ninth step is to splice the second splicing vector and the second keyword attention feature information in a predetermined order to obtain a fourth splicing vector.

[0066] Step 10: Generate vector similarity between the third splicing vector and the fourth splicing vector as the matching similarity, wherein the matching similarity may be cosine similarity.

[0067] Optionally, the attention feature information generation model can be trained by the following steps:

[0068] The first step is to obtain domain keyword sets corresponding to multiple fields. Each domain keyword set has a corresponding field. For example, the domain keyword set group may include: a domain keyword set corresponding to the real estate field, a domain keyword set corresponding to the medical field, and a domain keyword set corresponding to the government field.

[0069] In the second step, for each domain keyword set in the above domain keyword set group, the following generation steps are performed:

[0070] Sub-step 1: determine the domain keyword vector corresponding to each domain keyword in the domain keyword set to obtain the domain keyword vector set.

[0071] Sub-step 2: randomly extracting a random number of domain keywords from the above domain keyword vector set to generate a domain keyword vector group, thereby obtaining a domain keyword vector group set.

[0072] Sub-step 3, input each domain keyword vector group in the above-mentioned domain keyword vector group set into a pre-trained keyword semantic information generation model to generate keyword semantic information and obtain a keyword semantic information set. Among them, the keyword semantic information can characterize the common key semantic content between the corresponding domain keyword vectors in the domain keyword vector group. In practice, the common key semantic content can be the common features between the keywords in each domain. For example, the keywords in each domain include: fast software speed, low software call cost, and small software size. The corresponding common features are high software operation efficiency and high cost performance.

[0073] Sub-step 4, vectorizing each keyword semantic information in the above keyword semantic information set to generate an attention feature information set corresponding to the domain keyword vector group.

[0074] Sub-step 5, combining the domain keyword vector group and the attention feature information set to generate a training data subset.

[0075] The third step is to determine each obtained training data subset as a training data set.

[0076] The fourth step is to train the initial attention feature information generation model based on the above training data set to generate an attention feature information generation model.

[0077] As an example, the above-mentioned execution entity can perform model training on the initial attention feature information generation model based on the above-mentioned training data set by using the back propagation method to generate an attention feature information generation model.

[0078] As one of the invention points, another technical problem "how to match the similarity between the evaluation direction and the evaluation keyword set" is solved. Based on this, the present application combines the keyword attention feature information and direction feature information under the attention direction with the keyword feature information and direction feature information itself, so that the combined feature information includes not only the attention feature information but also the own feature information. Therefore, the vector similarity is determined based on the combined third splicing vector and the fourth splicing vector to accurately generate the matching similarity.

[0079] Step 102: In response to determining that there is no such event, determine from the event processing database at least one event whose corresponding event content similarity with the first target event is greater than the target similarity.

[0080] In some embodiments, in response to determining that it does not exist, the execution subject may determine from the event processing database at least one event whose corresponding event content similarity with the first target event is greater than the target similarity. The event content may include: the processing content corresponding to the event and the event domain information involved. The target similarity may be a pre-set similarity value. The similarity between the event corresponding event content of the at least one event and the first target event content is greater than the target similarity.

[0081] As an example, first, determine the processing content and event domain information corresponding to the first target event, as the first processing content and the first event domain information, respectively. Then, query at least one initial event that matches the first event domain information from the above event processing database. Then, determine the content similarity between the processing content corresponding to the above at least one initial event and the first processing content (specifically, it can be determined by extracting text feature information through a temporal neural network), and obtain at least one content similarity. Finally, filter out at least one event whose corresponding content similarity is greater than the target similarity from at least one initial event.

[0082] In some optional implementations of some embodiments, after step 102, the steps further include:

[0083] The execution subject may send the event content and event query information corresponding to the first target event to a second event processing platform corresponding to at least one second region. The second region has a regional jurisdiction relationship or a regional proximity relationship with the first region. Each second region has a corresponding second event processing platform. For example, the first region is Tangshan City, Hebei Province, and the corresponding second region may be Shijiazhuang City, Hebei Province, or Tianjin City.

[0084] Step 103: Determine at least one event description position of the first target event in at least one specification file based on the at least one event.

[0085] In some embodiments, the above-mentioned execution subject can determine at least one event description position of the above-mentioned first target event in at least one specification file based on the above-mentioned at least one event. Among them, the at least one specification file can be at least one standardized file for event handling in various fields specified by the first region. For example, at least one specification file may include: a list of rights and responsibilities, a three-determination plan file, and a national standard specification basis file. The event description position may be the file position where the first target event is described in a specification in the corresponding specification file. That is, the corresponding position area of ​​the event description position may be a processing specification requirement corresponding to the first target event that is described in detail. The specification file in at least one specification file may have a one-to-one correspondence with at least one event description position, or there may be at least one event description position corresponding to a specification file, or there may be no corresponding event description position for a specification file.

[0086] As an example, first, for each event in at least one event, an event description position set of the event in at least one specification file is determined. Then, each event description position in the obtained at least one event description position set is deduplicated to generate a deduplicated event description position set. Finally, the deduplicated event description position set is determined as at least one event description position.

[0087] Step 104: extracting event specification information corresponding to the first target event from the at least one specification file according to the at least one event description position.

[0088] In some embodiments, the execution subject may extract event specification information corresponding to the first target event from the at least one specification file according to the at least one event description location, wherein the event specification information may be specification information corresponding to the first target event that indicates how to perform event processing.

[0089] As an example, first, for each event description position in at least one event description position, the first step is to determine the event description content of the above event description position on the corresponding specification file. The second step is to determine whether the similarity between the event description content and the event content corresponding to the target event is greater than a predetermined similarity. The third step is to determine the event description content as the event specification sub-information corresponding to the first target event when the similarity is determined to be greater than the predetermined similarity. Finally, the obtained event specification sub-information is integrated to generate event specification information.

[0090] Step 105, based on the above event specification information, using a pre-trained event processing information generation model, generates first event processing information, a first event processing image and a first event processing audio.

[0091] In some embodiments, the above-mentioned execution subject can generate the first event processing information, the first event processing image and the first event processing audio according to the above-mentioned event specification information using a pre-trained event processing information generation model. Among them, the event processing information generation model can be a neural network model that generates information related to event processing. The first event processing information can be a textual representation of the processing flow corresponding to the first target event. The first event processing image can be an image in graphic form that represents how the first target event is processed. The first event processing audio can be an audio in audio form that represents how the first target event is processed. In practice, the event processing information generation model can be a large language model based on the Transformer model.

[0092] As an example, first, the execution subject may generate first event processing prompt information in text form corresponding to the first event processing information, second event processing prompt information in text form corresponding to the first event processing image, and third event processing prompt information in text form corresponding to the first event processing audio. The first event processing prompt information may be: "Generate the first event processing information corresponding to the first target event according to the event specification information, and the first event processing information matches the first event processing image and the first event processing audio corresponding content". The second event processing prompt information may be: "Generate the first event processing image corresponding to the first target event according to the event specification information, and the first event processing image matches the first event processing information and the first event processing audio corresponding content". The third event processing prompt information may be: "Generate the first event processing audio corresponding to the first target event according to the event specification information, and the first event processing audio matches the first event processing information and the first event processing image corresponding content". Then, the event specification information, the first event processing prompt information, the second event processing prompt information, and the third event processing prompt information are input into the event processing information generation model to generate the first event processing information, the first event processing image, and the first event processing audio.

[0093] In some optional implementations of some embodiments, after step 105, the steps further include:

[0094] The first step is, in response to determining that feedback information for the above-mentioned first target event sent by at least one second event processing platform is received, and based on the above-mentioned feedback information, the above-mentioned first event processing information, the above-mentioned first event processing image and the above-mentioned first event processing audio are subjected to event processing logic verification. Among them, the feedback information can be the third event processing information, the third event processing image and the third event processing audio for the first target event stored by each second event processing platform. The feedback information may include: at least one event feedback information corresponding to at least one second event processing platform. The event feedback information includes: the third event processing information, the third event processing image and the third event processing audio corresponding to the second event processing platform.

[0095] As an example, for each event feedback information in at least one event feedback information, first, determine the first information gap between the third event processing information and the first event processing information, the second information difference between the third event processing image and the first event processing image, and the third information difference between the third event processing audio and the first event processing audio. Then, in response to determining that the first information difference is less than the first predetermined difference value, the second information difference is less than the second predetermined difference value, and the third information difference is less than the third predetermined difference value, generate first information representing that the verification has passed. In response to determining that the first information difference is greater than or equal to the first predetermined difference value, the second information difference is greater than or equal to the second predetermined difference value, and / or the third information difference is greater than or equal to the third predetermined difference value, generate second information representing that the verification has passed. Finally, in response to determining that the number of information corresponding to the first information divided by the number of information corresponding to the second information is greater than or equal to the target value, generate information representing that the event processing logic verification has passed. In response to determining that the number of information corresponding to the first information divided by the number of information corresponding to the second information is less than the target value, generate information representing that the event processing logic verification has not passed.

[0096] In the second step, in response to determining that the event processing logic has passed the verification, continue to execute the publishing step for the above-mentioned first event processing information, the above-mentioned first event processing image and the above-mentioned first event processing audio.

[0097] Step 106: publish the first event processing information, the first event processing image and the first event processing audio to the first event processing platform.

[0098] In some embodiments, the execution entity may publish the first event processing information, the first event processing image, and the first event processing audio to the first event processing platform.

[0099] Step 107, in response to determining that the target user clicks on the event processing control in the event processing page corresponding to the first event processing platform, a text display pop-up window corresponding to the first event processing information is popped up, as well as an image control corresponding to the first event processing image and an audio control corresponding to the first event processing audio are displayed.

[0100] In some embodiments, in response to determining that the target user clicks on the event processing control in the event processing page corresponding to the first event processing platform, the execution subject may pop up a text display pop-up window corresponding to the first event processing information, as well as an image control corresponding to the first event processing image and an audio control corresponding to the first event processing audio. Among them, the event processing page may be a page for processing an event. Specifically, in an event query scenario, the event processing page may be an event query page, that is, a page for querying event-related event processing information. The event processing control may be a page control representing the query event processing information. The text display pop-up window may be a pop-up window that displays the text content corresponding to the first event processing information. The image control may be a page control representing the query of the first event processing image. Here, by clicking the image control, the corresponding first event processing image may pop up. The audio control may be a page control representing the query of the first event processing audio. Here, by clicking the audio control, the first event processing audio may be played.

[0101] In some optional implementations of some embodiments, the event processing evaluation information is generated based on the event processing comments corresponding to the first target event.

[0102] Optionally, after step 107, the steps further include:

[0103] In the first step, in response to the determination of existence, the event processing evaluation information is input into the pre-selected and trained evaluation feature information extraction model to generate evaluation content under each event feature dimension. Among them, the evaluation feature information generation model can be a neural network model that generates evaluation content under each event feature dimension. The event feature dimension can represent each feature direction of the event evaluation. For example, for the first target event being a real estate event, the corresponding event feature dimension can be one of the following: event processing speed, event processing quality, and user evaluation star number. The evaluation content can be in the form of a label that can represent the degree of feature evaluation under the event feature dimension. The evaluation feature information generation model can include: a multi-layer series convolution layer and a classification layer corresponding to each event feature dimension. That is, for each event feature dimension, it includes: a first event feature dimension, a second event feature dimension, and a third event feature dimension. The corresponding classification layers include: a classification layer corresponding to the first event feature dimension, a classification layer corresponding to the second event feature dimension, and a classification layer corresponding to the third event feature dimension. Each classification layer is a classification layer connected in parallel to the multi-layer series convolution layer. The evaluation feature information extraction model can be trained by a conventional model training method. For example, the evaluation feature information extraction model can be trained by back propagation and gradient descent.

[0104] In the second step, in response to determining that each evaluation content meets the feature requirements under the corresponding event feature dimension, a pop-up link corresponding to the above event processing information is sent to a user terminal corresponding to the above target user.

[0105] In some optional implementations of some embodiments, after step 107, the steps further include:

[0106] The first step is to obtain the second event processing information, the second event processing image and the second event processing audio corresponding to the second target event in response to receiving the event processing verification information for the second target event in the event processing database. The second target event may be an event to be subjected to event processing verification. The event processing verification information may be request information for performing content verification on the event processing information, the event processing image and the event processing audio of the second target event.

[0107] The second step is to generate text processing prompt information corresponding to the second event processing information, image processing prompt information corresponding to the second event processing image, and audio processing prompt information corresponding to the second event processing audio. Among them, the text processing prompt information includes: information representing whether the event processing logic corresponding to the second event processing information is reasonable, and text reason information representing whether the output event processing logic is reasonable. The text reason information can be reason information in text form. The image processing prompt information includes: information representing whether the image content logic corresponding to the second event processing image is reasonable, and text reason information representing whether the output image content logic is reasonable. The audio processing prompt information includes: information representing whether the audio content logic corresponding to the second event processing audio is reasonable, and audio reason information representing whether the output audio content logic is reasonable.

[0108] The third step is to input the second event processing information, the second event processing image, the second event processing audio, the text processing prompt information, the image processing prompt information and the audio processing prompt information into a large language model for processing information logic verification to generate logic verification information corresponding to the second target event. The fourth step is to respond to the determination that the logic verification information indicates that there is a logic problem in the second event processing information, the second event processing image and the second event processing audio, and adjust the second event processing information, the second event processing image and the second event processing audio accordingly according to the output text reason information, audio reason information and image reason information to generate adjusted event processing information, adjusted event processing image and adjusted event processing audio.

[0109] As an example, first, the text reason information is input into the first semantic extraction model to generate the first reason feature information. Then, the audio reason information is input into the second semantic extraction model to generate the second reason feature information. Next, the image reason information is input into the third semantic extraction model to generate the third reason feature information. Furthermore, the first reason feature information, the second reason feature information and the third reason feature information are subjected to feature information fusion to generate fusion feature information. Next, the second event processing information and the fusion feature information are input into the generative model for generating event processing information to generate the adjusted event processing information. The second event processing image and the fusion feature information are input into the generative model for generating the event processing image to generate the adjusted event processing image. The second event processing audio and the fusion feature information are input into the generative model for generating the event processing audio to generate the adjusted event processing audio. The first semantic extraction model, the second semantic extraction model and the third semantic extraction model can be text semantic feature extraction models.

[0110] The fifth step is to replace the second event processing information, the second event processing image, and the second event processing audio corresponding to the second target event in the above event processing database with the above adjusted event processing information, the above adjusted event processing image, and the above adjusted event processing audio.

[0111] The above-mentioned various embodiments of the present disclosure have the following beneficial effects: through the event processing method based on multimodality of some embodiments of the present disclosure, the accurately generated event processing information, event processing images and event processing audio can be timely published to the event processing platform to provide an event processing method under multimodality. Specifically, the reason why the relevant event processing is not efficient and accurate is that the judgment may be inaccurate and inefficient due to differences in personal experience and subjective judgment, and there is also a lack of clear definition of whether the event has the conditions for handling. Based on this, the event processing method based on multimodality of some embodiments of the present disclosure, first, in response to the first event processing platform corresponding to the first region receiving the event query information of the target user for the first target event, querying from the event processing database whether there is event processing information corresponding to the above-mentioned first target event, wherein the above-mentioned event processing information includes: event processing path and event processing evaluation information. Here, after receiving the event query information of the first target event, it is first preliminarily determined by querying the event processing database whether there is corresponding event processing information, so as to improve the processing efficiency of the first target event. Then, in response to determining that it does not exist, at least one event whose corresponding event content similarity with the above-mentioned first target event is greater than the target similarity is determined from the above-mentioned event processing database. Here, by determining at least one similar event, the specification content related to the first target event is subsequently determined to facilitate determining how the first target event should be processed. Then, based on the above at least one event, at least one event description position of the above first target event in at least one specification file can be accurately determined. Then, based on the above at least one event description position, the event specification information corresponding to the above first target event is extracted from the above at least one specification file to determine the processing specification requirements related to the first target event. Furthermore, based on the above event specification information, using the pre-trained event processing information generation model, the first event processing information, the first event processing image and the first event processing audio under multimodality can be accurately generated, so as to facilitate the subsequent multimodal display of the processing results of the first target event and increase the diversity of the display. Further, the above first event processing information, the above first event processing image and the above first event processing audio are published to the above first event processing platform so as to facilitate feedback to the target user to view the corresponding event processing information. Finally, in response to determining that the target user clicks on the event processing control in the event processing page corresponding to the first event processing platform, a text display pop-up window corresponding to the first event processing information pops up, as well as an image control corresponding to the first event processing image and an audio control corresponding to the first event processing audio, so that the target user can view the processing status of the first target event in a multimodal manner.In summary, through the method of querying the event processing database and determining at least one similar event, as well as the use of the event processing information generation model, accurately generated event processing information, event processing images and event processing audio can be published to the event processing platform in a timely manner to provide a multi-modal event processing method.

[0112] Further references Figure 2 As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a multi-modal event processing device. These device embodiments are similar to Figure 1 Corresponding to the method embodiments shown, the multimodal-based event processing device can be specifically applied to various electronic devices.

[0113] like Figure 2 As shown, a multimodal event processing device 200 includes: a query unit 201, a first determination unit 202, a second determination unit 203, an extraction unit 204, a generation unit 205, a publishing unit 206 and a display unit 207. The query unit 201 is configured to query whether there is event processing information corresponding to the first target event from the event processing database in response to the first event processing platform corresponding to the first region receiving event query information from the target user for the first target event, wherein the event processing information includes: event processing path and event processing evaluation information; the first determination unit 202 is configured to determine at least one event whose corresponding event content similarity with the first target event is greater than the target similarity from the event processing database in response to determining that there is no event; the second determination unit 203 is configured to determine at least one event description position of the first target event in at least one specification file based on the at least one event; the extraction unit 204 is configured to, based on the at least one event description position, Event specification information corresponding to the above-mentioned first target event is extracted from the above-mentioned at least one specification file; the generation unit 205 is configured to generate first event processing information, a first event processing image and a first event processing audio according to the above-mentioned event specification information using a pre-trained event processing information generation model; the publishing unit 206 is configured to publish the above-mentioned first event processing information, the above-mentioned first event processing image and the above-mentioned first event processing audio to the above-mentioned first event processing platform; the display unit 207 is configured to respond to determining that the above-mentioned target user clicks on the event processing control in the event processing page corresponding to the above-mentioned first event processing platform, pop up a text display pop-up window corresponding to the above-mentioned first event processing information, and display an image control corresponding to the above-mentioned first event processing image and an audio control corresponding to the first event processing audio.

[0114] It can be understood that the units recorded in the multi-modal event processing device 200 are similar to those in the reference Figure 1Therefore, the operations, features and beneficial effects described above for the method are also applicable to the multimodal event processing device 200 and the units contained therein, and will not be described in detail here.

[0115] Reference below Figure 3 , which shows a structural schematic diagram of an electronic device (eg, an electronic device) 300 suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0116] like Figure 3 As shown, the electronic device 300 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 to a random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 are also stored. The processing device 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0117] Typically, the following devices may be connected to the I / O interface 305: an input device 306 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 308 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 309. The communication device 309 may allow the electronic device 300 to communicate with other devices wirelessly or by wire to exchange data. Figure 3 The electronic device 300 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead. Figure 3 Each block shown in the figure may represent one device, or may represent multiple devices as required.

[0118] In particular, according to some embodiments of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from the network through the communication device 309, or installed from the storage device 308, or installed from the ROM 302. When the computer program is executed by the processing device 301, the above-mentioned functions defined in the method of some embodiments of the present disclosure are executed.

[0119] It should be noted that the computer-readable medium in some embodiments of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device. In some embodiments of the present disclosure, the computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than computer readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0120] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0121] The above-mentioned computer-readable medium may be included in the above-mentioned electronic device; or it may exist independently without being installed in the electronic device. The above-mentioned computer-readable medium carries one or more programs. When the above-mentioned one or more programs are executed by the electronic device, the electronic device: in response to the first event processing platform corresponding to the first region receiving event query information from the target user for the first target event, inquire from the event processing database whether there is event processing information corresponding to the above-mentioned first target event, wherein the above-mentioned event processing information includes: event processing path and event processing evaluation information; in response to determining that it does not exist, determine from the above-mentioned event processing database at least one event whose corresponding event content similarity with the above-mentioned first target event is greater than the target similarity; based on the above-mentioned at least one event, determine at least one event description position of the above-mentioned first target event in at least one specification file. ; According to the above-mentioned at least one event description location, extract the event specification information corresponding to the above-mentioned first target event from the above-mentioned at least one specification file; according to the above-mentioned event specification information, use the pre-trained event processing information generation model to generate the first event processing information, the first event processing image and the first event processing audio; publish the above-mentioned first event processing information, the above-mentioned first event processing image and the above-mentioned first event processing audio to the above-mentioned first event processing platform; in response to determining that the above-mentioned target user clicks on the event processing control in the event processing page corresponding to the above-mentioned first event processing platform, pop up a text display pop-up window corresponding to the above-mentioned first event processing information, and display the image control corresponding to the above-mentioned first event processing image and the audio control corresponding to the first event processing audio.

[0122] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0123] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0124] The units described in some embodiments of the present disclosure may be implemented by software or by hardware. The described units may also be set in a processor, for example, may be described as: a processor includes a query unit, a first determination unit, a second determination unit, an extraction unit, a generation unit, a publishing unit and a display unit. Among them, the names of these units do not constitute a limitation on the unit itself under certain circumstances. For example, the second determination unit may also be described as "a unit that determines, based on the at least one event, the position of at least one event description of the above-mentioned first target event in at least one specification file."

[0125] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0126] The above descriptions are only some preferred embodiments of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the above features are replaced with (but not limited to) technical features with similar functions disclosed in the embodiments of the present disclosure.

Claims

1. A multimodal event processing method, comprising: In response to a first event processing platform corresponding to the first region receiving event query information for a first target event from a target user, querying from an event processing database whether there is event processing information corresponding to the first target event, wherein the event processing information includes: an event processing path and event processing evaluation information; In response to determining that there is no such event, determining from the event processing database at least one event whose corresponding event content similarity with the first target event is greater than a target similarity; Sending the event content and event query information corresponding to the first target event to a second event processing platform corresponding to at least one second region, wherein the second region has a regional jurisdiction relationship or a regional proximity relationship with the first region; Determine, based on the at least one event, at least one event description position of the first target event in at least one specification file; extracting event specification information corresponding to the first target event from the at least one specification file according to the at least one event description position; According to the event specification information, using a pre-trained event processing information generation model, generate first event processing information, a first event processing image, and a first event processing audio; In response to determining that feedback information regarding the first target event sent by at least one second event processing platform is received, performing event processing logic verification on the first event processing information, the first event processing image, and the first event processing audio according to the feedback information; In response to determining that the event processing logic passes the verification, continuing to execute the publishing step for the first event processing information, the first event processing image, and the first event processing audio; Publishing the first event processing information, the first event processing image, and the first event processing audio to the first event processing platform; In response to determining that the target user clicks on the event processing control in the event processing page corresponding to the first event processing platform, a text display pop-up window corresponding to the first event processing information pops up, as well as an image control corresponding to the first event processing image and an audio control corresponding to the first event processing audio.

2. The method according to claim 1, wherein: The querying from the event processing database whether there is event processing information corresponding to the first target event includes: Determine an event category and event range corresponding to the first target event; Generate a first database query statement for the event category and the event range; Using the first database query statement, determining whether there is an event identical to the first target event in the event processing database; In response to determining that there is corresponding at least one candidate event processing information, determining an event keyword set corresponding to the first target event; generating a second database query statement for the event keyword set; Using the second database query statement, determining whether there is candidate event processing information corresponding to the first target event in the at least one candidate event processing information; In response to determining that the first target event does not exist, event processing information indicating that the first target event does not exist in the event processing database is generated.

3. The method according to claim 1, wherein: The event processing evaluation information is generated based on the event processing comments corresponding to the first target event; as well as The method further comprises: In response to determining existence, inputting the event processing evaluation information into a pre-selected and trained evaluation feature information extraction model to generate evaluation content under various event feature dimensions; In response to determining that each evaluation content satisfies the feature requirements under the corresponding event feature dimension, a pop-up link corresponding to the event processing information is sent to a user terminal corresponding to the target user.

4. The method according to claim 1, wherein: The method further comprises: In response to receiving event processing verification information for a second target event in the event processing database, obtaining second event processing information, a second event processing image, and a second event processing audio corresponding to the second target event; Generate text processing prompt information corresponding to the second event processing information, image processing prompt information corresponding to the second event processing image, and audio processing prompt information corresponding to the second event processing audio, wherein the text processing prompt information includes: information characterizing whether the event processing logic corresponding to the second event processing information is reasonable, and text reason information characterizing whether the output event processing logic is reasonable; the image processing prompt information includes: information characterizing whether the image content logic corresponding to the second event processing image is reasonable, and text reason information characterizing whether the output image content logic is reasonable; the audio processing prompt information includes: information characterizing whether the audio content logic corresponding to the second event processing audio is reasonable, and audio reason information characterizing whether the output audio content logic is reasonable; Inputting the second event processing information, the second event processing image, the second event processing audio, the text processing prompt information, the image processing prompt information, and the audio processing prompt information into a large language model for processing information logic verification to generate logic verification information corresponding to the second target event; In response to determining that the logic check information indicates that there is a logic problem in at least one of the second event processing information, the second event processing image, and the second event processing audio, according to the output text reason information, audio reason information, and image reason information, the second event processing information, the second event processing image, and the second event processing audio are correspondingly adjusted to generate adjusted event processing information, adjusted event processing image, and adjusted event processing audio; The second event processing information, the second event processing image, and the second event processing audio corresponding to the second target event in the event processing database are replaced with the adjusted event processing information, the adjusted event processing image, and the adjusted event processing audio.

5. The method according to claim 3, wherein: The event processing evaluation information is generated by the following steps: Acquire a set of event processing comments on the first target event by each user on the first event processing platform within a predetermined time period; Using the obtained comment filtering regularization information, at least one event processing comment is filtered out from the event processing comment set; For each event handling comment in the at least one event handling comment, the following generating steps are performed: Determine a comment keyword set corresponding to the event handling comment; Obtain at least one evaluation direction information corresponding to the previous historical time period; Determine a matching similarity between the comment keyword set and each evaluation direction information in the at least one evaluation direction information to obtain at least one matching similarity; Filter out at most two evaluation direction information with the highest corresponding matching similarity from the at least one evaluation direction information as an evaluation direction information group; Inputting the event processing comments into the evaluation question information generation model group corresponding to the evaluation direction information group to generate an evaluation text information group; Grouping each evaluation text information in the obtained at least one evaluation text information group according to the evaluation direction information to obtain an evaluation text information group set; For each evaluation text information group in the evaluation text information group set, according to the corresponding matching similarity group, a key evaluation text information group is screened out from the evaluation text information group; The event processing evaluation information is generated according to the obtained key evaluation text information set.

6. A multimodal event processing device, comprising: A query unit is configured to query whether there is event processing information corresponding to the first target event from an event processing database in response to a first event processing platform corresponding to a first region receiving event query information from a target user for a first target event, wherein the event processing information includes: an event processing path and event processing evaluation information; A first determining unit is configured to, in response to determining that there is no event, determine from the event processing database at least one event whose corresponding event content similarity with the first target event is greater than a target similarity; The device further includes: sending the event content and event query information corresponding to the first target event to a second event processing platform corresponding to at least one second region, wherein the second region has a regional jurisdiction relationship or a regional proximity relationship with the first region; A second determining unit is configured to determine at least one event description position of the first target event in at least one specification file according to the at least one event; an extracting unit configured to extract event specification information corresponding to the first target event from the at least one specification file according to the at least one event description position; A generating unit configured to generate first event processing information, a first event processing image, and a first event processing audio according to the event specification information and using a pre-trained event processing information generating model; The device further includes: in response to determining that feedback information regarding the first target event sent by at least one second event processing platform is received, performing event processing logic verification on the first event processing information, the first event processing image, and the first event processing audio according to the feedback information; in response to determining that the event processing logic passes the verification, continuing to execute the publishing step for the first event processing information, the first event processing image, and the first event processing audio; a publishing unit, configured to publish the first event processing information, the first event processing image, and the first event processing audio to the first event processing platform; The display unit is configured to pop up a text display window corresponding to the first event processing information in response to determining that the target user clicks on the event processing control in the event processing page corresponding to the first event processing platform, and to display an image control corresponding to the first event processing image and an audio control corresponding to the first event processing audio.

7. An electronic device comprising: one or more processors; a storage device having one or more programs stored thereon, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 5.

8. A computer readable medium having a computer program stored thereon, wherein: When the program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images