Intelligent compilation system, method and program product for crisis management

Abstract

The present application belongs to the technical field of information processing, and particularly relates to an intelligent publishing system, method and program product for emergency management. The data acquisition module acquires reported emergencies and stored historical publication data. For the historical publication data, the characteristic information is extracted, the format structure in the characteristic information and the correlation between the format structures are programmed, so that the publication template can be generated. Then, the heat of the emergency is calculated. For the listed events as the generated publication events, the hot list events are processed. Based on the big data language model as the model for automatic generation of publication content, finally, the hot list events are input into the trained big data language model, and the required publication is directly output. The present application can solve the problem of low efficiency in the existing digital system for publication production.

Description

Technical Field

[0001] This invention belongs to the field of information processing technology, and in particular relates to intelligent editing systems, methods and program products for emergency management. Background Technology

[0002] To coordinate and promote the digital construction of the city, facilitate the interconnection of information systems and networks, the sharing of data on demand, and the efficient collaboration of business, an incident handling system was developed to integrate digital concepts, thinking, skills and qualities into all aspects of the performance of duties by urban units.

[0003] The event handling system differs from previous methods in its various functions. For example, the production and distribution of publications have been transformed from traditional paper printing and promotion of articles, books, and newspapers into online digital entities, such as WeChat official account pushes. Therefore, the application of the event handling system has brought great convenience to urban units.

[0004] The existing incident handling system typically relies on editors to edit and produce publications based on information reported by the reporting personnel, which is inefficient. Furthermore, for cases where a large amount of reported information needs to be compiled into publications for record-keeping, the large volume and long editing time often lead to an overload of editors' workload, affecting the smooth operation of the incident handling system. Summary of the Invention

[0005] The technical problem solved by this invention is to provide an intelligent journal editing system, method and program product for emergency management, so as to solve the problem of low efficiency in journal production in existing event handling systems.

[0006] The basic solution provided by this invention is an intelligent editing system for emergency management, comprising:

[0007] Data acquisition module: used to acquire emergency reports from personnel from the incident handling system, as well as historical publication data stored in the incident handling system;

[0008] Publication template generation module: used to filter the feature information of historical publication data, programmatically process the format structure of the feature information and the relationship between the format structures, and generate publication templates;

[0009] Event Hot List Generation Module: Used to generate an event hot list based on the acquired breaking events according to the preset hot list rules, and to use the events on the event hot list as the events for publication.

[0010] Emergency handling module: used to extract the event structure and event content of the journal's events, and to integrate the extracted event structure and event content according to the journal template to generate a new event structure and corresponding event content that corresponds to the journal template;

[0011] Event optimization module: Construct a big data language model, use historical publication data and publication templates as training samples to train the big data language model, and after training, input new event structures and corresponding event content into the big data language model according to the event hot list as priority;

[0012] Intelligent editing module: The big data language model optimizes the corresponding event content based on the new event structure, and the optimized event content and the new event structure generate the publication.

[0013] Furthermore, the data acquisition module includes an emergency event acquisition unit and a historical data storage unit. The emergency event acquisition unit acquires emergency events through multiple methods, and the historical data storage unit is used to store historical publication data.

[0014] Furthermore, in the publication template generation module, the characteristic information of historical publication data includes the type, layout, format, and format content of the historical publication;

[0015] The specific steps of programmatically processing the format structure of feature information and the correlation between format structures are as follows:

[0016] The structure and format content of the format for extracting feature information from historical publication data;

[0017] The importance of the format structure is calculated based on the layout of historical publications;

[0018] Relate the format structures according to their format content and importance.

[0019] Furthermore, the hot list generation module includes a rule-making unit for listing events, an event risk calculation unit, and a hot list ranking unit. The rule-making unit is used to formulate rules for listing events and to filter sudden events according to the rules for listing events. The event risk calculation unit is used to calculate the risk value of the filtered sudden events through an event risk calculation model. The hot list ranking unit is used to rank the sudden events according to the risk value and take the top N sudden events as the events to be published.

[0020] Furthermore, the expression for the event risk calculation model is as follows:

[0021]

[0022] Where Sa represents the initial score calculation, Normalize represents the normalization process, and W i The keyword indicating the risk of a hit, P i N represents the weight of the keyword. j The key number representing the risk of a hit, P j P represents the weight of the key number.r S represents the rule weight. m P represents the AI ​​model score. m P represents the weights of the AI ​​model. el S represents the float range corresponding to the event level. el S represents the base score corresponding to the event level. ul The score indicates the level of urgency.

[0023] Furthermore, the specific steps of formulating rules for listing events and screening emergencies based on these rules are as follows:

[0024] Screen for emergencies and remove those that are not included in the advanced management interface of the incident handling system;

[0025] If a reported emergency is processed by the advanced management interface of the event handling system, it will be automatically added to the list.

[0026] The event level of reported emergencies is determined, and those that fall within the preset event levels are automatically included in the list.

[0027] The system assesses the urgency of reported emergencies and automatically adds them to the list if they fall within a preset urgency level.

[0028] Set up daily, weekly, and monthly rankings as needed, and set a preset number of entries based on the timeliness of the daily, weekly, and monthly rankings.

[0029] Furthermore, the emergency handling module includes an event extraction unit and an integration unit. The event extraction unit is used to extract the event structure and event content of the journal editing event. The integration unit is used to associate the extracted event structure with the journal template so that the event structure and its event content correspond to the journal template.

[0030] Furthermore, the event optimization module includes a big data language model construction unit and a hot topic event preprocessing unit. The big data language model construction unit is used to construct a big data language model and use historical publication data and publication templates as training samples for format training of the big data language model. During the training process, it communicates with the Internet and uses Internet publication data as content beautification training samples to train the publication content.

[0031] The hot list event preprocessing unit is used to generate a big data language model processing order for hot list events according to their hot list ranking.

[0032] It also includes an auditing module, which is used to audit the generated publications by human or artificial intelligence.

[0033] The intelligent editing method for emergency management, applied to the aforementioned intelligent editing system for emergency management, includes:

[0034] Based on the incident handling system, we obtain the emergency incidents reported by personnel, as well as the historical publication data stored in the incident handling system;

[0035] The feature information of historical publications is filtered, and the format structure of the feature information and the relationship between the format structures are programmed to generate publication templates.

[0036] The acquired breaking news will be used to generate a hot list of events according to the preset hot list rules, and the events on the hot list will be used as events for publication.

[0037] Extract the event structure and content of the editing events, and integrate the extracted event structure and content according to the publication template to generate a new event structure and corresponding event content that correspond to the publication template.

[0038] A big data language model is constructed by using historical publication data and publication templates as training samples. After training, new event structures and corresponding event content are input into the big data language model based on the event hot list as the priority.

[0039] The big data language model optimizes the corresponding event content based on the new event structure, and the optimized event content and the new event structure generate the publication.

[0040] A program product includes a computer program or instructions that, when executed by a processor, implement the intelligent editing method for emergency management as described above.

[0041] The principle and advantages of this invention are as follows: In this application, for sudden events reported by users from the event processing system, how to automatically generate publications and select which events to generate publications, this application first obtains the reported sudden events and stored historical publication data through the data acquisition module. For the historical publication data, its feature information is extracted, and the format structure and the correlation between the format structures in the feature information are programmed to generate publication templates. The programmed nature facilitates subsequent processing by the big data language model. Then, the popularity of the sudden events is calculated, with the popularity based on the event risk value. The events that are on the list are used as the events for generating publications. Next, the hot events are processed, including event structure processing and event content processing. Based on the big data language model, the model is used to automatically generate publication content. During the construction process, the big data language model is initially trained through publication templates and historical publication data, so that the big data language model has the basic recognition ability of publication structure and publication content in the event processing system. Then, the publication content is retrained through the Internet, so that the model has good text content beautification ability. Finally, the hot events are input into the trained big data language model, and the required publications are directly output.

[0042] Therefore, the advantage of this application is that it adopts an automatic screening and generation mode for the publications of the event processing system, and the content of the generated publications is highly correlated with the publication types, structures and contents of the event processing system, making it suitable for automatic publication editing in the event processing system. Attached Figure Description

[0043] Figure 1 This is a functional block diagram of an embodiment of the present invention;

[0044] Figure 2 This is a flowchart of an embodiment of the present invention. Detailed Implementation

[0045] The following detailed description illustrates the specific implementation method:

[0046] The basic implementation examples are as follows: Figure 1 As shown: An intelligent journal editing system for emergency management includes a data acquisition module, a journal template generation module, an event hot list generation module, an emergency handling module, an event optimization module, an intelligent journal editing module, and an approval module, among which:

[0047] The data acquisition module is used to acquire personnel-reported emergencies from the incident handling system, as well as historical publication data stored in the incident handling system.

[0048] In this application, the data acquisition module includes an emergency event acquisition unit and a historical data storage unit. The emergency event acquisition unit is used to acquire emergency events in multiple ways, such as through voice input, convert them into text, and extract the corresponding content according to the structure of the emergency event and place it in the corresponding structure.

[0049] The historical data storage unit is used to store historical publication data in the event processing system. Specifically, the historical publication data is stored digitally. For example, previous print publications are stored in the historical data storage unit as images, and previous digital news can also be stored directly in the historical data storage unit.

[0050] The publication template generation module is used to filter the feature information of historical publication data, programmatically process the format structure of the feature information and the correlation between the format structures, and generate publication templates. In this embodiment, the feature information of historical publication data includes the type, layout, format and content of historical publications. Specifically, the types of historical publications include news, criminal, and livelihood publications; the layout of historical publications includes text layout, image layout, etc.; and the format of historical publications includes title type, title format, content format, etc.

[0051] The specific steps for programmatically implementing the format structure of feature information and the relationships between format structures are as follows:

[0052] First, the format structure and content of characteristic information are extracted from historical publication data. Then, the importance of the format structure is calculated according to the layout of the historical publications. The format structures are then correlated according to the format content and importance. Specifically, the importance of each section is calculated by analyzing the layout of the historical publications. The ratio of the format structure in each section is obtained by the content proportion, thereby generating the importance. The importance of the format structures in each section is compared, and the format structures with the comparison results within a preset range are correlated. This ensures that when generating new content corresponding to a particular format structure, its richness is determined by its importance, and the associated format structures do not need to have their importance calculated again. Instead, content of the same level is generated based on correlation.

[0053] The hot topic generation module is used to generate a hot topic list of acquired emergencies according to preset hot topic rules, and to use the events on the hot topic list as editorial events. The hot topic generation module includes a rule-making unit for inclusion in the list, an event risk calculation unit, and a hot topic ranking unit. The rule-making unit is used to formulate rules for inclusion in the list and to filter emergencies according to these rules. Specifically, the rules for inclusion in the list and the filtering process include:

[0054] Screen for emergencies and remove those that are not included in the advanced management interface of the incident handling system;

[0055] If a reported emergency is processed by the advanced management interface of the event handling system, it will be automatically added to the list.

[0056] The event level of reported emergencies is determined, and those that fall within the preset event levels are automatically included in the list.

[0057] The system assesses the urgency of reported emergencies and automatically adds them to the list if they fall within a preset urgency level.

[0058] Set up daily, weekly, and monthly rankings as needed, and set a preset number of entries based on the timeliness of the daily, weekly, and monthly rankings.

[0059] In this embodiment, the high-level management terminal of the event handling system represents emergencies that require management-level processing, indicating a high level of influence and severity. Therefore, events that can be processed by the high-level management terminal of the event handling system are among the selections for publication events. The level of the reported emergencies is judged, and high-level emergencies are selected as publication events, such as events with the level of major, serious, or extremely serious. The urgency of the reported emergencies is then judged, and events with high urgency are selected as publication events, such as emergencies with the urgency level of urgent or extremely urgent.

[0060] For the selected events, daily, monthly, and weekly lists are created, and the number of events listed is set, for example, 5 events for the daily list, 10 events for the weekly list, and 10 events for the monthly list. The purpose of creating the lists in this application is to facilitate the accurate control of the number of events for subsequent review.

[0061] The event risk calculation unit is used to calculate the risk value of the screened emergencies using the event risk calculation model. Specifically, the expression of the event risk calculation model is as follows:

[0062]

[0063] Where Sa represents the initial score calculation, Normalize represents the normalization process, and W i The keyword indicating the risk of a hit, P i N represents the weight of the keyword. j The key number representing the risk of a hit, P j P represents the weight of the key number. r S represents the rule weight. m P represents the AI ​​model score. m P represents the weights of the AI ​​model. el S represents the float range corresponding to the event level. el S represents the base score corresponding to the event level. ul The score indicates the level of urgency.

[0064] By using an event risk calculation model to calculate the risk value of events in the publication, it is easier for the subsequent hot list ranking unit to set the ranking of the listed events based on the risk value. When producing the publication, only the top N events are selected for production, which makes it easier for editors and reviewers to control the number of publications and the events on the daily, weekly and monthly lists, and avoid blindly compiling publications.

[0065] The emergency incident handling module is used to extract the event structure and content of the editing and publishing events, and to associate and integrate the extracted event structure and content according to the publication template to generate a new event structure and corresponding event content that corresponds to the publication template. The emergency incident handling module includes an event extraction unit and an integration unit. The event extraction unit is used to extract the event structure and event content of the editing and publishing events, and the integration unit is used to associate the extracted event structure with the publication template so that the event structure and its event content correspond to the publication template.

[0066] In this embodiment, the event extraction module extracts the event structure and content of the publication events. Specifically, the event structure includes title, description, time, source, location, involved parties, key areas, type and information level, and key figures. The event content corresponds to the content of each event structure. After associating and integrating them through the publication template, event structures that do not belong to the publication template are removed, thereby forming an event publication with the same format, structure, type, and content as the publication template. However, after the event publication initially corresponds to the publication template, its content is rough and simple, insufficient to form complete publication content. Therefore, it is optimized through the event optimization module. Specifically:

[0067] Event Optimization Module: This module constructs a big data language model, using historical publication data and templates as training samples. After training, it prioritizes new event structures and corresponding content based on the event popularity ranking. The module includes a big data language model construction unit and a popular event preprocessing unit. The big data language model construction unit builds the model and uses historical publication data and templates as training samples for format training. During training, it communicates with the internet to use internet-based publication data as content enhancement training samples. The popular event preprocessing unit generates a big data language model processing order for popular events based on their popularity ranking.

[0068] Meanwhile, the event optimization module also includes an event similarity comparison module. This module compares the current event content with historical events to determine similarity. Similarity can be determined based on factors such as event content keywords, key numbers, and type. After comparison, the historical event with the highest similarity is extracted into a big data language model. The big data language model then compares the current event with the optimized content of the historical event. If the optimization result of the current event is inferior to that of the historical event, prompt words are generated to guide the user to further optimize and generate the desired result, thereby improving the optimization results of each event.

[0069] In this embodiment, unlike existing big data language models, the big data language model of this application is initially trained on the basis of the original historical publication data, so that the big data language model can initially expand and supplement the publication content unique to the event processing system. On this basis, it is retrained through the Internet, which further improves the text content beautification and improvement capabilities of the big data language model. In this way, it can deeply improve and supplement the content of the initial event publication to solve its rough and simple problem.

[0070] Finally, the intelligent editing module uses the trained big data language model as a generator to generate the corresponding publications for events that are listed or selected as publications, greatly improving the editing efficiency of editors.

[0071] After the publication is generated, it is reviewed by the review module, either manually or intelligently. If the review process is insufficient, it can be further improved by the big data language model to meet the publication requirements of the event processing system.

[0072] like Figure 2 As shown, in another embodiment of this example, an intelligent editing method for emergency management is also included, which is applied to the aforementioned intelligent editing system for emergency management, and includes:

[0073] Based on the incident handling system, we obtain the emergency incidents reported by personnel, as well as the historical publication data stored in the incident handling system;

[0074] The feature information of historical publications is filtered, and the format structure of the feature information and the relationship between the format structures are programmed to generate publication templates.

[0075] The acquired breaking news will be used to generate a hot list of events according to the preset hot list rules, and the events on the hot list will be used as events for publication.

[0076] Extract the event structure and content of the editing events, and integrate the extracted event structure and content according to the publication template to generate a new event structure and corresponding event content that correspond to the publication template.

[0077] A big data language model is constructed by using historical publication data and publication templates as training samples. After training, new event structures and corresponding event content are input into the big data language model based on the event hot list as the priority.

[0078] The big data language model optimizes the corresponding event content based on the new event structure, and the optimized event content and the new event structure generate the publication.

[0079] In addition, a program product is included, comprising a computer-readable storage medium, wherein the computer program instructions, when executed by a processor, cause the processor to perform the steps of the intelligent journal editing method for emergency management provided in any embodiment of the present invention.

[0080] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of the present invention. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0081] The above are merely embodiments of the present invention. Commonly known structures and characteristics are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are aware of all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, under the guidance of this application, improve and implement this solution in combination with their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. An intelligent editing system for emergency management, characterized in that: include: Data acquisition module: used to acquire emergency reports from personnel from the incident handling system, as well as historical publication data stored in the incident handling system; Publication template generation module: used to filter the feature information of historical publication data, programmatically process the format structure of the feature information and the relationship between the format structures, and generate publication templates; Event Hot List Generation Module: Used to generate an event hot list based on the acquired breaking events according to the preset hot list rules, and to use the events on the event hot list as the events for publication. Emergency handling module: used to extract the event structure and event content of the journal's events, and to integrate the extracted event structure and event content according to the journal template to generate a new event structure and corresponding event content that corresponds to the journal template; Event optimization module: Construct a big data language model, use historical publication data and publication templates as training samples to train the big data language model, and after training, input new event structures and corresponding event content into the big data language model according to the event hot list as priority; Intelligent editing module: The big data language model optimizes the corresponding event content based on the new event structure, and the optimized event content and the new event structure generate the publication.

2. The intelligent editing system for emergency management according to claim 1, characterized in that: The data acquisition module includes an emergency event acquisition unit and a historical data storage unit. The emergency event acquisition unit acquires emergency events through multiple methods, and the historical data storage unit is used to store historical publication data.

3. The intelligent editing system for emergency management according to claim 2, characterized in that: In the publication template generation module, the characteristic information of historical publication data includes the type, layout, format, and format content of the historical publication; The specific steps of programmatically processing the format structure of feature information and the correlation between format structures are as follows: The structure and format content of the format for extracting feature information from historical publication data; The importance of the format structure is calculated based on the layout of historical publications; Relate the format structures according to their format content and importance.

4. The intelligent editing system for emergency management according to claim 3, characterized in that: The hot list generation module includes a rule-making unit for listing events, an event risk calculation unit, and a hot list ranking unit. The rule-making unit is used to formulate rules for listing events and to filter emergencies according to the rules for listing events. The event risk calculation unit is used to calculate the risk value of the filtered emergencies through an event risk calculation model. The hot list ranking unit is used to rank emergencies according to the risk value and take the top N emergencies as the events to be published.

5. The intelligent editing system for emergency management according to claim 4, characterized in that: The expression for the event risk calculation model is: Where Sa represents the initial score calculation, Normalize represents the normalization process, and W i The keyword indicating the risk of a hit, P i N represents the weight of the keyword. j The key number representing the risk of a hit, P j P represents the weight of the key number. r S represents the rule weight. m P represents the AI ​​model score. m P represents the weights of the AI ​​model. el S represents the float range corresponding to the event level. el S represents the base score corresponding to the event level. ul The score indicates the level of urgency.

6. The intelligent editing system for emergency management according to claim 5, characterized in that: The specific steps for formulating the rules for listing events and screening emergencies based on these rules are as follows: Screen for emergencies and remove those that are not included in the advanced management interface of the incident handling system; If a reported emergency is processed by the advanced management interface of the event handling system, it will be automatically added to the list. The event level of reported emergencies is determined, and those that fall within the preset event levels are automatically included in the list. The system assesses the urgency of reported emergencies and automatically adds them to the list if they fall within a preset urgency level. Set up daily, weekly, and monthly rankings as needed, and set a preset number of entries based on the timeliness of the daily, weekly, and monthly rankings.

7. The intelligent editing system for emergency management according to claim 6, characterized in that: The emergency handling module includes an event extraction unit and an integration unit. The event extraction unit is used to extract the event structure and event content of the journal's events. The integration unit is used to associate the extracted event structure with the journal template so that the event structure and its event content correspond to the journal template.

8. The # according to claim 7, characterized in that: The event optimization module includes a big data language model construction unit and a hot list event preprocessing unit. The big data language model construction unit is used to construct a big data language model and use historical publication data and publication templates as training samples for format training of the big data language model. During the training process, it communicates with the Internet and uses Internet publication data as content beautification training samples to train the publication content. The hot list event preprocessing unit is used to generate a big data language model processing order for hot list events according to their hot list ranking. It also includes an auditing module, which is used to audit the generated publications by human or artificial intelligence.

9. An intelligent editing method for emergency management, applied to the intelligent editing system for emergency management as described in any one of claims 1-8, characterized in that: include: Based on the incident handling system, we obtain the emergency incidents reported by personnel, as well as the historical publication data stored in the incident handling system; The feature information of historical publications is filtered, and the format structure of the feature information and the relationship between the format structures are programmed to generate publication templates. The acquired breaking news will be used to generate a hot list of events according to the preset hot list rules, and the events on the hot list will be used as events for publication. Extract the event structure and content of the editing events, and integrate the extracted event structure and content according to the publication template to generate a new event structure and corresponding event content that correspond to the publication template. A big data language model is constructed by using historical publication data and publication templates as training samples. After training, new event structures and corresponding event content are input into the big data language model based on the event hot list as the priority. The big data language model optimizes the corresponding event content based on the new event structure, and the optimized event content and the new event structure generate the publication.

10. A program product, characterized in that: It includes a computer program or instructions that, when executed by a processor, implement the intelligent editing method for emergency management as described in claim 9 above.

Images