Three-dimensional situation early warning method, system and device and storage medium

Abstract

The invention provides a three-dimensional situation early warning method, system and device and a storage medium, a work order is input into an early warning model with a pre-configured task, the early warning field, event information and early warning level of the work order are obtained, the configuration of the early warning model comprises the steps of obtaining an early warning index value of a historical work order, and the early warning index comprises time, space and a statistical period; analyzing the historical work order through a theme analysis algorithm to obtain an early warning event; analyzing the early warning event through an event analysis algorithm to obtain event information; and inputting the early warning index value into an early warning rule, and calculating an early warning level. According to the invention, the early warning indexes are divided into three dimensions of time, space and statistical period for comprehensive analysis to obtain the early warning event and the early warning level, which is helpful for capturing dynamic changes and complex relationships, and solves the limitation problem in situation early warning. And the problem of lack of flexibility in situation early warning is solved through a theme analysis algorithm and an event analysis algorithm.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a three-dimensional situation warning method, system, device and storage medium. Background Art

[0002] The mainstream methods for early warning of complaints and opinions in different fields include rule-based and machine learning-based methods. In rule-based methods, a series of rules and thresholds need to be defined to detect specific events or patterns. These rules rely on the knowledge of experts in the field or known patterns in historical data. For example, a patent for active early warning of unconventional situation risks (patent number CN117649113A) uses rules to determine the mechanism of risk warning. In machine learning-based methods, supervised learning or unsupervised learning and other technologies are used to train based on historical data, and predictions and early warnings are made based on the trained models. This method is suitable for complex and dynamically changing scenarios and scenarios that generate large amounts of data, such as medical diagnosis and traffic management.

[0003] Different scenarios represent different data, and the data type and complexity affect our technology selection. It is highly dependent on expert knowledge and known rules and patterns, has certain limitations, and cannot cope with complex and highly dynamic data situations. Methods based solely on machine learning require a large amount of labeled data for training, and have high requirements for data quality. To a certain extent, the model effect depends on data characteristics, has black box characteristics, and lacks flexibility. Summary of the invention

[0004] In view of this, the embodiments of the present application are committed to providing a three-dimensional situation warning method, system, device and storage medium to solve the limitations of the situation warning technology in the prior art, as well as the problems of lack of flexibility and low accuracy.

[0005] According to a first aspect of the present invention, the present invention provides a three-dimensional situation warning method, which inputs a work order into a warning model with pre-configured tasks to obtain the warning field, event information and warning level of the work order, and the configuration of the warning model includes: Obtain the warning indicator value of the historical work order, the warning indicator includes time, space and statistical period; Analyze the historical work orders through a topic analysis algorithm to obtain warning events; Analyze the warning event through an event analysis algorithm to obtain event information; The warning index value is input into the warning rule to calculate the warning level.

[0006] Preferably, analyzing the historical work orders by a topic analysis algorithm to obtain warning events includes: Identify and classify the title and content of the historical work order by using the predefined classification conditions in the topic analysis algorithm, and calculate the correlation between the content of the historical work order and the predefined classification conditions by using the BM25 algorithm to obtain the warning field of the historical work order; The warning event is obtained through the warning field.

[0007] Preferably, analyzing the warning event by an event analysis algorithm to obtain event information includes: Obtain keywords in the historical work order by keyword extraction, and obtain the event name and the location of the event based on the keywords; Analyze the warning events through thematic analysis and classify the warning events according to pre-classified types; Analyze the classified warning events through event clustering to obtain the content of event types; The pre-classified types include emergencies, persistent hot spots, repeated complaints, potential incidents, sensitive incidents and concentrated incidents.

[0008] Preferably, the inputting the warning indicator value into the warning rule to calculate the warning level includes: Counting the total number of requests within the time dimension of the event occurrence in the historical work order as the warning indicator value; The warning level is calculated based on the warning indicator value and the statistical period.

[0009] Preferably, the calculating the warning level by using the warning indicator value and the statistical period includes: When the statistical period is the same period, the increase ratio is obtained by comparing the total number of complaints from the start date of the event to the day of the warning with the total number of complaints in the same period of the previous period; When the statistical period is not the same period, the increase ratio is calculated by comparing the total number of complaints on the warning day with the average number of complaints during the statistical period; The warning level is determined based on the increase ratio.

[0010] Preferably, determining the warning level based on the increase ratio includes: When the increase ratio is lower than the first threshold, the warning level is low; When the increase ratio is greater than the first threshold and less than the second threshold, the warning level is medium; When the increase ratio is greater than the second threshold, the warning level is high; When the demand volume in the previous period and the statistical period is zero, the warning level is high.

[0011] Preferably, it also includes: The anomaly detection model is trained through historical work order data to output abnormal work orders that do not conform to normal behavior; The time, space, and event information are monitored in real time through the trained anomaly detection model, and when an abnormal work order that does not conform to normal behavior is identified, early warning information is provided.

[0012] According to a second aspect of the present invention, a three-dimensional situation warning system is also provided, comprising: An acquisition module is used to obtain the warning indicator value of the historical work order, wherein the warning indicator includes time, space and statistical period; The theme module analyzes the historical work orders through a theme analysis algorithm to obtain warning events; An event module analyzes the warning event through an event analysis algorithm to obtain event information; The early warning module inputs the early warning index value into the early warning rule and calculates the early warning level.

[0013] According to a third aspect of the present invention, an electronic device is also provided, comprising a processor and a memory, wherein the memory stores computer executable instructions that can be executed by the processor, and the processor executes the computer executable instructions to implement the three-dimensional situation warning method.

[0014] According to a fourth aspect of the present invention, a computer-readable storage medium is also provided, wherein the computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are called and executed by a processor, the computer-executable instructions prompt the processor to implement the three-dimensional situation warning method.

[0015] The effects provided in the content of the invention are only the effects of the embodiments, not all the effects of the invention. One of the above technical solutions has the following advantages or beneficial effects: According to the first aspect of the present invention, the present invention configures tasks for the warning model through a topic analysis algorithm, an event analysis algorithm and warning rules, inputs the work order into the configured warning model, and obtains comprehensive analysis data in three dimensions of time, space and statistical period. It is a comprehensive and systematic method in situation warning. It solves the problem of lack of flexibility in situation warning through the topic analysis algorithm and the event analysis algorithm. The comprehensive analysis of three dimensions helps to capture dynamic changes and complex relationships, and at the same time solves the limitation problem in situation warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic flow chart of a three-dimensional situation warning method according to an embodiment of the present invention is shown; Figure 2 The present invention shows Figure 1A schematic flow chart of a method for obtaining a warning event in step S200; Figure 3 The present invention shows Figure 1 A schematic flow chart of a method for obtaining event information in step S300; Figure 4 The present invention shows Figure 1 A schematic flow chart of the method for calculating the warning level in step S400; Figure 5 The present invention shows Figure 4 A schematic flow chart of the method for determining the warning level in step S420; Figure 6 A schematic flow chart of a three-dimensional situation warning system according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0017] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are described in detail below in conjunction with the accompanying drawings. It is to be understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for ease of description, only some structures related to the present application are shown in the accompanying drawings, rather than all structures. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0018] The terms "including" and "having" and any variations thereof in this application are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products or devices.

[0019] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0020] Figure 1 FIG. 4 shows a schematic flow chart of a three-dimensional situation warning method according to an embodiment of the present invention. Figure 1 As shown, the present invention provides a three-dimensional situation warning method, which inputs a work order into a warning model with pre-configured tasks to obtain the warning field, event information and warning level of the work order. The configuration of the warning model includes: Step S100, obtaining the warning indicator value of the historical work order, the warning indicator includes time, space and statistical period; Step S200, analyzing historical work orders through a topic analysis algorithm to obtain warning events; Step S300, analyzing the warning event through an event analysis algorithm to obtain event information; Step S400: input the warning index value into the warning rule to calculate the warning level.

[0021] By configuring the warning model tasks through the topic analysis algorithm, event analysis algorithm and warning rules, the work order is input into the configured warning model to obtain comprehensive analysis data in three dimensions of time, space and statistical period. This is a comprehensive and systematic method in situation warning. The topic analysis algorithm and event analysis algorithm solve the problem of lack of flexibility in situation warning. The comprehensive analysis of three dimensions helps to capture dynamic changes and complex relationships, and at the same time solves the limitation problem in situation warning.

[0022] In this step, the time dimension refers to the precision of the time range. Different types of demands have different reference time range precisions. For example, daily problems that have no time pattern require real-time monitoring and real-time cumulative statistics of the number of demands for the day; for example, periodic demands such as the college entrance examination need to be counted according to this type of periodic range.

[0023] The spatial dimension refers to the precision of the spatial scope, such as cities, districts, counties, streets, etc. Different types of demands refer to different areas of precision. For example, some consulting issues only need to be within the scope of provinces and cities, while some reports and complaints about public facilities must be accurate to the street or community.

[0024] The statistical period is the historical period of the change trend of the number of appeals corresponding to the time dimension that requires an alert. Conventional statistical periods include yesterday, the past week, the past month, the past year, and custom periods. The periodic statistical period is determined based on specific issues.

[0025] Figure 2 The present invention shows Figure 1 A schematic flow chart of the method for obtaining a warning event in step S200; Figure 2 As shown, in step S200, the historical work orders are analyzed by a topic analysis algorithm to obtain warning events, including: Step S210, identifying and classifying the titles and contents of the historical work orders by using the predefined classification conditions in the topic analysis algorithm, and calculating the correlation between the contents of the historical work orders and the predefined classification conditions by using the BM25 (Okapi BM25) algorithm to obtain the warning areas of the historical work orders; Step S220, obtaining a warning event through the warning field.

[0026] Rules are formed through predefined classification conditions, and preliminary classification can be achieved through the rules to ensure that obvious, fixed-model content can be accurately classified. The BM25 algorithm can then be used to measure the correlation between the work order content and predefined classification keywords, thereby achieving more precise classification and ensuring higher accuracy.

[0027] In this step, topic analysis is to classify the data into a defined classification system based on the work order content, keywords in the title and user-defined rules, and finally output a three-level classification.

[0028] Figure 3 The present invention shows Figure 1 A schematic flow chart of a method for obtaining event information in step S300; like Figure 3 As shown, in step S300, the warning event is analyzed by an event analysis algorithm to obtain event information, including: Step S310, obtaining keywords in the historical work order through keyword extraction, and obtaining the event name and the location of the event based on the keywords; Step S320, analyzing the warning events through thematic analysis, and classifying the warning events according to pre-classified types; Step S330, analyzing the classified warning events through event clustering to obtain the content of the event type; Among them, the pre-divided types include emergencies, persistent hot spots, repeated complaints, potential events, sensitive events and concentrated events.

[0029] Keyword extraction (Kleinberg) detects sudden keywords in text data to help identify important events. Topic analysis (LDA) discovers potential topics in document collections, and by calculating the distribution of words and documents, documents are classified according to pre-divided types. Event clustering (K-Means) is then used to group related events together to form event clusters. The combination of multiple algorithms can more accurately identify and analyze events.

[0030] In this step, event analysis locates an event based on the elements such as the person making the appeal, time, location, and appeal content in the hotline ticket. Based on different types of appeals and the data characteristics of the ticket, the event can be further subdivided into different event types, including emergencies, persistent hot spots, repeated complaints, potential events, sensitive events, and concentrated events. Event analysis will eventually output the event name, location, volume, and type of event.

[0031] In step S320, as shown in Table 1, the warning events are classified by rule judgment.

[0032]

[0033] Figure 4 The present invention shows Figure 1 A schematic flow chart of the method for calculating the warning level in step S400; like Figure 4 As shown, in step S400, the warning indicator value is input into the warning rule to calculate the warning level, including: Step S410, counting the total number of requests in the time dimension of the event occurrence in the historical work order as the warning indicator value; Step S420, calculating the warning level through the warning indicator value and the statistical period.

[0034] Figure 5 The present invention shows Figure 4 A schematic flow chart of the method for determining the warning level in step S420; like Figure 5 As shown, in step S420, the warning level is calculated by the warning indicator value and the statistical period, including: Step S421, when the statistical period is the same period, the total number of claims from the event start date to the warning date is compared with the total number of claims in the same period of the previous period to obtain the increase ratio; Step S422, when the statistical period is not the same period, the increase ratio is calculated by comparing the total number of demands on the warning day with the average number of demands within the statistical period; Step S423, determining the warning level based on the increase ratio.

[0035] The warning level rule refers to the rule for dividing the warning level according to the increase ratio of the number of complaints in different statistical periods. First, the increase ratio is calculated by comparing the cumulative number of complaints on the warning day with the number of complaints in the statistical period. When the statistical period is yesterday: the calculation method is to compare the cumulative number of complaints on the warning day with the number of complaints in the same period yesterday, and calculate the increase ratio. When the statistical period is the past week, the past month, or the past year: the calculation method is to calculate the increase ratio between the cumulative number of complaints on the warning day and the average value in the statistical period. When the statistical period is the same period: calculate the cumulative number of complaints from the start date of the period to the warning day compared with the cumulative number of complaints in the same period of the previous period. For example: the college entrance examination period is configured as May 7-July 7, and the warning day is June 1, then calculate the increase ratio of the cumulative number of complaints from May 7 to June 1 compared with the cumulative number of complaints from May 7 to June 1 last year.

[0036] Wherein, in step S430, determining the warning level based on the increase ratio includes: When the increase ratio is lower than the first threshold, the warning level is low; When the increase ratio is greater than the first threshold and less than the second threshold, the warning level is medium; When the increase ratio is greater than the second threshold, the warning level is high; When the demand volume in the previous period and the statistical period is zero, the warning level is high.

[0037] In this step, the warning level is determined according to the rules. The initial setting of the increase ratio is that if the increase ratio is between 0-5%, the warning level is low; if the increase ratio is between 5-10%, the warning level is medium; if the increase ratio is greater than 10%, the warning level is high. In addition, when the number of previous requests is 0 or the warning level associated with the event is high, the warning level is directly high. For example, an event occurs in the request of the work order, but the same event is not found in the training data, and the increase ratio cannot be calculated. Therefore, the warning level of this event is set to high.

[0038] Input the work order into the pre-configured task warning model to obtain the warning area, event information and warning level of the work order, as well as: The anomaly detection model is trained through historical work order data to output abnormal work orders that do not conform to normal behavior; The trained anomaly detection model monitors time, space, and event information in real time, and provides early warning information when abnormal work orders that do not conform to normal behavior are identified.

[0039] By introducing anomaly detection models to monitor work order data in real time, we can identify abnormal work orders that do not conform to normal behavior patterns and provide immediate warning information. By learning the normal patterns of historical data, we can detect anomalies in real-time data. This feature enhances the real-time and dynamic response capabilities of the system.

[0040] Among them, the work order data is preprocessed, feature vectors such as time, space, and event type are extracted and standardized, and used as the input of the automatic encoder in the anomaly detection model for training, and the model parameters are optimized by minimizing the reconstruction error.

[0041] The real-time work order data is reconstructed using the trained autoencoder and the reconstruction error is calculated. Work orders that exceed the set threshold are judged as abnormal work orders. The real-time work order data includes the results output by the topic analysis algorithm and the event analysis algorithm.

[0042] Figure 6 FIG. 4 shows a schematic flow chart of a three-dimensional situation warning system according to an embodiment of the present invention. Figure 6 As shown, a three-dimensional situation warning system is also provided, including: The acquisition module 610 acquires the warning indicator value of the historical work order, and the warning indicator includes time, space and statistical period; Theme module 620 analyzes historical work orders through a theme analysis algorithm to obtain warning events; The event module 630 analyzes the warning event through an event analysis algorithm to obtain event information; The warning module 640 inputs the warning index value into the warning rule and calculates the warning level.

[0043] In particular, an electronic device is also provided, the electronic device includes a processor and a memory, the memory stores computer executable instructions that can be executed by the processor, and the processor executes the computer executable instructions to implement the three-dimensional situation warning method, for example, to execute the above-described Figure 1 Steps S100 to S400 of the method described above are also performed. Figure 3 The method comprises steps S310 to S330.

[0044] In particular, a computer-readable storage medium is also provided, the computer-readable storage medium storing computer-executable instructions, when the computer-executable instructions are called and executed by the processor, the computer-executable instructions prompt the processor to implement the three-dimensional situation warning method, for example, to execute the above-described Figure 1 Steps S100 to S400 of the method described above are also performed. Figure 3 The method comprises steps S310 to S330.

[0045] The processor is used to complete various control logics of the device, which can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), a single-chip microcomputer, an ARM (Acorn RISCMachine) or other programmable logic device, a discrete gate or transistor logic, a discrete hardware component, or any combination of these components. In addition, the processor can also be any traditional processor, microprocessor or state machine. The processor can also be implemented as a combination of computing devices, for example, a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors combined with a DSP and / or any other such configuration.

[0046] The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system and applications required for at least one function; the data storage area may store data created according to the use of the device, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0047] As an example, non-volatile storage media can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) as external cache memory. By way of illustration and not limitation, RAM can be obtained in many forms such as synchronous RAM (SRAM), dynamic RAM, (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synchlink DRAM (SLDRAM), and direct Rambus RAM (DRRAM). The disclosed memory components or memories of the operating environment described herein are intended to include one or more of these and / or any other suitable types of memory.

[0048] Of course, those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware (such as a processor, a controller, etc.) through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium, and the computer program can include the processes of the above-mentioned method embodiments when executed. The storage medium can be a memory, a disk, a floppy disk, a flash memory, an optical storage device, etc.

[0049] The above are only some specific implementation methods of the present application, and any other improvements made based on the concept of the present application are deemed to be within the protection scope of the present application.

Claims

1. A three-dimensional situation warning method, characterized in that: Input the work order into the pre-configured task warning model to obtain the warning field, event information and warning level of the work order. The configuration of the warning model includes: Obtain the warning indicator value of the historical work order, the warning indicator includes time, space and statistical period; Analyze the historical work orders through a topic analysis algorithm to obtain warning events; Analyze the warning event through an event analysis algorithm to obtain event information; The warning indicator value is input into the warning rule to calculate the warning level.

2. The method according to claim 1, characterized in that: The historical work orders are analyzed by a topic analysis algorithm to obtain warning events, including: Identify and classify the title and content of the historical work order by using the predefined classification conditions in the topic analysis algorithm, and calculate the correlation between the content of the historical work order and the predefined classification conditions by using the BM25 algorithm to obtain the warning field of the historical work order; The warning event is obtained through the warning field.

3. The method according to claim 1, characterized in that The step of analyzing the warning event by an event analysis algorithm to obtain event information includes: Obtain keywords in the historical work order by keyword extraction, and obtain the event name and the location of the event based on the keywords; Analyze the warning events through thematic analysis and classify the warning events according to pre-classified types; Analyze the classified warning events through event clustering to obtain the content of event types; The pre-classified types include emergencies, persistent hot spots, repeated complaints, potential incidents, sensitive incidents and concentrated incidents.

4. The method according to claim 1, characterized in that: The step of inputting the warning indicator value into the warning rule and calculating the warning level includes: The total number of requests in the time dimension of the event occurrence in the historical work order is counted and the warning indicator value is calculated by exponential smoothing average method; The warning level is calculated based on the warning indicator value and the statistical period.

5. The method according to claim 4, characterized in that The calculating the warning level by using the warning indicator value and the statistical period includes: When the statistical period is the same period, the increase ratio is obtained by comparing the total number of complaints from the start date of the event to the day of the warning with the total number of complaints in the same period of the previous period; When the statistical period is not the same period, the increase ratio is calculated by comparing the total number of complaints on the warning day with the average number of complaints during the statistical period; The warning level is determined based on the increase ratio.

6. The method according to claim 5, characterized in that The determining the warning level based on the increase ratio includes: When the increase ratio is lower than the first threshold, the warning level is low; When the increase ratio is greater than the first threshold and less than the second threshold, the warning level is medium; When the increase ratio is greater than the second threshold, the warning level is high; When the demand volume in the previous period and the statistical period is zero, the warning level is high.

7. The method according to claim 1, characterized in that Also includes: The anomaly detection model is trained through historical work order data to output abnormal work orders that do not conform to normal behavior; The time, space, and event information are monitored in real time through the trained anomaly detection model, and when an abnormal work order that does not conform to normal behavior is identified, early warning information is provided.

8. A three-dimensional situation warning system, characterized in that: include: An acquisition module is used to obtain the warning indicator value of the historical work order, wherein the warning indicator includes time, space and statistical period; The theme module analyzes the historical work orders through a theme analysis algorithm to obtain warning events; An event module analyzes the warning event through an event analysis algorithm to obtain event information; The early warning module inputs the early warning index value into the early warning rule and calculates the early warning level.

9. An electronic device, characterized in that: The electronic device includes a processor and a memory, the memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement the three-dimensional situation warning method described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are called and executed by the processor, the computer-executable instructions prompt the processor to implement the three-dimensional situation warning method described in any one of claims 1 to 7.

Images