Urban user event-oriented resource scheduling method and device based on big data

Abstract

The embodiment of the invention discloses an urban user event-oriented resource scheduling method and device based on big data. A specific embodiment of the method comprises the following steps: preprocessing a real-time event data stream, a preset assessment rule base and a historical event database to obtain a preprocessed data set; performing deep semantic analysis and multi-label classification processing on the preprocessed data set to generate semantic understanding information; performing multi-dimensional elimination judgment and resource demand mapping processing on the semantic understanding information to generate event property and resource demand information; generating an effective governance event list based on the event property and the resource demand information, and constructing a corresponding resource demand thermodynamic diagram based on the effective governance event list; generating a dynamic resource allocation instruction set; and scheduling the corresponding cross-modal resources according to the dynamic resource allocation instruction set. According to the embodiment, accurate resource scheduling and efficient cross-modal resource configuration in the urban user event treatment process can be realized, and the time cost of event treatment is reduced.

Description

Technical Field

[0001] The embodiments disclosed herein relate to the field of computer technology, and more specifically to a resource scheduling method and apparatus for urban user events based on big data. Background Technology

[0002] Currently, resource scheduling methods for urban user events (e.g., hotline complaint cases) mainly rely on manual experience or simple assignment based on fixed rules. However, when using these methods for resource scheduling, the following technical problems often arise: In urban user incident governance, low data processing efficiency, insufficient accuracy in incident analysis, lack of priority guidance and poor matching of resource scheduling, and lack of dynamism in cross-modal resource allocation lead to slow incident response, low resource utilization, and easy occurrence of governance oversights or resource waste.

[0003] The information disclosed in this background section is only intended to enhance the understanding of the background of the inventive concept, and therefore may contain information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion later. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0005] Some embodiments of this disclosure propose a resource scheduling method, apparatus, electronic device, and computer-readable medium based on big data for urban user events, in order to solve one or more of the technical problems mentioned in the background section above.

[0006] Firstly, some embodiments of this disclosure provide a resource scheduling method for urban user events based on big data, including: in response to receiving processing information for urban user events, preprocessing the collected real-time event data stream related to urban user events, a preset assessment rule base, and a historical event database to obtain a preprocessed dataset; performing deep semantic parsing and multi-label classification processing on the preprocessed dataset to generate semantic understanding information, wherein the semantic understanding information includes: event nature classification labels, core demand vectors, and processing estimated complexity information; and performing multi-dimensional elimination judgment and resource demand mapping processing on the semantic understanding information to... The process generates event nature and resource requirement information, which includes: event priority information generated based on big data; based on the event nature and resource requirement information, it generates a list of effective governance events and constructs a corresponding resource requirement heatmap based on the list of effective governance events; based on the resource requirement heatmap and a preset resource mapping library, it generates a dynamic resource allocation instruction set with priority information; and according to the dynamic resource allocation instruction set, it schedules corresponding cross-modal resources to execute governance operations for corresponding city user events, wherein the cross-modal resources include at least one of the following: computing resources, transmission resources, and material reserve resources.

[0007] Secondly, some embodiments of this disclosure provide a resource scheduling device for urban user events based on big data, including a preprocessing unit configured to, in response to receiving processing information for urban user events, preprocess the collected real-time event data stream related to urban user events, a preset assessment rule base, and a historical event database to obtain a preprocessed dataset; a parsing and classification unit configured to perform deep semantic parsing and multi-label classification processing on the preprocessed dataset to generate semantic understanding information, wherein the semantic understanding information includes: event nature classification labels, core demand vectors, and processing estimated complexity information; and a mapping unit configured to perform multi-dimensional elimination judgment and resource demand mapping processing on the semantic understanding information to generate... The system generates event nature and resource demand information, including event priority information generated based on big data; a first generation unit is configured to generate a list of effective governance events based on the event nature and resource demand information, and to construct a corresponding resource demand heatmap based on the list of effective governance events; a second generation unit is configured to generate a dynamic resource allocation instruction set with priority information based on the resource demand heatmap and a preset resource mapping library; and a scheduling unit is configured to schedule corresponding cross-modal resources according to the dynamic resource allocation instruction set to perform governance operations for corresponding city user events, wherein the cross-modal resources include at least one of the following: computing resources, transmission resources, and material reserve resources.

[0008] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, such that when the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any implementation of the first aspect.

[0009] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method as described in any implementation of the first aspect.

[0010] The above-described embodiments of this disclosure have the following beneficial effects: Through the resource scheduling method for urban user events based on big data, as described in some embodiments of this disclosure, precise resource scheduling and efficient cross-modal resource allocation are achieved in the urban user event governance process, reducing the time cost of event processing. Specifically, the reasons for low resource scheduling accuracy, low cross-modal resource allocation efficiency, and excessively high time costs are: a lack of ability to intelligently analyze, dynamically prioritize, and collaboratively map massive event data, resulting in resource mismatch and response delays. Based on this, the resource scheduling method for urban user events based on big data, as described in some embodiments of this disclosure, firstly, in response to receiving processing information for urban user events, preprocesses the collected real-time event data stream related to urban user events, the preset assessment rule base, and the historical event database to obtain a preprocessed dataset. Transforming the real-time event stream, the preset assessment rule base, and the historical event database into a well-organized preprocessed dataset provides a unified and clean data foundation for subsequent intelligent analysis, solving the problem of analytical accuracy caused by the messiness and inconsistent formats of the original data. Then, deep semantic parsing and multi-label classification are performed on the preprocessed dataset to generate semantic understanding information, which includes event nature classification labels, core demand vectors, and estimated processing complexity information. Extracting this semantic understanding information overcomes the limitations of traditional keyword matching, enabling accurate identification of complex event natures. Next, multi-dimensional elimination judgment and resource demand mapping are performed on the semantic understanding information to generate event nature and resource demand information, which includes event priority information generated based on big data. Combining rule matching and historical case correction, a judgment result including quantified resource demand and event priority information is generated, deeply integrating business logic (whether to eliminate) with resource planning (how much resource is needed), achieving a leap from event identification to resource demand prediction. Furthermore, based on the event nature and resource demand information, a list of effective governance events is generated, and a corresponding resource demand heatmap is constructed based on this list. Effective governance events are sorted based on event priority information, and the resource demand is spatially visualized using the resource demand heatmap. Next, based on the aforementioned resource demand heatmap and pre-defined resource mapping library, a dynamic resource allocation instruction set with priority information is generated. Combining the visualized resource demand heatmap with the pre-defined resource mapping library generates a specific, executable, and priority-embedded dynamic resource allocation instruction set, transforming the scheduling strategy into precise operational commands that drive the coordinated action of various resources. Finally, according to the aforementioned dynamic resource allocation instruction set, corresponding cross-modal resources are scheduled to execute governance operations for corresponding city user events. These cross-modal resources include at least one of the following: computing resources, transmission resources, and material reserve resources.By scheduling cross-modal resources according to the dynamic resource allocation instruction set, the collaborative and efficient allocation of multiple types of resources can be achieved, eliminating resource mismatch and response delay issues, accurately matching the event governance needs of urban users, and significantly reducing the time cost of event processing. Attached Figure Description

[0011] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0012] Figure 1 This is a flowchart of some embodiments of the resource scheduling method for urban user events based on big data according to this disclosure; Figure 2 This is a schematic diagram of the structure of some embodiments of the resource scheduling device for urban user events based on big data according to this disclosure; Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0013] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0014] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0015] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0016] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0017] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0018] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] refer to Figure 1 The diagram illustrates a flow 100 of some embodiments of a resource scheduling method for urban user events based on big data according to the present disclosure. This resource scheduling method for urban user events based on big data includes the following steps: Step 101: In response to receiving processing information for city user events, preprocess the collected real-time event data stream related to city user events, the preset assessment rule base and the historical event database to obtain a preprocessed dataset.

[0020] In some embodiments, the executing entity (e.g., an electronic device) of the above-described resource scheduling method for urban user events based on big data can be hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster composed of multiple servers or terminal devices, or as a single server or a single terminal device. When the computing device is software, it can be installed in the hardware devices listed above. It can be implemented as multiple software programs or software modules to provide distributed services, or as a single software program or software module. No specific limitations are made here.

[0021] In other embodiments, the aforementioned executing entity may, in response to receiving processing information regarding urban user events, preprocess the collected real-time event data stream related to the urban user events, a preset assessment rule base, and a historical event database to obtain a preprocessed dataset. The aforementioned urban user events may be events related to urban governance issues that are participated in or triggered by users in the city, such as a citizen's complaint about road flooding through a hotline platform. The aforementioned processing information may be instructions or feedback information that trigger the event processing flow, such as an event handling initiation notification message from the urban management department. The aforementioned real-time event data stream may be a sequence of urban user requests. For example, the aforementioned real-time event data stream may be a real-time message queue from a hotline platform. The aforementioned preset assessment rule base may refer to a predefined set of rules used to evaluate the quality and responsibility of event processing, and may include various assessment standards and classification criteria. For example, the aforementioned preset assessment rule base may be a database including specific rules such as "urban appearance and environmental sanitation issues must be responded to within 24 hours" and "repeated complaints require a higher response level." The aforementioned historical event database may be a structured data warehouse storing processed urban user events and related information. For example, the aforementioned historical event database could be a database of all resolved complaint cases within the past year, which could include information such as event description, handling department, processing time, and processing result. The aforementioned preprocessed dataset could be a set of numerical features formed after cleaning, correlation, rule fusion, and standardized encoding of real-time event data streams.

[0022] In some optional implementations of certain embodiments, the aforementioned execution entity may, in response to receiving processing information regarding city user events, preprocess the collected real-time event data stream related to city user events, the preset assessment rule base, and the historical event database to obtain a preprocessed dataset, which may include the following steps: The first step involves updating the aforementioned real-time event data stream based on preset rejection rules to obtain the original heterogeneous data stream. This original heterogeneous data stream can be an unprocessed collection of initial event data from diverse sources with inconsistent formats. For example, it could include raw data in mixed formats such as text complaints, on-site photos, voice recordings, and structured work orders pushed from different departmental systems. In practice, firstly, preset rejection rules are applied to the received real-time event stream. For example, invalid events such as "withdrawn" and "test submitted" are automatically filtered out, and events already marked as "not included in the assessment" on the municipal platform are simultaneously marked. Then, the events that were not initially rejected are aggregated to form the original heterogeneous data stream.

[0023] The second step involves cleaning the original heterogeneous data stream to obtain a cleaned event data stream. This cleaned event data stream can be a standardized set of event data that has undergone data cleaning, noise removal, error correction, and standardization. In practice, first, the anomalies, missing data, and duplicate data in the original heterogeneous data stream are identified. Then, erroneous data is corrected, missing data is filled in, and duplicate data is deleted. Finally, the standardized cleaned event data stream is output.

[0024] The third step involves associating the cleaned event data stream with relevant cases in the historical event database to construct an event association graph. These relevant cases can be processed cases in the historical event database that share similarities with the current event in terms of issue type, location, and involved parties. For example, if the current complaint is "elevator malfunction in Building 3 of Community A," then "elevator maintenance record for Building 1 of Community A last month" and "past complaints from the same complainant regarding elevators in Community A" in the historical event database are both relevant cases. The event association graph can be a structured network representing the relationships between the current event and historical events. Nodes in the event association graph can represent the current event and relevant historical cases retrieved from the historical database. Edges in the event association graph can represent the relationships between events, for example, based on similarity in geographical location, issue type, or time series. In practice, first, key features of the cleaned event stream (e.g., location, issue type, core demands) are extracted. Then, similarity retrieval and matching are performed in the historical event database. Finally, taking the current event as the central node, and using each relevant historical event retrieved as a related node, directed connections are established to construct an event association graph.

[0025] The fourth step involves fusing the aforementioned event association graph with the pre-defined assessment rule base to generate a rule-enhanced dataset. This rule-enhanced dataset can be an enhanced dataset containing rule labels, formed by associating and fusing event data with assessment rules. In practice, first, the current event node and its associated historical event nodes in the event association graph are traversed. Then, applicable specific assessment rules are matched from the assessment rule base based on event characteristics (e.g., type, location). Finally, these rule clauses (e.g., processing time limit, responsible department, assessment criteria) are appended as new attribute labels to the data records of the current event and its associated events, forming the rule-enhanced dataset.

[0026] The fifth step involves encoding the textual, spatiotemporal, and rule-label information in the rule-enhanced dataset to generate a preprocessed dataset. The textual information can be unstructured text content from event descriptions, representing the core part requiring semantic analysis. The spatiotemporal information can be the time and location of the event, a key dimension for spatial analysis and temporal processing. The rule-label information can be structured labels assigned to events based on an assessment rule base. For example, the rule-label information could be labels for event A: "Event Type: Urban Appearance and Environment," "Responsible Unit: Street Sanitation Office," "Processing Time Limit: 24 Hours," and "Priority: High." In practice, first, the textual information is segmented and vectorized (e.g., using a BERT model to generate word vectors). Then, the spatiotemporal information is encoded, such as converting geographic locations into geographic grid codes and time into timestamp sequences. Finally, the rule-label information is one-hot encoded or embedded encoded to generate the preprocessed dataset.

[0027] Step 102: Perform deep semantic parsing and multi-label classification on the preprocessed dataset to generate semantic understanding information.

[0028] In some embodiments, the aforementioned execution entity can perform deep semantic parsing and multi-label classification processing on the aforementioned preprocessed dataset to generate semantic understanding information. This semantic understanding information includes: event nature classification labels, core demand vectors, and estimated processing complexity information. The deep semantics can be the deeper meaning, contextual relationships, and true intent implied in the text, going beyond the literal meaning. For example, in the complaint description "My doorstep is almost flooded with garbage," the deep semantics are not the literal "flooded," but rather express the deeper demand that "the garbage accumulation is very serious, the situation is urgent, and immediate cleanup is needed." The multi-label classification can be a task that simultaneously classifies an event into multiple related categories. For example, a case can be classified into multiple labels such as "urban environment," "urgent," and "repeated complaint." The event nature classification labels can be standardized labels that identify the management field or type to which the event belongs, such as "environmental sanitation -> garbage collection," or "property management -> elevator malfunction." The core demand can be the deep meaning of the most essential intent and key information of the urban user event. For example, the core semantics of "the noise is so loud I can't sleep" is "dissatisfaction with the resting environment and a demand to resolve the noise source," rather than simply describing "there is sound." The aforementioned core demand vector can be a semantic feature representing the core demand of the event using mathematical vectors. The aforementioned estimated processing complexity information can be a quantitative assessment of the resources, time, and coordination difficulty required to handle the event. For example, when the estimated processing complexity information is 85 points (out of 100), it indicates that multi-departmental collaboration is required and the process will be time-consuming. For example, regarding "construction work at the site downstairs at night, multiple complaints have been ineffective," the semantic understanding information can include: event nature classification tags: noise pollution, construction management, repeated complaints; core demand vector: [demand intensity: 0.9, timeliness: 0.8, involved parties: construction party / urban management] (vectorized representation); estimated processing complexity: high (score 75 / 100).

[0029] In some optional implementations of certain embodiments, the aforementioned execution entity may perform deep semantic parsing and multi-label classification processing on the aforementioned preprocessed dataset to generate semantic understanding information, which may include the following steps: The first step is to generate a core semantic vector for the event based on the text information in the preprocessed dataset. This core semantic vector can be generated by compressing the text information into a high-dimensional numerical vector representing its core meaning. For example, a 200-word complaint description could be transformed into a 512-dimensional vector. In practice, firstly, the word vectors from the preprocessed dataset are input into a pre-trained language model (e.g., the BERT model). Then, deep semantic features of the text are extracted using a multi-layer attention mechanism. Finally, pooling is performed on the hidden states output by the model to generate a fixed-length "core semantic vector for the event".

[0030] The second step is to generate event nature classification labels and core demand vectors based on the aforementioned core semantic vectors of the events. The aforementioned event nature can be the fundamental attribute of urban user events within the scope of government administration or the functional area classification to which they belong. For example, the event nature of "blocking passageways with piled-up items in residential building corridors" can be classified as "fire safety hazard". In practice, firstly, the obtained event core semantic vector is input into a classifier (which can be a multi-label fully connected neural network, including an input layer (receiving the event core semantic vector generated in the previous step (e.g., 512-dimensional)), hidden layers (composed of one or more fully connected layers (e.g., 256-dimensional, 128-dimensional), each layer undergoing non-linear transformation through an activation function (e.g., ReLU) to progressively extract deep features related to classification), and an output layer (a fully connected layer including N neurons (N equal to the total number of preset property categories), each neuron corresponding to one category, typically using a Sigmoid activation function, independently outputting the probability (between 0 and 1) that the event belongs to that category. Categories with probabilities exceeding a set threshold (e.g., 0.5) are selected as the "event property classification label"). The connection relationship of the classifier can be that the event core semantic vector flows sequentially from the input layer through each hidden layer and finally reaches the output layer, with each layer being fully connected). Then, the event property classification label is determined through the classifier, and the multi-label classification result is output, for example, ["public facilities", "safety hazards"]. Simultaneously, through attention mechanisms (e.g., multi-head self-attention), a more refined semantic vector representing only the core appeal is extracted from the core semantic vector of the event. Finally, these are integrated into event nature classification labels and core appeal vectors.

[0031] The third step involves generating estimated processing complexity information based on the aforementioned event nature classification labels and core appeal vectors, using a pre-trained complexity prediction model. This pre-trained complexity prediction model can be a machine learning model used to predict the processing difficulty of new events. For example, it could be a regression model that takes the event nature classification labels and core appeal vectors as input and outputs estimated processing complexity information. This pre-trained complexity prediction model may include: an input layer (receiving the one-hot encoded event nature classification labels and core appeal vectors, outputting a fused feature vector), a hidden layer (consisting of 2 to 3 fully connected layers, each taking the output of the previous layer as input, undergoing non-linear transformation through an activation function, and outputting high-order abstract features), and an output layer (taking the hidden layer output (DNN) or fused feature vector (XGBoost) as input, and outputting multi-dimensional estimated processing complexity information through a regression head). This pre-trained complexity prediction model can be an XGBoost model or a DNN model.

[0032] The fourth step involves integrating the aforementioned event nature classification labels, core appeal vectors, and estimated processing complexity information into semantic understanding information. In practice, firstly, the event nature classification labels and core appeal vectors are standardized in format. Then, they are integrated with the estimated processing complexity information. Finally, structured semantic understanding information is generated.

[0033] Step 103: Perform multi-dimensional elimination judgment and resource demand mapping processing on the semantic understanding information to generate event nature and resource demand information.

[0034] In some embodiments, the aforementioned executing entity can perform multi-dimensional elimination judgment and resource demand mapping processing on the aforementioned semantic understanding information to generate event nature and resource demand information. This event nature and resource demand information includes event priority information generated based on big data. The multi-dimensional elimination judgment can be a comprehensive assessment based on a preset assessment rule base, a historical event database, and semantic understanding information to determine whether to eliminate a case. For example, regarding the complaint of "garbage accumulation in the community," matching the preset assessment rule base reveals that it is the responsibility of the sanitation company (not the direct responsibility of the street / township), and a preliminary judgment is made that it "will not be included in the assessment." Then, querying the historical event database reveals similar issues requiring coordination with the sanitation department, and the judgment is revised to "handled on credit." Finally, based on the semantic understanding information, the event nature label "handled on credit" is generated, and the resource demand is mapped to "2 man-days." The aforementioned event nature and resource demand information can be structured information that integrates event nature, quantifies resource demand, and event priority information. The aforementioned resource demand can be a quantitative description of the resources required to handle urban user events. For example, the aforementioned event nature and resource demand information could be: {Nature: Municipal emergency repair, Demand: {1 engineering vehicle, Priority: High}}. The aforementioned big data can be a complete collection of urban event data, including real-time events, historical handling, spatiotemporal attributes, system operation, and other multi-dimensional data. For example, the aforementioned big data could be the complete collection of urban complaints over the past year. The aforementioned event priority information can be a ranking of event handling based on multi-dimensional data, such as marking missing manhole covers as "Level 1 Priority" and street vendors as "Level 2 Priority".

[0035] In some optional implementations of certain embodiments, the aforementioned execution entity can perform multi-dimensional elimination judgment and resource demand mapping processing on the aforementioned semantic understanding information to generate event nature and resource demand information. The event nature and resource demand information includes: event priority information generated based on big data, which may include the following steps: The first step involves matching the event nature classification labels and core demand vectors from the semantically understood information with a pre-defined assessment rule base to generate a preliminary elimination result. This preliminary elimination result can be the initial elimination result obtained after matching the event semantic information with the assessment rules. In practice, firstly, the event nature classification labels (e.g., "noise disturbance") and core demand vectors are extracted from the semantically understood information. Then, they are matched with a pre-defined assessment rule base (e.g., "complaints about nighttime construction noise must be responded to within 2 hours"). Finally, a preliminary elimination result is generated.

[0036] The second step involves revising the initial exclusion judgment based on the aforementioned historical event database to generate an enhanced judgment. This enhanced judgment can be derived by modifying the initial result using historical data. In practice, firstly, the revision logic of similar judgment cases in the historical event database is retrieved. Then, the rationality of the initial exclusion judgment is compared and analyzed. Finally, deviations are corrected to obtain the enhanced judgment. For example, the initial judgment might be "a tree has fallen." However, a historical search reveals that similar events on non-municipal roads were handled by the property owner, thus the enhanced judgment is "exclusion."

[0037] The third step involves determining the event nature label for each event based on the enhanced judgment results and the pre-defined assessment rule base, thus obtaining an event nature label set. This event nature label set can be a set of standard labels used to ultimately characterize the event category and its handling status. For example, the event nature label could be "municipal facility malfunction -> missing manhole cover". In practice, firstly, valid handling events are selected based on the enhanced judgment results. Then, standardized nature labels are applied to the valid events using the assessment rule base. Finally, the various labeling results are integrated to obtain the event nature label set.

[0038] The fourth step involves generating quantitative resource requirement indicators for each event based on the event nature tag set and the estimated processing complexity information from the semantic understanding information. These quantitative resource requirement indicators include estimated processing time and computational resource consumption level. The quantitative resource requirement information can be a numerical description of the types and quantities of resources required to process the event. For example, the quantitative resource requirement information could be {estimated processing time: 4 hours, computational resources: 2 GPUs, coordination departments: 2}. The estimated processing time can be the predicted working time required to complete the event handling. The computational resource consumption level can be the predicted level of computational power requirements for the corresponding computational task. For example, for the "bridge crack assessment" task involving 3D modeling, the computational resource consumption level is 5 (the highest level). In practice, firstly, basic resource types are matched from the resource template library based on the "event nature tag set". Then, the resource quantity is quantitatively estimated using parameters from the "estimated processing complexity information" (e.g., problem size, impact range). Finally, the "quantitative resource requirement information" containing specific numerical values ​​is output.

[0039] The fifth step involves generating event priority information based on the aforementioned quantitative resource demand information, the event spatiotemporal attribute information from the aforementioned big data, and the real-time operational status data from the aforementioned big data. The aforementioned event spatiotemporal attribute information can include the time and geographical location of the event. The aforementioned real-time operational status data can be the city's current dynamic operational parameters, such as the current emergency team attendance rate (85%), the inventory of repair materials, and the traffic congestion index. In practice, firstly, the "quantitative resource demand information," "event spatiotemporal attribute information" (e.g., whether it occurs during a sensitive period), and "real-time operational status data" (e.g., current resource busyness) are collected. Then, these features are input into a pre-trained priority prediction model (e.g., a gradient boosting tree-based prediction model). Finally, priority information for event handling is generated.

[0040] The sixth step involves structurally integrating the aforementioned event nature tag set, quantitative resource requirement information, and event priority information to obtain structured event nature and resource requirement information. In practice, first, a standard data structure (e.g., JSON Schema) is established. Then, the "event nature tag set," "quantitative resource requirement information," and "event priority information" are used as different fields and filled into this structure. Finally, a "structured event nature and resource requirement information" object is generated.

[0041] In some optional implementations of certain embodiments, the execution entity may generate event priority information based on the aforementioned quantitative resource requirement information, the event spatiotemporal attribute information in the aforementioned big data, and the real-time running status data in the aforementioned big data. This may include the following steps: The first step is to extract a priority-related feature set from the aforementioned big data, based on the quantitative resource demand information, event spatiotemporal attribute information, and real-time operational status data. This priority-related feature set can be a set of features selected from the big data that are highly correlated with judging the urgency of the event and the order of processing. In practice, firstly, the three types of data—quantitative resource demand, event spatiotemporal attributes, and real-time operational status—are integrated. Then, priority-related features for each dimension are extracted from the big data. Finally, these features are integrated to form a standardized priority-related feature set. For example, by integrating data on missing manhole covers, features such as 2 working hours, morning peak hours, and a team attendance rate of 85% are extracted to form the feature set for this event.

[0042] The second step involves inputting the aforementioned priority-related feature set into a pre-trained priority prediction model to generate event priority information. This pre-trained priority prediction model can be a machine learning model (e.g., a gradient boosting tree model) trained on historical scheduling data to predict the urgency of new events. The input to this priority prediction model can be the priority-related feature set, and the output can be event priority information. The structure of this priority prediction model may include an input layer (input is the priority-related feature set, output is a feature vector set), a hidden layer (using the ReLU activation function for non-linear transformation to extract deep features, input is the aforementioned feature vector set, output is a fused feature vector), and an output layer (using the Softmax function, input is the fused feature vector, output is event priority information).

[0043] Step 104: Based on the nature of the events and resource demand information, generate a list of effective governance events, and construct a corresponding resource demand heatmap based on the list of effective governance events.

[0044] In some embodiments, the aforementioned implementing entity can generate a list of effective governance events based on the nature of the events and resource demand information, and construct a corresponding resource demand heatmap based on the list of effective governance events. The list of effective governance events can be an ordered set of events included in the actual handling process. The effective governance events can be urban user events that have been confirmed and included in the actual handling process. For example, an effective governance event could be a citizen complaint about "garbage accumulation in XX community," which, after being determined not to be excluded, is included in the actual event handling process. The resource demand heatmap can be a visual layer on a geographic base map that uses color depth to intuitively display the intensity and concentration of resource demand in different areas. For example, the city center area on the map is displayed as a dark red "hotspot," indicating that this area has a large number of high resource demand events.

[0045] In addressing the technical challenges mentioned above, the application scenario—urban grid-based integrated governance and emergency command and dispatch—requires rapid processing of massive and dynamic citizen demands. This often presents the following technical problems: traditional methods struggle to quickly identify regional issues, resource demand clusters, and evolving trends from a large volume of discrete events, leading to low resource allocation efficiency, delayed responses, and wasted coordination resources. Considering the following requirements for this application scenario—macro-level situation visualization, demand aggregation analysis, dynamic updates, and spatial positioning—we have decided to adopt the following solution: In some optional implementations of certain embodiments, the aforementioned executing entity may generate a list of valid governance events based on the aforementioned event nature and resource requirement information, and construct a corresponding resource requirement heatmap based on the aforementioned list of valid governance events, which may include the following steps: The first step is to filter the real-time event data stream based on the event nature tags in the event nature and resource requirement information to obtain an initial valid event set. This initial valid event set can be a preliminary, unsorted collection of events selected solely based on event nature tags. In practice, first, the "event nature tags" in the "event nature and resource requirement information" are read to identify events tagged "included in assessment" or "on account." Then, based on the IDs of these events, the corresponding complete event records are extracted from the "real-time event data stream." Finally, all extracted events are aggregated to form the initial valid event set.

[0046] The second step involves sorting the events in the initial valid event set based on their nature and resource requirements, generating a valid event governance list with priority information. This valid event governance list can be the final list of events to be processed, formed by sorting the initial valid event set by priority. In practice, first, the priority value of each event in the initial set is determined (e.g., P0 > P1 > P2). Then, all events are sorted in descending order of priority. Finally, the valid event governance list is generated.

[0047] The third step involves clustering the events in the effective event governance list based on their geographical location and resource requirement similarity to generate resource requirement clusters. These resource requirement clusters can be grouped together from events with similar resource requirements and geographical proximity. In practice, first, the geographical location and resource requirement characteristics (e.g., required working hours, equipment type) of the events in the effective event governance list are extracted. Then, a clustering algorithm (e.g., DBSCAN) is used to group events with similar geographical locations and resource requirement characteristics into one category. Finally, multiple resource requirement clusters are generated. For example, five events in a certain area that both require "pipe dredging" and "large pumping equipment" can be clustered into an "emergency drainage requirement cluster."

[0048] The fourth step involves weighting each resource demand cluster in the aforementioned resource demand cluster set to generate a resource demand cluster set with weighted information. In practice, firstly, a total weight is determined for each resource demand cluster. Then, this total weight can be calculated based on the number of events within the cluster, the total priority score, or the total estimated working hours. Finally, a weight value is assigned to each cluster. For example, a cluster containing 3 P0-level events has a greater weight than a cluster containing 5 P2-level events.

[0049] The fifth step involves rasterizing the geographic region corresponding to the real-time event data stream to obtain a standardized grid set. This standardized grid set can be a collection of regularly sized grid cells dividing the geographic region corresponding to the real-time event data stream. In practice, first, the geographic region corresponding to the real-time event data stream is determined. Then, the grid size is set (e.g., 1 square kilometer). Finally, the region is completely covered with a set of regular square grid cells, each with a unique ID and geographic extent, forming a standardized grid set.

[0050] The sixth step involves mapping the aforementioned resource demand clusters to the aforementioned grid set to generate a grid demand distribution map. This grid demand distribution map can be a numerical distribution map of demand values ​​formed by allocating the total (weighted) demand of each resource demand cluster to its respective grid. In practice, firstly, each resource demand cluster is mapped to its specific grid based on the geographical location of its event. Then, the weight values ​​of all clusters within a grid are summed to obtain the total demand value for that grid. Finally, a distribution map is generated where each grid corresponds to a single demand value, serving as the grid demand distribution map.

[0051] Step 7: Smooth the above-mentioned grid demand distribution map to generate a smooth demand density surface. This smooth demand density surface can be a surface model that smooths the grid distribution map, allowing demand intensity to transition continuously in space. In practice, firstly, the data in the grid distribution map is discrete (one value per grid). Then, a spatial interpolation algorithm (e.g., Kriging interpolation) is used to estimate the regions between grids, ensuring a continuous and smooth spatial transition of demand values. Finally, the smooth demand density surface is generated.

[0052] Step 8: Perform color gradient rendering on the smoothed demand density surface to generate a visual heatmap layer. This visual heatmap layer can be an overlayable map layer generated by rendering the smoothed density surface using a color gradient. In practice, first, a color mapping rule is defined (e.g., blue for low demand, red for high demand). Then, a corresponding color is assigned to each point based on its smoothed demand density value. Finally, a semi-transparent color layer is generated on the base map to serve as the visual heatmap layer.

[0053] The ninth step involves overlaying and merging the aforementioned visualization heatmap layers to generate a resource demand heatmap. In practice, this is done first, based on the corresponding base geographic information map (e.g., a street map). Then, the generated visualization heatmap layer is precisely overlaid onto the base map and blended for rendering. Finally, a resource demand heatmap is generated, where darker areas represent more urgent resource demands.

[0054] The above-described operational steps, combined with step 106, constitute an inventive point of this disclosure, solving the technical problem mentioned in the background art: "Traditional methods struggle to quickly identify regional problems, resource demand clusters, and evolution trends from massive discrete events, leading to low resource scheduling efficiency, delayed response, and wasted coordination and scheduling resources." The reasons for these technical problems are as follows: data is discrete and isolated, lacking spatial aggregation analysis; it is impossible to intuitively and in real-time display the distribution and evolution of demands; and it is difficult to quantify the urgency of demands in different regions and of different types. This invention, by clustering effective events according to geographical location and demand similarity, and weighting and mapping them into a visual heatmap, achieves a macroscopic, real-time, and quantitative visual perception of urban governance resource demands, saving the time cost required for coordinating resource allocation.

[0055] Step 105: Based on the resource demand heatmap and the preset resource mapping library, generate a dynamic resource allocation instruction set with priority information.

[0056] In some embodiments, the aforementioned executing entity can generate a dynamic resource allocation instruction set containing priority information based on the aforementioned resource demand heatmap and a preset resource mapping library. The preset resource mapping library can be a database recording the responsibilities, resource (personnel, vehicles, materials) types, quantities, locations, and statuses of various departments. For example, the preset resource mapping library could record "District Sanitation Center: 10 water trucks (Location: XX garage, 8 available)". The aforementioned dynamic resource allocation instruction set can be action commands calculated in real-time based on the generated resource demand heatmap and issued to currently available resource units (e.g., idle water trucks of the sanitation department, on-duty emergency repair teams, available material pallets in warehouses, and elastically allocable GPU computing instances on the cloud platform), including specific execution time, location, and priority. For example, the above dynamic resource allocation instruction set could be: {Instruction 1: Object: Sanitation department water truck, Destination: [Longitude 116.1, Latitude 39.3] (corresponding to the center of the "Dust Pollution" hotspot area in the resource demand heatmap), Task: Perform 30 minutes of road dust suppression operation, Start time: 14:45, Priority: High}, {Instruction 1: Object: Street emergency repair team (Group: Water and Electricity), Destination: Building 3 of Sunshine Community (corresponding to the "Water and Electricity Repair" hotspot location in the resource demand heatmap), Task: Handle water pipe burst, Start time: Immediate, Priority: Emergency}, {Instruction 3, Object: Cloud computing platform, Action: Allocate resources, Specification: 2 GPU instances, Purpose: Run video analysis model to process traffic congestion video streams in hotspot areas, Duration: 1 hour, Priority: Medium}.

[0057] In some optional implementations of certain embodiments, the execution entity may generate a dynamic resource allocation instruction set containing priority information based on the resource demand heatmap and a preset resource mapping library, which may include the following steps: The first step, based on the aforementioned resource demand heatmap, is to determine the corresponding resource demand characteristic information. This information includes: resource demand intensity information, spatial distribution pattern information, and trend information. The resource demand characteristic information can be quantitative features extracted from the heatmap to guide resource matching, such as {Intensity: High (peak demand 90), Pattern: Linear clustering (distributed along main roads), Trend: Rapid increase (demand increased by 50% in the past hour)}. The resource demand intensity information can be a numerical value quantifying the size of resource demand in a specific area. The spatial distribution pattern information can describe the clustering characteristics of resource demand in geographic space, such as patterns like "point bursts" (single hotspots), "band-like distribution" (along rivers), or "area diffusion" (the entire area). The trend information can describe the rate and direction of change in resource demand over time. Examples: trends such as "sharp increase," "stable," or "slow decrease." In practice, firstly, image analysis is performed on the "resource demand heatmap" to extract the average pixel intensity of each hotspot area as resource demand intensity information. Then, morphological analysis is used to identify the shapes of hotspot areas (e.g., circles, stripes) to determine spatial distribution patterns. Finally, by comparing the current heatmap with historical heatmaps (six days prior to the previous period), the demand change gradient is calculated to obtain trend characteristics.

[0058] The second step involves matching the aforementioned resource demand characteristics with the pre-defined resource mapping library to generate matching results. These matching results can be potential resource supply solutions derived from comparing the demand characteristics with the resource library. For example, the matching result could be "high-intensity linear demand" matched with "mobile patrol rapid response teams." In practice, firstly, the extracted resource demand characteristics are compared with the service capability characteristics (e.g., coverage, response speed, processing capacity) of each resource record in the pre-defined resource mapping library to calculate similarity. Then, for each demand hotspot area, several resource records with the highest similarity are selected from the library. Finally, these pairing relationships are output as the matching results.

[0059] The third step involves generating a preliminary resource allocation plan based on the matching results. This plan includes resource type information, quantity estimation information, and target location information. Specifically, the preliminary resource allocation plan can be a preliminary suggestion of resource type, quantity, and target location based on the matching results. The resource type information can be the specific type of resource required, such as "high-pressure water pump" or "aerial work vehicle." The quantity estimation information can be a preliminary estimate of the quantity of each type of resource required, such as "cable: 100 meters." The target location information can be the geographical location where the resource needs to be dispatched. In practice, firstly, based on the matching results, the "resource type information" (e.g., fire truck, ambulance) that needs to be dispatched is determined. Then, based on the demand intensity information and spatial distribution pattern information, the quantity estimation information of each type of resource required is determined. Finally, the geographical center or coverage area of ​​the demand hotspot is used as the target location information to form the preliminary resource allocation plan.

[0060] The fourth step involves integrating and optimizing the event priority information from the effective governance event list with the preliminary resource allocation plan to generate an optimized resource allocation scheme. This optimized scheme can be a more refined and rational resource allocation and routing plan formed by integrating event priorities. In practice, firstly, based on the "effective governance event list," detailed "event priority information" and its geographical location are obtained for each event in the list. Then, resources from the preliminary resource allocation plan are prioritized for allocation to the highest-priority events located within the target area of ​​the plan. Finally, the proportion and order of resource allocation are adjusted according to priority, resulting in the optimized resource allocation scheme.

[0061] The fifth step involves determining the dynamic scheduling strategy based on the optimized resource allocation plan and real-time load information. The real-time load information can be the current task load and availability status of each department and resource unit, such as "Available sanitation vehicles: 15," or "Current computing cluster load rate: 65%." The dynamic scheduling information can be a strategy that includes specific execution logic, determined after comprehensively considering the plan and real-time load. In practice, first, the system's real-time load information (e.g., the availability of each resource pool) is obtained. Then, based on the requirements of the optimization plan and the real-time load, the specific scheduling strategy is determined. Finally, the strategy information is generated. For example, if the real-time load shows that a certain resource pool is under strain, the dynamic scheduling strategy information might be determined as "batch scheduling" and "activating backup resource points."

[0062] The sixth step is to transform the aforementioned dynamic scheduling policy information into an initial instruction set. This initial instruction set can be a set of raw operation instructions, not yet embedded with priorities, derived from the dynamic scheduling policy information. In practice, first, each policy in the dynamic scheduling policy is parsed. Then, each policy is transformed into one or more specific operation instructions that can be understood by the underlying execution system. Finally, the initial instruction set is formed.

[0063] Step 7: Embed the aforementioned event priority information into the initial instruction set to generate a dynamic resource allocation instruction set with priority information. In practice, first, the priority of the events related to the instruction is obtained. Then, the priority is added as a metadata field to each initial instruction. Finally, the final dynamic resource allocation instruction set is generated. For example, the final instruction is: {Instruction: Dispatch vehicle to G1001, Resource: Water truck, Quantity: 1, Priority: P0}.

[0064] Step 106: Based on the dynamic resource allocation instruction set, schedule the corresponding cross-modal resources to execute the governance operations for the corresponding city user events.

[0065] In some embodiments, the aforementioned executing entity can schedule corresponding cross-modal resources according to the aforementioned dynamic resource allocation instruction set to execute governance operations for corresponding urban user events. The aforementioned cross-modal resources include at least one of the following: computing resources, transmission resources, and material reserve resources. These cross-modal resources can be resource categories with vastly different properties and forms of existence, but which need to be used collaboratively in governance tasks. For example, cloud computing servers for analyzing events, government dedicated networks for transmitting instructions, and emergency supplies for on-site handling. The aforementioned governance operations can be specific actions performed to respond to and resolve urban user events, such as dispatching sanitation vehicles for cleaning or arranging repair teams to repair facilities. The aforementioned computing resources can refer to computing power used for data processing and model running, such as cloud virtual machines (VMs) and graphics processing unit (GPU) clusters. The aforementioned transmission resources can be network communication capabilities used for data transmission, such as 5G network slicing and government dedicated network bandwidth. The aforementioned material reserve resources can be physical materials and transportation capacity used for physical handling, such as sandbags, generators, and transport vehicles.

[0066] In addressing the technical challenges of the aforementioned background technologies, and considering the specific application scenario—urban emergency response and joint handling of complex events such as floods and large-scale event support—which requires simultaneous allocation of computing power for analysis, communication for command, and materials and personnel for on-site operations, the following technical issues often arise: cross-modal heterogeneous resources (e.g., computing resources, transmission resources, material reserve resources) belong to different systems, and the scheduling instructions and execution systems are fragmented, leading to difficulties in cross-modal resource collaboration, lengthy response chains, excessively long response times, and low efficiency in handling urban users' time. Based on the following requirements for this application scenario: unified instructions and collaboration, differentiated execution, and process monitorability, we have decided to adopt the following solution: In some optional implementations of certain embodiments, the aforementioned execution entity can schedule corresponding cross-modal resources according to the aforementioned dynamic resource allocation instruction set to perform governance operations for corresponding city user events. The aforementioned cross-modal resources include at least one of the following: computing resources, transmission resources, and material reserve resources, and may include the following steps: The first step, based on the event priority information and resource type identifier in the above dynamic resource allocation instruction set, is to perform the following differentiated scheduling operations: In the first sub-step, in response to the existence of a computing resource identifier in the aforementioned resource type identifier, the estimated complexity information is processed centrally based on the aforementioned dynamic resource allocation instructions, and computing nodes with corresponding computing power are scheduled to construct a distributed computing link. The aforementioned computing resource identifier can be a marker in the instructions used to specify the required computing resource type. The aforementioned distributed computing link can be a computing cluster dynamically constructed for processing complex tasks, consisting of multiple computing nodes working collaboratively through a network. For example, the aforementioned distributed computing link could be 10 GPU servers interconnected through a high-speed RDMA network to form a computing cluster for processing city user events. In practice, firstly, if a "computing resource identifier" exists in the resource type identifier, the associated estimated processing complexity information is determined. Then, based on the complexity level (e.g., high, medium, low), nodes with matching computing power (e.g., GPU model, memory size) are selected from the cloud computing resource pool or edge computing nodes. Finally, a data pipeline or task parallel link is established between these nodes to form a distributed computing link.

[0067] Sub-step two: In response to the existence of a transmission resource identifier in the aforementioned resource type identifier, based on the target location coordinate set and event priority information in the aforementioned dynamic resource allocation instruction set, differentiated network bandwidth and transmission paths are configured for nodes at different levels to establish a cross-level data transmission channel. The aforementioned transmission resource identifier can be a marker in the instruction used to specify the required network service quality and type. The aforementioned transmission path can be the sequence of network nodes and links that data traverses from source to destination. The aforementioned cross-level data transmission channel can be an end-to-end data path established across different administrative or network levels (e.g., city -> district -> site). In practice, firstly, if a "transmission resource identifier" exists in the resource type identifier, the target location coordinate set and event priority information are determined. Then, for nodes at different levels such as city, district, and site, differentiated network bandwidth (e.g., ensuring high bandwidth for on-site high-definition video backhaul) and transmission paths (e.g., selecting low-latency leased lines) are configured according to their location and priority. Finally, an end-to-end cross-level data transmission channel is established.

[0068] Step three involves responding to the existence of a "reserve resource identifier" in the aforementioned resource type identifier. Based on the resource type information, quantity estimation information, and target location information in the aforementioned dynamic resource allocation instruction set, a material allocation and transportation capacity coordination plan is generated. The aforementioned "reserve resource identifier" can be a marker in the instruction used to specify the required material type and standards. The aforementioned material allocation and transportation capacity coordination plan can be detailed information regarding material outbound, transportation, and delivery. For example, the aforementioned material allocation and transportation capacity coordination plan could be: transporting 50 tents from warehouse X to resettlement point Y, undertaken by transportation company Z, via route R. In practice, firstly, if a "reserve resource identifier" exists in the resource type identifier, the resource type information, quantity estimation information, and target location information are determined. Then, the material inventory database and transportation capacity pool are queried to generate a detailed outbound list, loading plan, transportation route, and responsible unit. Finally, an executable material allocation and transportation capacity coordination plan is generated.

[0069] The second step involves generating a resource collaboration logic diagram based on the aforementioned distributed computing link, cross-level data transmission channel, and material allocation and transportation coordination scheme. This resource collaboration logic diagram can be a flowchart describing the dependencies and timing of cross-modal resources during task execution. For example, it could show a process of "analyzing data first -> issuing instructions -> finally allocating materials." In practice, firstly, the generated distributed computing link, cross-level data transmission channel, and material allocation and transportation coordination scheme are designated as nodes. Then, the data dependencies, timing relationships, and logical constraints between them (e.g., "some instructions cannot be issued before the calculation and analysis are completed") are determined, and these nodes are connected with directed edges. Finally, a resource collaboration logic diagram describing the overall resource collaboration relationships is generated.

[0070] The third step involves generating a resource collaboration execution workflow based on the aforementioned resource collaboration logic diagram and the event priority information in the dynamic resource allocation instruction set. This workflow can be a sequence of specific, executable tasks that the resource collaboration logic diagram can be transformed into, and automatically scheduled, executed, and monitored by the workflow engine. For example, the workflow could be defined in Airflow as: Task 1 (Initiate risk control model calculation) -> Task 2 (Generate risk report and push it) -> Task 3 (Trigger high-risk area material scheduling). In practice, firstly, the "resource collaboration logic diagram" is transformed into a process definition (e.g., a DAG graph) that can be recognized by a workflow engine (e.g., Apache Airflow). Then, the event priority information in the dynamic resource allocation instruction set is injected into the engine as the execution priority parameter for the entire workflow. Finally, the resource collaboration execution workflow is generated.

[0071] The fourth step involves executing the corresponding resource scheduling operations in parallel based on the resource collaborative execution workflow described above, to complete the governance operations for the corresponding city user events. In practice, this is done by first breaking down the parallel task nodes of the resource collaborative execution workflow; then, synchronously executing cross-modal resource scheduling operations on a node-by-node basis; and finally, completing the scheduling and implementing the city user event governance operations.

[0072] The above-described operational steps, as an inventive point of this disclosure, solve the technical problem mentioned in the background art: "Cross-modal heterogeneous resources (e.g., computing resources, transmission resources, and material reserve resources) belong to different systems, and the scheduling instructions and execution system are fragmented, resulting in difficulties in cross-modal resource coordination, lengthy response chains, excessively long response times, and low efficiency in processing urban users' time." The reasons for the above technical problems are as follows: In emergency situations, there is a lack of a standard interface for unified digital description and scheduling of various core resources; traditional scheduling relies on manual coordination, which cannot meet the timeliness requirements of multi-task concurrency; and the execution status of heterogeneous resources is scattered, lacking global visualization and monitoring methods. This invention, by automatically parsing emergency instructions and driving the generation of collaborative workflows oriented towards computing, transmission, and materials, achieves rapid, parallel, and integrated scheduling and deployment of cross-modal core resources in the early stages of an emergency, saving emergency response time and reducing resource waste caused by poor coordination.

[0073] The above-described embodiments of this disclosure have the following beneficial effects: Through the resource scheduling method for urban user events based on big data, as described in some embodiments of this disclosure, precise resource scheduling and efficient cross-modal resource allocation are achieved in the urban user event governance process, reducing the time cost of event processing. Specifically, the reasons for low resource scheduling accuracy, low cross-modal resource allocation efficiency, and excessively high time costs are: a lack of ability to intelligently analyze, dynamically prioritize, and collaboratively map massive event data, resulting in resource mismatch and response delays. Based on this, the resource scheduling method for urban user events based on big data, as described in some embodiments of this disclosure, firstly, in response to receiving processing information for urban user events, preprocesses the collected real-time event data stream related to urban user events, the preset assessment rule base, and the historical event database to obtain a preprocessed dataset. Transforming the real-time event stream, the preset assessment rule base, and the historical event database into a well-organized preprocessed dataset provides a unified and clean data foundation for subsequent intelligent analysis, solving the problem of analytical accuracy caused by the messiness and inconsistent formats of the original data. Then, deep semantic parsing and multi-label classification are performed on the preprocessed dataset to generate semantic understanding information, which includes event nature classification labels, core demand vectors, and estimated processing complexity information. Extracting this semantic understanding information overcomes the limitations of traditional keyword matching, enabling accurate identification of complex event natures. Next, multi-dimensional elimination judgment and resource demand mapping are performed on the semantic understanding information to generate event nature and resource demand information, which includes event priority information generated based on big data. Combining rule matching and historical case correction, a judgment result including quantified resource demand and event priority information is generated, deeply integrating business logic (whether to eliminate) with resource planning (how much resource is needed), achieving a leap from event identification to resource demand prediction. Furthermore, based on the event nature and resource demand information, a list of effective governance events is generated, and a corresponding resource demand heatmap is constructed based on this list. Effective governance events are sorted based on event priority information, and the resource demand is spatially visualized using the resource demand heatmap. Next, based on the aforementioned resource demand heatmap and pre-defined resource mapping library, a dynamic resource allocation instruction set with priority information is generated. Combining the visualized resource demand heatmap with the pre-defined resource mapping library generates a specific, executable, and priority-embedded dynamic resource allocation instruction set, transforming the scheduling strategy into precise operational commands that drive the coordinated action of various resources. Finally, according to the aforementioned dynamic resource allocation instruction set, corresponding cross-modal resources are scheduled to execute governance operations for corresponding city user events. These cross-modal resources include at least one of the following: computing resources, transmission resources, and material reserve resources.By scheduling cross-modal resources according to the dynamic resource allocation instruction set, the collaborative and efficient allocation of multiple types of resources can be achieved, eliminating resource mismatch and response delay issues, accurately matching the event governance needs of urban users, and significantly reducing the time cost of event processing.

[0074] Further reference Figure 2 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a resource scheduling device based on big data and oriented towards urban user events. These device embodiments are similar to... Figure 1 Corresponding to the method embodiments shown, this big data-based resource scheduling device for urban user events can be specifically applied to various electronic devices.

[0075] like Figure 2 As shown, a resource scheduling device 200 for urban user events based on big data includes: a preprocessing unit 201, a parsing and classification unit 202, a mapping unit 203, a first generation unit 204, a second generation unit 205, and a scheduling unit 206. The preprocessing unit 201 is configured to: in response to receiving processing information for urban user events, preprocess the collected real-time event data stream related to urban user events, a preset assessment rule base, and a historical event database to obtain a preprocessed dataset. The parsing and classification unit 202 is configured to perform deep semantic parsing and multi-label classification processing on the preprocessed dataset to generate semantic understanding information, wherein the semantic understanding information includes: event nature classification labels, core demand vectors, and processing estimated complexity information. The mapping unit 203 is configured to: perform multi-dimensional elimination judgment and resource demand mapping processing on the semantic understanding information to generate event nature and resource demand information, wherein the event nature and resource demand information includes: event priority information generated based on big data. The first generation unit 204 is configured to generate a list of valid governance events based on the aforementioned event nature and resource demand information, and to construct a corresponding resource demand heatmap based on the aforementioned list of valid governance events. The second generation unit 205 is configured to generate a dynamic resource allocation instruction set with priority information based on the aforementioned resource demand heatmap and a preset resource mapping library. The scheduling unit 206 is configured to schedule corresponding cross-modal resources according to the aforementioned dynamic resource allocation instruction set to execute governance operations for corresponding city user events, wherein the aforementioned cross-modal resources include at least one of the following: computing resources, transmission resources, and material reserve resources.

[0076] It is understandable that the units and references recorded in this big data-based resource scheduling device 200 for urban user events... Figure 1The steps in the described method correspond accordingly. Therefore, the operations, features, and beneficial effects described above for the method are also applicable to the resource scheduling device 200 based on big data and oriented towards urban user events, and the units contained therein, and will not be repeated here.

[0077] The following is for reference. Figure 3 It shows a schematic diagram of the structure of an electronic device (e.g., an electronic device) 300 suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0078] like Figure 3 As shown, the electronic device 300 may include a processing unit (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 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0079] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0080] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0081] It should be noted that, in some embodiments of this disclosure, the computer-readable medium described above may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0082] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0083] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: respond to receiving processing information regarding urban user events; preprocess the collected real-time event data stream related to urban user events, the preset assessment rule base, and the historical event database to obtain a preprocessed dataset; perform deep semantic parsing and multi-label classification processing on the preprocessed dataset to generate semantic understanding information, wherein the semantic understanding information includes: event nature classification labels, core demand vectors, and processing estimated complexity information; and perform multi-dimensional elimination judgment and resource demand mapping on the semantic understanding information. The system generates event nature and resource demand information, which includes: event priority information generated based on big data; a list of effective governance events is generated based on the event nature and resource demand information, and a corresponding resource demand heatmap is constructed based on the list of effective governance events; a dynamic resource allocation instruction set with priority information is generated based on the resource demand heatmap and a preset resource mapping library; and corresponding cross-modal resources are scheduled according to the dynamic resource allocation instruction set to execute the governance operations of the corresponding city user events, wherein the cross-modal resources include at least one of the following: computing resources, transmission resources, and material reserve resources.

[0084] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0085] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0086] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor may be described as including a preprocessing unit, a parsing and classification unit, a mapping unit, a first generation unit, a second generation unit, and a scheduling unit. The names of these units do not necessarily limit the specific unit itself; for example, the preprocessing unit may also be described as "a unit that, in response to receiving processing information related to urban user events, preprocesses the collected real-time event data stream related to urban user events, a preset assessment rule base, and a historical event database to obtain a preprocessed dataset."

[0087] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0088] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A resource scheduling method based on big data and oriented towards urban user events, comprising: In response to receiving processing information for city user events, the system preprocesses the collected real-time event data streams related to city user events, the preset assessment rule base, and the historical event database to obtain a preprocessed dataset. Deep semantic parsing and multi-label classification are performed on the preprocessed dataset to generate semantic understanding information, wherein the semantic understanding information includes: event nature classification labels, core appeal vectors, and processing estimated complexity information; The semantic understanding information is subjected to multi-dimensional elimination judgment and resource demand mapping processing to generate event nature and resource demand information, wherein the event nature and resource demand information includes: event priority information generated based on big data; Based on the nature of the events and the resource demand information, a list of effective governance events is generated, and a corresponding resource demand heatmap is constructed based on the list of effective governance events. Based on the resource demand heatmap and the preset resource mapping library, a dynamic resource allocation instruction set with priority information is generated. According to the dynamic resource allocation instruction set, the corresponding cross-modal resources are scheduled to perform the governance operations of the corresponding city user events. The cross-modal resources include at least one of the following: computing resources, transmission resources, and material reserve resources.

2. The method according to claim 1, wherein, The response to receiving processing information for city user events involves preprocessing the collected real-time event data stream related to city user events, the preset assessment rule base, and the historical event database to obtain a preprocessed dataset, including: The real-time event data stream is updated based on a preset elimination rule to obtain the original heterogeneous data stream. The original heterogeneous data stream is cleaned to obtain a cleaned event data stream; The cleaned event data stream is associated with relevant cases in the historical event database to construct an event association graph; The event association graph is fused with the preset assessment rule base to generate a rule-enhanced dataset; The textual information, spatiotemporal information, and rule label information in the rule-enhanced dataset are encoded to generate a preprocessed dataset.

3. The method according to claim 1, wherein, The process of performing deep semantic parsing and multi-label classification on the preprocessed dataset to generate semantic understanding information includes: Based on the text information in the preprocessed dataset, generate the core semantic vector of the event; Based on the core semantic vector of the event, generate event nature classification labels and core demand semantic vectors; Based on the event nature classification labels and the core demand semantic vectors, a pre-trained complexity prediction model is used to generate processing estimated complexity information. The event nature classification labels, the core demand semantic vectors, and the processing estimated complexity information are integrated into semantic understanding information.

4. The method according to claim 1, wherein, The process of performing multi-dimensional elimination and resource requirement mapping on the semantic understanding information to generate event nature and resource requirement information includes: The event nature classification labels and core demand vectors in the semantic understanding information are matched with a preset assessment rule base to generate preliminary elimination judgment results. Based on the historical event database, the preliminary elimination judgment result is revised to generate an enhanced judgment result; Based on the enhanced judgment results and the preset assessment rule base, the event nature label of each event is determined to obtain the event nature label set; Based on the event nature tag set and the processing estimated complexity information in the semantic understanding information, a quantitative resource requirement index is generated for each event to obtain quantitative resource requirement information. The quantitative resource requirement index includes: estimated processing time and computing resource consumption level. Event priority information is generated based on the quantified resource demand information, the event spatiotemporal attribute information in the big data, and the real-time operating status data in the big data; The event nature label set, the quantitative resource requirement information, and the event priority information are structurally integrated to obtain structured event nature and resource requirement information.

5. The method according to claim 4, wherein, The event priority information generated based on the quantified resource demand information, the event spatiotemporal attribute information in the big data, and the real-time operational status data in the big data includes: Based on the quantitative resource demand information, event spatiotemporal attribute information, and real-time operational status data, priority-related feature sets are extracted from the big data. The priority-related feature set is input into a pre-trained priority prediction model to generate event priority information.

6. The method according to claim 1, wherein, The step of generating a dynamic resource allocation instruction set with priority information based on the resource demand heatmap and a preset resource mapping library includes: Based on the resource demand heatmap, the corresponding resource demand characteristic information is determined, wherein the resource demand characteristic information includes: resource demand intensity information, spatial distribution pattern information, and change trend characteristic information; The resource demand feature information is matched with the preset resource mapping library to generate a matching result; Based on the matching results, a preliminary resource allocation plan is generated, which includes: resource type information, quantity estimation information, and target location information; The event priority information in the effective governance event list is integrated and optimized with the preliminary resource allocation plan to generate an optimized resource allocation plan; Based on the optimized resource allocation scheme and real-time load information, determine the dynamic scheduling strategy information; The dynamic scheduling strategy information is converted into an initial instruction set; The event priority information is embedded in the initial instruction set to generate a dynamic resource allocation instruction set containing priority information.

7. A resource scheduling device based on big data and oriented towards urban user events, comprising: The preprocessing unit is configured to, in response to receiving processing information for urban user events, preprocess the collected real-time event data stream related to urban user events, the preset assessment rule base and the historical event database to obtain a preprocessed dataset. The parsing and classification unit is configured to perform deep semantic parsing and multi-label classification on the preprocessed dataset to generate semantic understanding information, wherein the semantic understanding information includes: event nature classification labels, core appeal vectors, and processing estimated complexity information; The mapping unit is configured to perform multi-dimensional elimination judgment and resource demand mapping processing on the semantic understanding information to generate event nature and resource demand information, wherein the event nature and resource demand information includes: event priority information generated based on big data; The first generation unit is configured to generate a list of effective governance events based on the nature of the events and the resource demand information, and to construct a corresponding resource demand heatmap based on the list of effective governance events. The second generation unit is configured to generate a dynamic resource allocation instruction set with priority information based on the resource demand heatmap and a preset resource mapping library. The scheduling unit is configured to schedule corresponding cross-modal resources according to the dynamic resource allocation instruction set to perform governance operations for corresponding city user events, wherein the cross-modal resources include at least one of the following: computing resources, transmission resources, and material reserve resources.

8. An electronic device, comprising: One or more processors; Storage device, on which one or more programs are stored, When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-6.

9. A computer-readable medium having a computer program stored thereon, wherein, When the program is executed by the processor, it implements the method as described in any one of claims 1-6.

Images