Resource allocation method, apparatus, device, and medium

Abstract

The application relates to the technical field of natural language processing, and discloses a resource allocation method and device, equipment and a medium, which comprise the following steps: performing text preprocessing on an acquired initial planning text to obtain a project planning text; performing text disassembly on the project planning text by using a pre-trained semantic segmentation model to obtain a sub-activity sequence; wherein the sub-activity sequence comprises at least one sub-activity item and a time node corresponding to the sub-activity item; performing keyword extraction on each sub-activity item to obtain initial resource labels of the sub-activity items; performing resource mapping on the initial resource labels to determine resource requirement data of the sub-activity items; performing requirement optimization according to the time node and the resource requirement data to obtain a resource requirement list; and performing resource allocation according to the resource requirement list. The application can be applied to the activity planning scene in the fields of finance and insurance and medicine, and can significantly improve the accuracy and rationality of resource scheduling and allocation when planning activities.

Description

Technical Field

[0001] This invention relates to the field of natural language processing technology, and in particular to a resource allocation method, apparatus, device, and medium. Background Technology

[0002] In scenarios such as quarterly marketing, holiday promotions, anniversary celebrations, product launches, and internal training, it is often necessary to organize a large number of events simultaneously. For example, in the financial insurance sector, there are many insurance product launches, quarterly performance briefings, or customer appreciation events at the beginning of a quarter; in the medical field, there are events such as new drug clinical trial launch meetings, hospital academic exchange conferences, or remote surgeries.

[0003] Different activities typically involve multiple stages in their approval, planning, execution, and resource allocation processes, including staff arrangements, venue coordination, material preparation, and timeline management. After receiving the activity plan, staff usually need to manually read the content, further break down the activity flow, organize the activity stages, and analyze the corresponding resource requirements. However, when multiple activities are carried out simultaneously, inaccurate and unreasonable resource allocation can easily occur. Therefore, improving the accuracy and rationality of resource scheduling and allocation during activity planning has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a resource allocation method, apparatus, device, and medium to solve the technical problem of inaccurate resource allocation.

[0005] Firstly, a resource allocation method is provided, including: Obtain the initial planning text and perform text preprocessing on the initial planning text to obtain the project planning text; The project planning text is decomposed using a pre-trained semantic segmentation model to obtain a sequence of sub-activities; wherein, the sequence of sub-activities includes at least one sub-activity project and the time node corresponding to the sub-activity project; Keyword extraction is performed on each of the sub-activity items to obtain the initial resource tags for each sub-activity item; Using a pre-built project resource mapping database, resource mapping is performed on each of the initial resource tags to determine the resource requirement data for each of the sub-activity projects; Based on the time points and the resource demand data, demand optimization is performed to obtain a resource demand list; Resources are allocated based on the resource requirements list.

[0006] Secondly, a resource allocation device is provided, comprising: The text preprocessing module is used to obtain the initial planning text and perform text preprocessing on the initial planning text to obtain the project planning text; The text decomposition module is used to decompose the project planning text using a pre-trained semantic segmentation model to obtain a sub-activity sequence; wherein, the sub-activity sequence includes at least one sub-activity project and the time node corresponding to the sub-activity project; The keyword extraction module is used to extract keywords from each of the sub-activity items to obtain the initial resource tags for each of the sub-activity items. The resource mapping module is used to map resources to each of the initial resource tags using a pre-built project resource mapping database, and to determine the resource requirement data of each of the sub-activity projects. The demand optimization module is used to optimize the demand based on the time nodes and the resource demand data to obtain a resource demand list. The resource allocation module is used to allocate resources according to the resource requirement list.

[0007] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the resource allocation method described above.

[0008] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the resource allocation method described above.

[0009] The above-mentioned resource allocation method, device, computer equipment, and storage medium achieve the following: First, the initial planning text is acquired and preprocessed to obtain the project planning text. Then, a semantic segmentation model is used to automatically decompose the activity planning text, which can automatically identify different sub-activity items and corresponding time nodes from unstructured activity text, reducing comprehension bias caused by manual decomposition of activity processes and improving the efficiency of activity content parsing. At the same time, through keyword extraction and the cooperation of a resource mapping database, the resource requirement data corresponding to each sub-activity can be automatically determined, reducing omissions or errors when manually calculating resource requirements. Furthermore, resource requirements are optimized in conjunction with time nodes, enabling comprehensive analysis of resource requirements for different time periods, thereby improving the accuracy of the resource requirement list. Finally, resource allocation is performed based on the optimized resource requirement list, which can improve the accuracy of resource allocation in scenarios with multiple parallel activities. Attached Figure Description

[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a schematic diagram of an application environment for a resource allocation method according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating a resource allocation method in one embodiment of the present invention; Figure 3 yes Figure 2 A schematic diagram of a specific implementation of step S40; Figure 4 yes Figure 2 A flowchart illustrating a specific implementation of step S44; Figure 5 yes Figure 4 A schematic diagram of a specific implementation method for step S442; Figure 6 This is a schematic diagram of a resource allocation device in one embodiment of the present invention; Figure 7 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention; Figure 8 This is another structural schematic diagram of a computer device according to one embodiment of the present invention. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] The resource allocation method provided in this embodiment of the invention can be applied to, for example, Figure 1In this application environment, the client communicates with the server via a network. The server can obtain the initial planning text from the client and preprocess it to obtain the project planning text. Then, using a semantic segmentation model, the event planning text is automatically decomposed, which can automatically identify different sub-events and corresponding time nodes from unstructured event text, reducing comprehension bias caused by manual decomposition of event processes and improving the efficiency of event content parsing. At the same time, through keyword extraction and the cooperation of a resource mapping database, the resource requirement data corresponding to each sub-event can be automatically determined, reducing omissions or errors when manually calculating resource requirements. Furthermore, by combining time nodes to optimize resource requirements, resource requirements for different time periods can be analyzed in a comprehensive manner, thereby improving the accuracy of the resource requirement list. Finally, resources are allocated based on the optimized resource requirement list, which can improve the accuracy of resource allocation in scenarios with multiple parallel events. The client can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server can be implemented using a standalone server or a server cluster consisting of multiple servers. The invention will be described in detail below through specific embodiments.

[0014] Please see Figure 2 As shown, Figure 2 A flowchart illustrating a resource allocation method provided in an embodiment of the present invention includes the following steps: S10: Obtain the initial planning text and perform text preprocessing on the initial planning text to obtain the project planning text.

[0015] In one embodiment, the initial planning text is the activity plan submitted through the enterprise activity approval system, activity management platform, or office system. The initial planning text includes the activity name, activity objectives, activity process, activity time, activity location, participants, and resource requirements.

[0016] For example, the initial planning document is an insurance marketing campaign plan, as follows: Spring Festival Insurance Promotion Activities Event Date: [Date]

[0017] Activity objective: To increase sales of family insurance products during the Spring Festival.

[0018] Sub-activity 1: Online promotional activities To promote insurance products through short video platforms, we need 4 new media operators, 2 live broadcast hosts, and equipment for producing promotional materials.

[0019] Sub-activity 2: Offline promotion activities Setting up an insurance consulting booth in a shopping mall; requires 6 sales personnel, 10 sets of promotional display stands, 5000 promotional leaflets, and booth space.

[0020] In one embodiment, text preprocessing is performed on the initial planning text, including at least one of the following processing methods: performing format unification processing on the initial planning text, for example, unifying date formats, unifying time expression methods, and unifying punctuation marks; performing word segmentation processing on the initial planning text to divide continuous text into a plurality of semantic words; performing stop word filtering on the initial planning text to remove words with no actual business meaning such as "de", "le", and "yiji"; performing special character cleaning on the initial planning text to delete garbled characters, repeated spaces or invalid symbols; In one embodiment, the project planning text is standardized text data after text preprocessing. Compared with the initial planning text, the text expression in the project planning text is more unified, and redundant information and invalid characters are removed.

[0021] For example: "Spring Festival insurance promotion activity; Online promotion activity, time: XX / XX / XXXX, resource requirements: 4 new media operators, 2 live streaming personnel, and promotional equipment; Offline promotion activity, time: XX / XX / XXXX, resource requirements: 6 sales personnel, 10 sets of promotional display stands, 5000 promotional leaflets, and booth space; S20: Disassembling the project planning text by using a pre-trained semantic segmentation model to obtain a sub-activity sequence; wherein the sub-activity sequence includes at least one sub-activity item and a time node corresponding to the sub-activity item; In one embodiment, the semantic segmentation model is a text semantic analysis model constructed based on natural language processing technology, which is used to perform syntactic analysis and semantic structure recognition on the project planning text, so as to realize activity content disassembly. The semantic segmentation model is implemented by using a pre-trained deep learning model, such as a semantic analysis model based on Transformer architecture or a BERT model.

[0022] In one embodiment, the semantic segmentation model is implemented by a text semantic analysis network based on Transformer architecture, which is used to perform syntactic analysis and semantic disassembly on the project planning text, so as to identify sub-activity items and time nodes in the project planning text. Specifically, the semantic segmentation model includes: a text input layer, a text vectorization layer, a context encoding layer, a semantic feature extraction layer, a sequence labeling layer and a result output layer.

[0023] The sub-activity item is an activity execution unit disassembled from the project planning text. For example, online promotion, offline promotion, customer check-in, live streaming promotion and gift distribution, etc.

[0024] Time nodes are time information associated with sub-activities, used to characterize the execution time of sub-activities. Time nodes include activity start time, activity end time, activity execution date, or time range.

[0025] For example, consider the following project planning text: "Online promotion activities, time: 2026-01-10 to 2026-01-20; Offline promotion activities, time: 2026-01-15 to 2026-01-25."

[0026] After decomposing the text using a semantic segmentation model, we obtain: (Online promotion activities, timeframe: 2026-01-10 to 2026-01-20) and (Offline promotion activities, timeframe: 2026-01-15 to 2026-01-25).

[0027] In one embodiment, the sub-activity sequence is an ordered data set consisting of multiple sub-activity items and corresponding time nodes obtained by semantic decomposition of the text. For example, after semantic decomposition, the above activities form the following sub-activity sequence: {(Online promotion activity, 2026-01-10 to 2026-01-20), (Offline promotion activity, 2026-01-15 to 2026-01-25), (Customer appreciation activity, 2026-01-22)}.

[0028] Step S20 also includes the following steps: S21: Using a semantic segmentation model, perform syntactic analysis on the project planning text to obtain sub-activity items and time nodes; S22: Determine the activity time boundaries for each sub-activity based on the sub-activity items and time nodes; S23: Construct a sequence based on the sub-activity items and activity time boundaries to obtain the sub-activity sequence.

[0029] For steps S21-S23, the activity time boundary is the time range information of the corresponding sub-activity item, which is used to limit the start and end time of the corresponding sub-activity item.

[0030] For example: "Online promotional activity"; time period: "January 10, 2026 to January 20, 2026"; then the corresponding activity time boundary is: start time: 2026-01-10; end time: 2026-01-20.

[0031] In one embodiment, the sub-activity sequence is structured data constructed based on multiple sub-activity items and their corresponding time boundaries. For example, based on the following activity content: "Online promotional event, from January 10, 2026 to January 20, 2026;" Offline promotional activities, from January 15, 2026 to January 25, 2026; Customer appreciation event, date: January 22, 2026. The sub-activity sequence is as follows: {(Online promotion activity, 2026-01-10 to 2026-01-20), (Offline promotion activity, 2026-01-15 to 2026-01-25), (Customer appreciation activity, 2026-01-22 00:00 to 2026-01-22 23:59)}.

[0032] By using a semantic segmentation model to perform syntactic analysis on the project planning text, the sub-activity items and their corresponding time nodes in the project planning text are obtained; by determining the time range of the time nodes corresponding to the sub-activity items, the corresponding activity time boundaries are formed; finally, by combining the sub-activity items and the activity time boundaries to construct a sub-activity sequence, the unstructured activity text is converted into structured activity data. Step S23 also includes the following steps: S231: Based on the initial planning text and the pre-set knowledge graph, perform dependency analysis on the sub-activity items to determine the prerequisite dependencies of each sub-activity item; S232: Treat each sub-activity item as a graph node, and use the prerequisite dependencies as directed edges. Construct a graph based on the graph nodes and directed edges to obtain the initial activity graph network. S233: Perform loop detection on the initial activity graph network. If there are no closed loops in the initial activity graph network, generate a sub-activity sequence.

[0033] For steps S231 to S233, the preset knowledge graph is a pre-built activity dependency knowledge base used to store the process dependencies between different activities.

[0034] For example, in an insurance marketing campaign scenario, the pre-defined knowledge graph includes the following campaign-dependent knowledge: Before the "live promotion" campaign begins, the "promotional material production" must be completed first. Before the "Customer Appreciation Event" begins, "Venue Reservation" must be completed. The reason for introducing knowledge graphs is that in the actual event planning process, some prerequisites may be assumed by planners to be common-sense procedures and therefore not explicitly written into the initial planning document. For example, the planning document may only state: "A customer appreciation event will be held on January 22nd," but it may not explicitly state: "The meeting room reservation and venue setup need to be completed in advance."

[0035] Therefore, by introducing a pre-defined knowledge graph, based on historical activity knowledge and activity dependency rules, implicit preceding activity relationships can be identified, thereby improving the completeness of activity process analysis.

[0036] In one embodiment, dependency analysis includes: using natural language processing technology to identify action semantics, execution objects, and process keywords in each sub-activity item; then, matching the identification results with activity dependency rules in a preset knowledge graph; and thereby determining the prerequisite dependencies between different sub-activity items.

[0037] For example, when the sub-activity "Customer Appreciation Activity" is identified, if the preset knowledge graph shows that "Customer Appreciation Activity" depends on "Venue Setup Completed", then it is determined that "Venue Setup" is a prerequisite activity for "Customer Appreciation Activity".

[0038] In one embodiment, graph nodes are data nodes corresponding to each sub-activity. For example, "Online Promotion Activity," "Offline Promotion Activity," and "Customer Appreciation Activity" are all graph nodes.

[0039] A directed edge is a directional connection edge that represents the sequential dependency relationship between two sub-activity items.

[0040] For example, if "Promotional Material Production" needs to be executed before "Online Live Stream Promotion," then a directed edge is created: "Promotional Material Production" → "Online Live Stream Promotion." The direction of the directed edge indicates the order in which the activities are executed.

[0041] In one embodiment, loop detection involves performing circular dependency checks on the directed edge relationships in the initial activity graph network to determine whether a closed-loop structure exists within the activity dependencies. A closed loop indicates that an activity depends on itself again after following a dependency path. For example: Activity A depends on Activity B; Activity B depends on Activity C; Activity C depends on Activity A. This forms the following closed loop: A → B → C → A. That is, Activity A requires Activity B to be completed before it can be executed; Activity B requires Activity C to be completed before it can be executed; however, Activity C requires Activity A to be completed first, thus preventing the activity flow from executing normally.

[0042] Therefore, when a loop is detected in the initial activity graph network, it means that there is no circular conflict in the dependencies between the sub-activities, thus forming a legal activity execution order.

[0043] S30: Extract keywords for each sub-activity to obtain initial resource tags for each sub-activity; In one embodiment, keyword extraction includes: performing word segmentation on each sub-activity item, performing part-of-speech tagging on the word segmentation results, extracting activity type keywords and resource requirement keywords based on the part-of-speech tagging results, and generating corresponding initial resource tags.

[0044] In one embodiment, the initial resource tags include: an activity type tag and a specific resource requirement tag. For example, for a sub-activity project: "Promoting insurance products through a live streaming platform requires 2 hosts and live streaming equipment," the extracted initial resource tags are: Activity type tag: Live Streaming Promotion; Specific resource requirement tags: Hosts; Live Streaming Equipment.

[0045] S40: Using a pre-built project resource mapping database, map each initial resource tag to determine the resource requirement data for each sub-activity project; The project resource mapping database is a database that records activity type tags, the specific resources used by the activity in historical data, and the quantity of each resource.

[0046] Resource mapping refers to looking up the activity type label in the initial resource label in a table to obtain the resources required for that type of activity and the corresponding quantity of resources in historical data, thereby determining the resource requirement data for each sub-activity project.

[0047] Please see Figure 3 As shown, Figure 3 A flowchart illustrating step S40 provided in an embodiment of the present invention includes the following steps: S41: Look up the project resource mapping database according to the activity type to obtain historical activity records; S42: Extract information from the historical activity records to obtain historical resource requirements; S43: Extract information from the initial resource requirements to obtain the first resource requirements; S44: Based on the historical resource requirements, supplement the initial resource requirements to obtain the second resource requirements; S45: Combine the first resource requirement and the second resource requirement to obtain the resource requirement data.

[0048] In one embodiment, information extraction involves retrieving resource entries and their corresponding quantities from historical activity records and converting them into a unified key-value pair format to obtain historical resource requirements. For example, "The live event uses 2 cameras and 1 host" is parsed as: {Cameras: 2, Host: 1}.

[0049] Resource replenishment is the process of supplementing the primary resource requirement based on historical resource requirements, used to identify the missing resource items and their quantity information in the current resource requirement.

[0050] Specifically, when certain resource types exist in historical resource requirements but are not included in the first resource requirement, the resource type and its corresponding quantity are added as supplementary resources to the current resource set.

[0051] For example: Historical resource requirements: {Cameras: 2 units, Anchor: 1 person, Lighting equipment: 2 sets} Primary resource requirements: {Camera: 1 unit, Streamer: 1 person} The resource replenishment result is: {Lighting equipment: 2 sets}, which is the second resource requirement.

[0052] The resource demand data is the final set of resource demands obtained by merging the first resource demand and the second resource demand.

[0053] For example: First resource requirement: {Camera: 1 unit, Anchor: 1 person}, Second resource requirement: {Lighting equipment: 2 sets}, After merging, the resource requirement data is: {Camera: 1 unit, Anchor: 1 person, Lighting equipment: 2 sets}.

[0054] By merging the first and second resource requirements, the resource requirement results include both the explicit requirements of current activities and the implicit experience information of historical activities, thereby improving the accuracy and rationality of the resource requirement results.

[0055] Please see Figure 4 As shown, Figure 4 A flowchart illustrating step S44 provided in an embodiment of the present invention includes the following steps: S441: Based on the initial resource requirements, historical resource requirements are filtered to determine alternative resource requirements; wherein, alternative resource requirements exist in historical resource requirements but not in the initial resource requirements; S442: Based on historical and initial resource demands, predict the quantity of candidate resources to obtain the quantity of candidate resources; S443: Treat the quantity and demand of reserve resources as the second resource demand.

[0056] The alternative resource requirements are a set of resources selected from historical resource requirements. The characteristic of this resource set is that the types of resources included in the resource set exist in the historical resource requirements, but are not reflected in the initial resource requirements.

[0057] Specifically, the screening process for alternative resource requirements is as follows: each resource type in the historical resource requirements is compared with the resource types in the initial resource requirements. When a certain resource type exists in the historical resource requirements but does not exist in the initial resource requirements, that resource type is identified as a alternative resource requirement.

[0058] For example: if the historical resource requirements are {camera, lighting equipment, anchor, booth}, and the initial resource requirements are {camera, anchor}, then the alternative resource requirements are {lighting equipment, booth}.

[0059] In one embodiment, the prediction of the number of candidate resources can be made by: predicting based on the average resource usage of the same or similar activity types in historical activities; predicting based on proportional mapping based on activity scale parameters (such as the number of participants and the duration of the activity); or predicting the number of resources by fitting a regression model or machine learning model.

[0060] For example, if the number of lighting equipment sets used in similar live-streaming promotional events in the past is on average 2 sets, then it can be predicted that the current event will have 2 sets of lighting equipment.

[0061] The second resource requirement is a set of resource requirements obtained by structuring and combining the candidate resource requirements and their corresponding quantities. For example, the second resource requirement is represented as: {Lighting equipment: 2 sets, Exhibition booth: 1}.

[0062] By using the above methods, resource types that are implicit but not explicitly present in historical experience are introduced into the initial resource requirements, and they are reasonably quantified through quantity prediction, thereby improving the accuracy of resource requirement expression.

[0063] Please see Figure 5 As shown, Figure 5 A flowchart illustrating step S442 provided in an embodiment of the present invention includes the following steps: Step S4421: Obtain the quantity information of the candidate resource demand in the historical resource demand to obtain the historical resource quantity; Step S4422: Divide historical resource requirements according to candidate resource requirements to obtain dependent variable resource requirements; Step S4423: Calculate the candidate resource demand based on the dependent variable resource demand to obtain the correlation coefficient; Step S4424: Based on the initial resource demand and the correlation coefficient, the demand for candidate resources is predicted to obtain the quantity of candidate resources.

[0064] In one embodiment, the process of predicting the quantity of candidate resource demand is implemented based on the structural correspondence between historical resource demand and current initial resource demand.

[0065] Specifically, in step S4421, the resource usage quantity corresponding to the candidate resource demand is extracted from the historical activity record to obtain the historical resource quantity, which is used to characterize the actual consumption level of the candidate resource in the historical activity.

[0066] For example, if historical resource requirements include resource sets A, B, C, and D, where A and D are candidate resource requirements, then the usage quantity of A and D in historical activities is extracted as the historical resource quantity.

[0067] In one embodiment, in step S4422, the historical resource demand is divided according to the candidate resource demand, which is to divide the historical resource demand into two sets of resources according to the resource type, namely dependent variable resource demand and non-dependent variable resource demand.

[0068] Among them, the dependent variable resource demand is the set of resources that already exist in the current initial resource demand, such as BC; the non-dependent variable resource demand is the set of resources that did not appear in the initial resource demand in the history, such as AD.

[0069] In one embodiment, in step S4423, for example, in historical activities: based on the quantities of resources B and C and resource A, the correlation coefficient between resources BC and resource A is calculated, and the correlation coefficient between resource D is calculated based on the quantities of resources BC, thus obtaining the correlation coefficients between A and BC, and between D and BC respectively.

[0070] In one embodiment, in step S4424, the demand for candidate resources is predicted based on the initial resource demand and the correlation coefficient. This involves deriving and calculating the resource quantity of candidate resource demands (such as A and D) based on the currently known dependent variable resource demand (such as BC) and its corresponding correlation coefficient.

[0071] By utilizing the correlation coefficients between historical resources, and with only a portion of the resources known, we can quantitatively predict the missing resources, thereby improving the completeness and rationality of resource demand forecasting and the accuracy of resource demand modeling.

[0072] S50: Optimize requirements based on time nodes and resource demand data to obtain a resource demand list; The requirement optimization aims to unify and integrate resource requirements while meeting the constraints of normal execution of each sub-activity, thereby reducing redundant resource allocation and resources.

[0073] Step S50 may also include the following steps: S51: Based on resource demand data and time nodes, conduct demand analysis to obtain overlapping demand data; S52: Optimize the resource demand data based on the overlapping demand data to obtain a resource demand list.

[0074] In one embodiment, the overlapping demand data is resource conflict relationship data obtained by jointly analyzing time nodes and resource demand data. Specifically, the overlapping demand data includes at least two scenarios: The first scenario is a conflict-type overlapping requirement where the time overlaps and the resource type is the same, that is, multiple sub-activities occupy the same type of resource at the same time interval. In this case, the resource cannot be reused. The second scenario is a reuse-type overlapping requirement where the time does not overlap but the resource type is the same, that is, the same resource is used by different sub-activities at different time nodes. In this case, the resource does not conflict in the time dimension and has the conditions for cross-sub-activities to be reused.

[0075] Therefore, resources that meet the conditions for cross-sub-activity reuse are compressed, thereby reducing the amount of resources allocated, improving resource utilization, and thus improving the accuracy of resource allocation.

[0076] S60: Allocate resources according to the resource requirements list.

[0077] Resources are allocated based on the resource demand list, for example, by issuing work orders on the system.

[0078] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0079] In one embodiment, a resource allocation device is provided, which corresponds one-to-one with the resource allocation method in the above embodiments. For example... Figure 6 As shown, the resource allocation device includes a text preprocessing module 101, a text decomposition module 102, a keyword extraction module 103, a resource mapping module 104, a demand optimization module 105, and a resource allocation module 106. Detailed descriptions of each functional module are as follows: The text preprocessing module 101 is used to obtain the initial planning text and perform text preprocessing on the initial planning text to obtain the project planning text; The text decomposition module 102 is used to decompose the project planning text using a pre-trained semantic segmentation model to obtain a sub-activity sequence; wherein, the sub-activity sequence includes at least one sub-activity project and the time node corresponding to the sub-activity project; Keyword extraction module 103 is used to extract keywords for each sub-activity item to obtain the initial resource tags for each sub-activity item; The resource mapping module 104 is used to map resources to each initial resource tag using a pre-built project resource mapping database, and to determine the resource requirement data of each sub-activity project. The requirement optimization module 105 is used to optimize requirements based on time nodes and resource requirement data to obtain a resource requirement list. Resource allocation module 106 is used to allocate resources according to the resource requirement list.

[0080] In one embodiment, the text decomposition module 102 is specifically used for: Using a semantic segmentation model, syntactic analysis is performed on the project planning text to obtain sub-activity items and time nodes; Based on the sub-activity items and time nodes, determine the activity time boundaries for each sub-activity item; The sequence of sub-activities is constructed based on the sub-activity items and activity time boundaries to obtain the sub-activity sequence.

[0081] In one embodiment, the process of constructing a sequence based on sub-activity items and activity time boundaries to obtain a sub-activity sequence further includes: Based on the initial planning text and the pre-set knowledge graph, a dependency analysis is performed on the sub-activity items to determine the prerequisite dependencies of each sub-activity item. Each sub-activity item is treated as a graph node, and the prerequisite dependencies are used as directed edges. The graph is constructed based on the graph nodes and directed edges to obtain the initial activity graph network. Loop detection is performed on the initial activity graph network. If no closed loop is found in the initial activity graph network, a sub-activity sequence is generated.

[0082] In one embodiment, the resource mapping module 104 is specifically used for: The project resource mapping database is searched according to the type of activity to obtain historical activity records; Information is extracted from historical activity records to determine historical resource needs; Information is extracted from the initial resource requirements to obtain the first resource requirement; Based on historical resource requirements, the initial resource requirements are supplemented to obtain the second resource requirements; The first and second resource requirements are combined to obtain resource requirement data.

[0083] The second resource requirement, derived by supplementing the initial resource requirement based on historical resource needs, also includes: Historical resource requirements are filtered based on initial resource requirements to determine alternative resource requirements; alternative resource requirements exist in historical resource requirements but not in initial resource requirements. Based on historical and initial resource demands, the quantity of candidate resources is predicted to be available. The quantity and demand of reserve resources are considered as the second resource demand.

[0084] This includes forecasting the quantity of candidate resources based on historical and initial resource demands to obtain the quantity of candidate resources; it also includes: Obtain the quantity information of alternative resource demand in historical resource demand to obtain the historical resource quantity; Based on the alternative resource demand, historical resource demand is divided to obtain the dependent variable resource demand; The correlation coefficient is obtained by calculating the alternative resource demand based on the dependent variable resource demand; Demand forecasting is performed on the demand for candidate resources based on the initial resource demand and correlation coefficients to obtain the quantity of candidate resources.

[0085] In one embodiment, the demand optimization module 106 is specifically used for: Demand analysis is performed based on resource demand data and time points to obtain overlapping demand data; Based on overlapping demand data, the resource demand data is optimized to obtain a resource demand list.

[0086] This invention provides a resource allocation device. First, it acquires an initial planning text and preprocesses it to obtain a project planning text. Then, it uses a semantic segmentation model to automatically decompose the activity planning text, automatically identifying different sub-activities and corresponding time nodes from unstructured activity text. This reduces comprehension biases caused by manual activity process decomposition and improves the efficiency of activity content parsing. Simultaneously, through keyword extraction and the use of a resource mapping database, it automatically determines the resource requirements for each sub-activity, reducing omissions or errors during manual resource requirement statistics. Furthermore, by optimizing resource requirements based on time nodes, it can comprehensively analyze resource requirements for different time periods, thereby improving the accuracy of the resource requirement list. Finally, it allocates resources based on the optimized resource requirement list, improving the accuracy of resource allocation in scenarios with multiple concurrent activities.

[0087] Specific limitations regarding the resource allocation device can be found in the limitations of the resource allocation method above, and will not be repeated here. Each module in the aforementioned resource allocation device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0088] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 7As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the computer program is executed by the processor, it implements a resource allocation method, a server-side function, or steps.

[0089] In one embodiment, a computer device is provided, which may be a client, and its internal structure diagram may be as follows: Figure 8 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements a resource allocation method, client-side functions, or steps. In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: Obtain the initial planning text and perform text preprocessing on the initial planning text to obtain the project planning text; The project planning text is decomposed using a pre-trained semantic segmentation model to obtain a sequence of sub-activities; wherein, the sequence of sub-activities includes at least one sub-activity project and the time node corresponding to the sub-activity project. Keyword extraction is performed on each sub-activity to obtain the initial resource tags for each sub-activity. Using a pre-built project resource mapping database, resource mapping is performed on each initial resource tag to determine the resource requirement data for each sub-activity project; Based on time points and resource demand data, demand optimization is performed to obtain a resource demand list; Resources are allocated based on the resource requirements list.

[0090] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: Obtain the initial planning text and perform text preprocessing on the initial planning text to obtain the project planning text; The project planning text is decomposed using a pre-trained semantic segmentation model to obtain a sequence of sub-activities; wherein, the sequence of sub-activities includes at least one sub-activity project and the time node corresponding to the sub-activity project. Keyword extraction is performed on each sub-activity to obtain the initial resource tags for each sub-activity. Using a pre-built project resource mapping database, resource mapping is performed on each initial resource tag to determine the resource requirement data for each sub-activity project; Based on time points and resource demand data, demand optimization is performed to obtain a resource demand list; Resources are allocated based on the resource requirements list.

[0091] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0092] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0093] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0094] It should be noted that any AI models, software tools, or components not belonging to this company appearing in the embodiments of this application are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this application has been authorized (with the knowledge and consent) by the relevant parties or has been fully authorized by all parties, and the executing entity may obtain it through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals.

[0095] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A resource allocation method, characterized in that, The method includes: Obtain the initial planning text and perform text preprocessing on the initial planning text to obtain the project planning text; The project planning text is decomposed using a pre-trained semantic segmentation model to obtain a sequence of sub-activities; wherein, the sequence of sub-activities includes at least one sub-activity project and the time node corresponding to the sub-activity project; Keyword extraction is performed on each of the sub-activity items to obtain the initial resource tags for each sub-activity item; Using a pre-built project resource mapping database, resource mapping is performed on each of the initial resource tags to determine the resource requirement data for each of the sub-activity projects; Based on the time points and the resource demand data, demand optimization is performed to obtain a resource demand list; Resources are allocated based on the resource requirements list.

2. The method according to claim 1, characterized in that, The process of using a pre-trained semantic segmentation model to decompose the project planning text into a sequence of sub-activities includes: Using the semantic segmentation model, the project planning text is syntactically analyzed to obtain sub-activity projects and time nodes; Based on the sub-activity items and the time nodes, determine the activity time boundaries for each sub-activity item; The sub-activity sequence is obtained by constructing a sequence based on the sub-activity items and the activity time boundaries.

3. The method according to claim 2, characterized in that, The step of constructing the sequence based on the sub-activity items and the activity time boundaries to obtain the sub-activity sequence includes: Based on the initial planning text and the preset knowledge graph, a dependency analysis is performed on the sub-activity items to determine the prerequisite dependencies of each sub-activity item; Each of the sub-activity items is used as a graph node, and the prerequisite dependencies are used as directed edges. A graph is constructed based on the graph nodes and the directed edges to obtain an initial activity graph network. Loop detection is performed on the initial activity graph network. If the initial activity graph network does not have any closed loops, the sub-activity sequence is generated.

4. The method according to claim 1, characterized in that, The initial resource tags include activity types and initial resource requirements. The step of using a pre-built project resource mapping database to map resources to each of the initial resource tags and determine the resource requirement data for each of the sub-activity projects includes: Based on the activity type, the project resource mapping database is looked up to obtain historical activity records; Information is extracted from the historical activity records to obtain historical resource requirements; Information is extracted from the initial resource requirements to obtain the first resource requirements; Based on the historical resource requirements, the initial resource requirements are supplemented to obtain the second resource requirements; The first resource requirement and the second resource requirement are combined to obtain the resource requirement data.

5. The method according to claim 4, characterized in that, The step of supplementing the initial resource requirement based on the historical resource requirement to obtain the second resource requirement includes: Based on the initial resource requirements, the historical resource requirements are filtered to determine the candidate resource requirements; wherein, the candidate resource requirements exist in the historical resource requirements but not in the initial resource requirements; Based on the historical resource demand and the initial resource demand, the quantity of the candidate resource demand is predicted to obtain the quantity of candidate resources. The quantity of candidate resources and the demand for candidate resources are taken as the second resource demand.

6. The method according to claim 5, characterized in that, The step of predicting the quantity of candidate resources based on the historical resource demand and the initial resource demand to obtain the quantity of candidate resources includes: Obtain the quantity information of the candidate resource demand in the historical resource demand to obtain the historical resource quantity; Based on the candidate resource requirements, the historical resource requirements are divided to obtain the dependent variable resource requirements; The correlation coefficient is obtained by calculating the candidate resource demand based on the dependent variable resource demand; Based on the initial resource demand and the correlation coefficient, the demand for the candidate resources is predicted to obtain the quantity of the candidate resources.

7. The method according to claim 1, characterized in that, The step of optimizing the resource requirements based on the time points and the resource requirement data to obtain a resource requirement list includes: Based on the resource demand data and the time points, demand analysis is performed to obtain overlapping demand data; The resource requirement data is optimized based on the overlapping requirement data to obtain the resource requirement list.

8. A resource allocation device, characterized in that, include: The text preprocessing module is used to obtain the initial planning text and perform text preprocessing on the initial planning text to obtain the project planning text; The text decomposition module is used to decompose the project planning text using a pre-trained semantic segmentation model to obtain a sub-activity sequence; wherein, the sub-activity sequence includes at least one sub-activity project and the time node corresponding to the sub-activity project; The keyword extraction module is used to extract keywords from each of the sub-activity items to obtain the initial resource tags for each of the sub-activity items. The resource mapping module is used to map resources to each of the initial resource tags using a pre-built project resource mapping database, and to determine the resource requirement data of each of the sub-activity projects. The demand optimization module is used to optimize the demand based on the time nodes and the resource demand data to obtain a resource demand list. The resource allocation module is used to allocate resources according to the resource requirement list.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the resource allocation method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the resource allocation method as described in any one of claims 1 to 7.

Images