AI-based cultural transmission economic benefit optimization system

Abstract

The invention relates to the technical field of cultural transmission, in particular to an AI-based cultural transmission economic benefit optimization system, which comprises a cultural value extraction module, a transmission efficiency analysis module, a resource adaptation optimization module, a benefit evaluation generation module and a credibility assignment module. According to the method, the multi-dimensional data change trend in the cultural transmission activity is obtained, dynamic matching and mapping projection are performed in combination with the spatial distribution characteristics of the audience behavior thermodynamic diagram, the core value point and the transmission coverage area are accurately identified, the key transmission node and the resource allocation weight are positioned, and the economic benefit is clearly optimized; and credibility evaluation is introduced to distinguish the difference degree of optimization effects. According to the method, the transmission effect capturing precision can be remarkably improved, resource waste is reduced, and the operation stability and economy of cultural transmission are enhanced.

Description

Technical Field

[0001] The present invention relates to the field of cultural communication technology, and in particular to an AI-based cultural communication economic benefit optimization system. Background Art

[0002] With the rapid development of artificial intelligence technology, AI-based optimization systems have been widely used in many fields to improve economic benefits and resource utilization efficiency. However, in the field of cultural communication, existing optimization technologies still have significant limitations, especially in terms of how to effectively combine economic benefits with cultural communication goals, and a mature technical solution has not yet been formed. In the existing technology, for example, a system for realizing the multi-objective economic benefit optimization of building product combinations with publication number CN110288156B uses the Octopus plug-in to independently control the quantity of each building product, thereby achieving multi-objective optimization of maximizing total profit and approaching the total building area and base area to the target. However, this technical solution mainly focuses on the economic benefit optimization of building product combinations. Its optimization goals are concentrated on economic indicators, and it lacks consideration of the effect of cultural value communication. Therefore, it cannot be directly applied to cultural communication scenarios. In addition, the optimization model of this solution is relatively complex, and it may face difficulties in data acquisition and model adaptation when applied to the field of cultural communication.

[0003] Another existing technology, with publication number CN114172175B, is a method for collaborative optimization of hydrogen storage configuration and control to improve the economic benefits of wind farms. This method establishes a steady-state mathematical model of the wind / hydrogen / storage coupling system and uses a convex optimization method to simultaneously optimize the system configuration and control, thereby improving the economic benefits of wind farms. Although this technical solution has demonstrated strong optimization capabilities in the field of new energy, its optimization goals mainly revolve around the economic benefits and technical performance of the energy system, and do not involve the evaluation of social benefits and cultural influence related to cultural communication. Therefore, this solution is difficult to meet the needs of dual optimization of economic benefits and cultural communication effects in cultural communication. In addition, although the solution's treatment of nonlinear efficiency improves the optimization accuracy, there is no precedent for similar cultural communication efficiency modeling in the field of cultural communication, and it may be difficult to directly migrate and apply it.

[0004] The above issues demonstrate that existing optimization technologies still face significant shortcomings in their application to the field of cultural communication, particularly in the lack of targeted technical solutions for synergistically optimizing economic benefits and cultural communication effectiveness. Cultural communication requires not only a focus on economic benefits but also on its social benefits and influence, placing higher demands on optimization technologies. Therefore, there is an urgent need for an optimization system that can combine the social and economic benefits of cultural communication to construct an optimization model suitable for cultural communication scenarios, thereby improving the overall effectiveness of cultural communication and meeting the demand for efficient and intelligent optimization systems in the modern cultural communication field. Summary of the Invention

[0005] The purpose of this invention is to solve the shortcomings of the existing technology and propose an AI-based cultural communication economic benefit optimization system.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an AI-based cultural communication economic benefit optimization system, the system comprising:

[0007] The cultural value extraction module obtains the cultural theme characteristics of each activity in the historical communication activity data, compares the theme popularity change value with the audience feedback vector, determines whether the change direction of the theme popularity has consistent characteristics, and generates the core value trend value of cultural communication;

[0008] The communication efficiency analysis module calls the cultural communication core value trend value, combines it with the behavior heat map provided by the audience behavior monitoring device, and generates communication efficiency distribution information based on the cross-projection contour of the heat map coverage area and the communication path;

[0009] The resource adaptation optimization module obtains the cultural communication resource allocation structure data based on the communication efficiency distribution information, performs spatial overlap judgment on the communication path and the resource allocation image, identifies the communication segments with uneven resource allocation, and generates a resource adaptation optimization block coordinate set;

[0010] The benefit evaluation generation module calls the resource adaptation optimization block coordinate set, obtains the number index of the coverage node at the propagation level, marks the output change trend difference item, inputs the difference item as an external optimization feature into the optimization input sequence, and generates optimization impact economic benefit evaluation information.

[0011] As a further solution of the present invention, the core value trend value of cultural communication includes the amplitude of fluctuation of topic popularity, density of change of value direction, consistency level of audience feedback and identification mark of topic mutation; the communication efficiency distribution information includes the main axis angle distribution of overlapping area, communication coverage breadth index, audience behavior classification level and projection area number mapping table; the resource adaptation optimization block coordinate set is specifically the node projection intersection boundary point, the correspondence between the communication area and the resource arrangement, the path occlusion block index and the spatial overlap density factor; the optimization impact economic benefit evaluation information includes the external resource interference feature sequence, the node-level response trend symbol sequence, the neural network prediction value set and the evaluation segment index structure.

[0012] As a further solution of the present invention, the cultural value extraction module includes:

[0013] The topic feature extraction submodule obtains the cultural theme features of each activity in the historical communication activity data, extracts the behavioral boundary points in each activity whose attention is greater than the set topic popularity threshold, establishes a topic point set, and constructs a topic boundary line structure sequence based on the relative positions of the topic points;

[0014] The heat change calculation submodule calls the theme boundary line structure sequence, extracts the coordinate values ​​of corresponding boundary point pairs in adjacent activity theme line sequences, calculates relative displacement vectors, constructs an angle sequence set based on the angles between points, performs ratio operations on the boundary angle sequences of continuous activities, calculates the mean value of the boundary change gradient, screens the sections where the angle change exceeds the change gradient threshold, and generates a theme direction change gradient value set;

[0015] The feedback consistency judgment submodule calls the theme direction change gradient value set, retrieves the average feedback vector recorded by the audience behavior monitoring device, performs projection operation on the direction change gradient, compares the angle between it and the feedback direction, and screens the segments based on the set consistency threshold to obtain the proportion of the number of segments whose theme change direction is consistent with the feedback direction, and obtains the core value trend value of cultural communication.

[0016] As a further solution of the present invention, the transmission efficiency analysis module includes:

[0017] The behavior direction extraction submodule calls the core value trend value of cultural communication, obtains the behavior heat map recorded by the audience behavior monitoring device, identifies the boundary of the medium behavior area in the heat map, determines the behavior distribution contour based on the set of isovalue contour lines, extracts the direction distribution of the boundary line and performs vectorized fitting in the position order to generate a behavior main axis direction sequence;

[0018] The angle matching calculation submodule obtains the angular relationship between the vector segment and the theme direction segment marked in the core value trend value of cultural communication based on the main axis direction sequence of the behavior, sequentially compares the angles between the two sets of direction vectors in the spatial coordinate plane, selects vector pairs whose angle deviation values ​​are lower than the set threshold of the propagation angle, calculates and obtains the regional direction deviation index value, and uses the blocks with deviation values ​​lower than the propagation angle threshold as matching areas to generate a direction matching propagation distribution value set;

[0019] The spatial cross-construction submodule matches the propagation distribution value set according to the direction, retrieves the propagation path information within the time period, establishes a propagation vector projection layer in a unified spatial coordinate system, locates the projection boundary intersection block between the propagation path and the marked behavior area, extracts the boundary line index value of the overlapping block and generates a spatial geometric coverage group to establish the propagation efficiency distribution information.

[0020] As a further solution of the present invention, the resource adaptation optimization module includes:

[0021] The resource arrangement extraction submodule obtains the communication efficiency distribution information, collects the cultural communication resource allocation structure data, extracts the two-dimensional spatial position index corresponding to the resource number grid, spatially locates the arrangement rows and numbering sequence of the resources in the actual site, and generates a resource spatial position set based on the orientation angle value of the resource receiving surface;

[0022] The propagation path mapping submodule calls the resource spatial location set, collects the propagation altitude angle, propagation intensity measurement value, and resource surface orientation azimuth, calculates the propagation path direction corresponding to the resource receiving surface based on the propagation direction vector, projects the path direction to the plane coordinate, determines whether there is overlap with the coverage range of the propagation efficiency distribution information, obtains the spatial path segment where the propagation and resource areas overlap, and generates path resource overlap distribution data;

[0023] The intersection block identification submodule extracts the coordinate point set corresponding to the resource segment based on the path resource overlap distribution data, identifies the numbered area to which the coordinate point belongs in the resource arrangement space, cross-judges the area boundary with the resource path boundary, extracts the resource number index corresponding to the intersection point, determines the resource number of the discontinuous propagation area and the corresponding coordinate area, and establishes a resource adaptation optimization block coordinate set.

[0024] As a further solution of the present invention, the benefit evaluation generation module includes:

[0025] The node signal acquisition submodule calls the resource adaptation optimization block coordinate set, extracts the corresponding resource number index at the propagation level, collects the unit resource output value and propagation count value recorded by resource monitoring, constructs the propagation status data group of each resource at the current moment, and generates a resource operation signal set;

[0026] The trend difference marking submodule obtains the benefit output values ​​of the resources in the propagation area and the resources in the adjacent non-propagation area at the same time point based on the resource operation signal set, compares the difference in change direction, identifies signal items with inconsistent directions, records the difference items in the form of symbolic values, classifies and organizes them into a separate input factor sequence, and obtains a benefit trend difference vector group;

[0027] The optimization information generation submodule calls the benefit trend difference vector group, inputs the neural network input sequence structure completed by historical period sample training, inputs the vector group and the constructed input factor items in parallel, performs combined optimization deduction based on the input factors of communication intensity, resource density and audience behavior data, obtains the optimization value distribution results at the corresponding time point, and generates optimization impact economic benefit evaluation information.

[0028] As a further embodiment of the present invention, the system further comprises:

[0029] The credibility assignment module calls the optimization impact economic benefit evaluation information, obtains the optimization value and the actual output value of the resource at each optimization time point, compares the degree of deviation between the optimization value and the actual output value, and compares it with the set confidence deviation threshold, divides the deviation segment above the threshold into intervals according to the level division benchmark, marks each level division segment as a credibility level identifier, and generates credibility level information for the optimization coverage area;

[0030] The trust level information of the optimized coverage area specifically refers to the deviation interval level index, the credibility label mapping result, the deviation fluctuation trend group and the credibility judgment result number table.

[0031] As a further solution of the present invention, the credibility assignment module includes:

[0032] The deviation value calculation submodule calls the optimization impact economic benefit evaluation information, obtains the optimized output sequence and the actual output sequence of each resource in the optimization time period, constructs a deviation difference set between the corresponding sequence indexes, and normalizes the deviation ratio based on the optimization output to obtain a normalized benefit deviation index.

[0033] The deviation segmentation submodule calls and sets the confidence deviation threshold according to the normalized benefit deviation index, groups and classifies the index sequence according to the upper limit of the threshold, establishes index labels for the time periods corresponding to the groups, and numbers the deviation levels corresponding to the groups in sequence to obtain the deviation level mapping interval value;

[0034] The trusted level labeling submodule calls the deviation level mapping interval value, establishes the trusted level interval corresponding benchmark based on the time mapping table of the level interval and the resource number, outputs the trusted label result set corresponding to the resource in each time period, and outputs the trusted level information of the optimized coverage area.

[0035] Compared with the prior art, the advantages and positive effects of the present invention are:

[0036] In the present invention, by obtaining the trend of multi-dimensional data changes in cultural communication activities and calculating the consistency characteristics of topic popularity changes and audience feedback, it is possible to accurately identify the core value points of cultural communication, determine the dynamic evolution of the communication coverage area, and significantly improve the accuracy of capturing communication effects. Combined with the spatial distribution characteristics of the audience behavior heat map, dynamic matching and mapping projection are performed, and the adaptation area of ​​communication resources to communication nodes is accurately delineated in space, the actual area scope and influence level of communication coverage are clarified, and the refined identification of the disturbance area of ​​the communication path is realized. Based on the spatial mapping and cross-calculation of the communication path and resource layout, the discontinuous resource segments of the communication are specifically locked, and the precise determination is made. By positioning key communication nodes and resource allocation weights, the economic benefits of specific communication scenarios are clearly optimized, making the optimization results highly targeted. By establishing a neural network evaluation model for optimization impact, the communication disturbance characteristics are directly incorporated into the optimization input, effectively capturing subtle changes in short-term benefit fluctuations, greatly improving the accuracy of optimization response, and introducing credibility evaluation based on the economic benefit optimization results. Through deviation level annotation, the degree of difference in optimization effects is effectively distinguished, which improves the reliability of the actual application of the optimization plan, enables cultural communication institutions to accurately adjust fluctuating benefits, reduce resource waste, reduce redundant configuration, and enhance the operational stability and economy of cultural communication. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is a system flow chart of the present invention;

[0038] Figure 2 This is a flowchart of the acquisition of the cultural value extraction module of the present invention;

[0039] Figure 3 This is a flowchart of obtaining the transmission efficiency analysis module of the present invention;

[0040] Figure 4 This is a flowchart for obtaining the resource adaptation optimization module of the present invention;

[0041] Figure 5 The acquisition flow chart of the benefit evaluation generation module of the present invention;

[0042] Figure 6 This is a flow chart of obtaining the credibility assignment module of the present invention. DETAILED DESCRIPTION

[0043] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0044] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as an "exemplary" in the present invention should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of the word "exemplary" is intended to present concepts in a concrete manner. Furthermore, in the embodiments of the present invention, "and / or" can mean both or either of the two.

[0045] In the embodiments of the present invention, the terms "image" and "picture" may be used interchangeably. It should be noted that, when the distinction between them is not emphasized, their intended meanings are the same. The terms "of," "corresponding," and "corresponding" may be used interchangeably. It should be noted that, when the distinction between them is not emphasized, their intended meanings are the same.

[0046] In the embodiments of the present invention, sometimes a subscript such as W1 may be written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are the same.

[0047] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0048] See also Figure 1 The present invention provides a technical solution: an AI-based cultural communication economic benefit optimization system, the system comprising:

[0049] The cultural value extraction module obtains the cultural theme characteristics of each activity in the historical communication activity data, compares the theme popularity change value with the audience feedback vector, determines whether the change direction of the theme popularity has consistent characteristics, and generates the core value trend value of cultural communication;

[0050] The communication efficiency analysis module uses the core value trend value of cultural communication, combines it with the behavior heat map provided by the audience behavior monitoring equipment, and generates communication efficiency distribution information based on the cross-projection outline of the heat map coverage area and the communication path;

[0051] The resource adaptation optimization module obtains the cultural communication resource allocation structure data based on the communication efficiency distribution information, makes a spatial overlap judgment between the communication path and the resource allocation image, identifies the communication segments with unbalanced resource allocation, and generates a resource adaptation optimization block coordinate set;

[0052] The benefit evaluation generation module calls the resource adaptation optimization block coordinate set, obtains the number index of the coverage node at the propagation level, marks the output change trend difference item, inputs the difference item as the external optimization feature into the optimization input sequence, and generates the optimization impact economic benefit evaluation information;

[0053] The credibility assignment module calls the optimization impact economic benefit evaluation information, obtains the optimization value and the actual output value of the resource at each optimization time point, compares the degree of deviation between the optimization value and the actual output value, and compares it with the set confidence deviation threshold. The deviation segment above the threshold is divided into intervals according to the level division benchmark, and each level division segment is marked as a credibility level identifier to generate credibility level information of the optimization coverage area.

[0054] The core value trend values ​​of cultural communication include the fluctuation amplitude of topic popularity, the density of value direction changes, the consistency level of audience feedback and the identification mark of topic mutation. The distribution information of communication efficiency includes the main axis angle distribution of overlapping areas, the communication coverage breadth index, the audience behavior classification level and the projection area number mapping table. The resource adaptation optimization block coordinate set is specifically the node projection intersection boundary point, the correspondence between the communication area and the resource arrangement, the path block index and the spatial overlap density factor. The optimization of the economic benefit evaluation information includes the external resource interference feature sequence, the node-level response trend symbol sequence, the neural network prediction value set and the evaluation segment index structure. The credibility level information of the optimized coverage area specifically refers to the deviation interval level index, the credibility label mapping result, the deviation fluctuation trend group and the credibility judgment result number table.

[0055] See also Figure 2 , the cultural value extraction module includes:

[0056] The topic feature extraction submodule obtains the cultural theme features of each activity in the historical communication activity data, extracts the behavioral boundary points in each activity whose attention is greater than the set topic popularity threshold, establishes a topic point set, constructs a topic boundary line sequence based on the relative positions of the topic points, and generates a topic boundary line structure sequence;

[0057] Retrieve data records of specific activities from the historical communication activity database, such as the activity theme numbered HD001. The activity record includes the activity time, activity location, activity introduction, and audience message text data. Perform word segmentation on the audience message text, extract high-frequency words, and match them with the preset cultural theme vocabulary entries to identify the core cultural theme of this event. At the same time, the system analyzes the average length of stay of the audience in each exhibition area. For example, the average stay time of "keyword A" in exhibition area A is 3 minutes, the average stay time of "keyword B" in exhibition area B is 8 minutes, the average stay time of "keyword C" in exhibition area C is 10 minutes, the average stay time of "keyword D" in exhibition area D is 7 minutes, and the average stay time of "keyword E" in exhibition area E is 4 minutes. Set the theme heat threshold to 5 minutes. This threshold is based on the statistical analysis of similar activity data in the past year and is set at the 70th percentile of the stay time. The specific setting process is: collect 50 cultural relics in the past year. The average residence time data of each exhibition area of ​​the exhibition is collected, with a total of 300 exhibition area data points. These data points are sorted from low to high. The value of the 300th × 0.7 = 210th data point is 5.1 minutes, which is rounded to 5 minutes. The behavioral boundary points of exhibition areas B, C, and D with an average residence time greater than 5 minutes are their geometric center coordinates on the exhibition plan, such as exhibition area B coordinates (10, 25), exhibition area C coordinates (15, 30), and exhibition area D coordinates (20 ,28), establish the theme point set P = {(10,25), (15,30), (20,28)}, and according to the usual flow order of visitors in the theme points in the actual exhibition, that is, B to C, C to D, construct the theme boundary line sequence, that is, the line segment L1 connecting (10,25) and (15,30), and the line segment L2 connecting (15,30) and (20,28), and generate the theme boundary line structure sequence S1 = (L1,L2).

[0058] The heat change calculation submodule calls the theme boundary line structure sequence, extracts the coordinate values ​​of the corresponding boundary point pairs in the adjacent activity theme line sequence, calculates the relative displacement vector, constructs an angle sequence set based on the angle between the points, performs ratio operations on the boundary angle sequences of continuous activities, obtains the mean value of the boundary change gradient, filters the segments where the angle change exceeds the change gradient threshold, and generates a theme direction change gradient value set;

[0059] Call the theme boundary line structure sequence S1 and obtain the adjacent next activity, such as HD002, whose theme boundary line structure sequence S2 is generated in a similar way to HD001. The corresponding theme point set is P′={(12,26),(17,31),(22,29)}, and the formed line sequence is S2=(L1',L2'), where L1' connects (12,26) and (17,31), and L2' connects (17,31) and (22,29). Then, extract the coordinate values ​​of the corresponding boundary point pairs in the adjacent activity theme line sequence, for example, point (10,25) in S1 corresponds to point (12,26) in S2, point (15,30) corresponds to point (17,31), and point (20,28) corresponds to point (22,29). Calculate these correspondences. The relative displacement vectors between point pairs, for example, the displacement vector V1 of the first point pair is (12-10, 26-25) = (2, 1), the displacement vector V2 of the second point pair is (17-15, 31-30) = (2, 1), and the displacement vector V3 of the third point pair is (22-20, 29-28) = (2, 1). Based on the angles between the displacement vectors of each boundary point under the same theme in continuous activities, an angle sequence set is constructed. For example, in the change of the theme from HD001 to HD002, since the directions of the displacement vectors V1, V2, and V3 are the same, the angle between them is 0 degrees. If HD003 exists, its corresponding point displacement vectors are V1″ = (1, 2), V2″ = (1, 2), and V3″ = (1, 2). Then, in the change from HD002 to HD003, the angle between V1 and V1″ is degrees, forming an angle sequence A = {0, 36.87, ...}. A ratio operation is performed on the boundary angle sequence of continuous activities, such as A. Here, when the angle is 0, in order to avoid division by zero error, it is regarded as a minimum value, such as 0.01 degrees, or the change amount is directly calculated. Here, the angle change amount is directly calculated. For example, the first change is 0 degrees, and the second change is 36.87 degrees. The calculation obtains the mean gradient of the boundary change. For example, if there are three activity changes, the angles are 0 degrees, 36.87 degrees, and 10 degrees respectively, then the gradient mean is (0 + 36.87 + 10) / 3 ≈ 15.62 degrees. The segments where the angle change exceeds the change gradient threshold are screened. The change gradient threshold is set to 20 degrees. The setting of this threshold refers to cases in historical data where the subject direction has significantly deviated. The angle change values ​​of these cases are counted, and the 80th percentile is taken as the threshold. The experimental verification process is as follows: 100 sets of subject direction change angle data for two consecutive activities are selected. It is found that 80% of the "significant deflection" cases (manually marked by experts) have an angle change greater than 18 degrees. To increase sensitivity, it is set to 20 degrees. If the angle change between a certain activity is 36.87 degrees, the segment is screened out to generate a subject direction change gradient value set, such as {36.87}, which means that the direction of the subject has changed significantly between the activities.

[0060] The feedback consistency judgment submodule calls the theme direction change gradient value set, retrieves the average feedback vector recorded by the audience behavior monitoring device, performs a projection operation on the direction change gradient, compares the angle between it and the feedback direction, and screens segments based on the set consistency threshold to obtain the proportion of segments where the theme change direction is consistent with the feedback direction, thereby obtaining the core value trend value of cultural communication;

[0061] The gradient value set of the topic direction change is called, for example, including the value 36.87 degrees, which represents the degree of direction change of the topic during the period from activity HD002 to HD003. At the same time, the average feedback vector about the topic in the corresponding time period recorded by the audience behavior monitoring device is called. The vector is obtained by analyzing the audience's sentiment on activity-related posts on social media. For example, 1000 comments are analyzed, with 700 positive comments, 200 neutral comments, and 100 negative comments. The sentiment value is quantified as positive +1, neutral 0, and negative -1, then the average feedback value is (700× 1+200×0+100×(-1)) / 1000=0.6. The direction can be assumed to be the preset positive feedback direction. For example, it is represented by the unit vector (1,0). The gradient value of the subject direction change is 36.87 degrees. Its vector can be represented as, for example, the vector obtained by deflecting 36.87 degrees from the direction (1,0) (cos(36.87°), sin(36.87°))≈(0.8,0.6). The direction change gradient vector (0.8,0.6) is projected on the average feedback vector (1,0). The projection value is Compare the angle between the gradient vector (0.8, 0.6) and the feedback direction vector (1, 0). Degrees, segment screening is performed based on the set consistency threshold, and the consistency threshold is set to 45 degrees. This threshold is determined by analyzing cases in historical data where audience feedback significantly improved after theme adjustment. 85% of the cases are selected, and the angle between the direction change and the feedback direction is found to be less than 45 degrees. Therefore, this value is set. If the angle is 36.87 degrees and is less than 45 degrees, the segment is considered consistent. The proportion of segments with consistent theme change directions and feedback directions is obtained. Assuming that among the 10 different themes or different time periods analyzed for direction changes, 8 segments have angles less than 45 degrees, then the proportion of consistent segments is 8 / 10=0.8, and the core value trend value of cultural communication is 0.8.

[0062] See also Figure 3 ,The propagation efficiency analysis module includes:

[0063] The behavior direction extraction submodule uses the core value trend value of cultural communication to obtain the behavior heat map recorded by the audience behavior monitoring device, identifies the boundaries of the medium behavior area in the heat map, determines the behavior distribution profile based on the set of isovalue contour lines, extracts the boundary line direction distribution, and performs vectorized fitting according to the position sequence to generate the behavior main axis direction sequence;

[0064] The calculated core value trend value of cultural communication is called, for example, 0.8. This value is used as a reference here and is not directly involved in the calculation. It mainly obtains the behavior heat map recorded by the audience behavior monitoring equipment, such as thermal imagers or WiFi probes deployed in cultural memorials. For example, a tourist density heat map of a specific area of ​​10 meters x 10 meters uses different colors to represent density, indicating high-density areas and blue indicating low-density areas. The boundaries of the medium behavior areas in the heat map are identified. For example, the contour lines with density values ​​of 5, 10, and 15 people per square meter are extracted. The behavior distribution contour is determined based on these contour line sets. For example, the contour line of 10 people / square meter forms an approximately elliptical closed contour C1, and several samples on this contour C1 are extracted. For example, a point is sampled every 0.5 meters, and the tangent direction vector of the contour line at each sampling point is calculated. For example, the tangent vector at point P1 is (0.707, 0.707), and the tangent vector at point P2 is (0.6, 0.8). These tangent vectors are vectorized and fitted in order of position. For example, for the tangent vector set of all sampling points on contour C1, by calculating its principal component direction, a principal axis direction vector, such as (0.65, 0.76), is obtained, which represents the main movement or gathering direction of the people in the area. If there are multiple significant gathering areas or paths, multiple such principal axis direction vectors will be generated to form a behavioral principal axis direction sequence, for example, sequence B = [(0.65, 0.76), (0.8, 0.6)].

[0065] The angle matching calculation submodule obtains the angular relationship between the vector segment and the theme direction segment marked in the core value trend value of cultural communication based on the main axis direction sequence of the behavior. It then compares the angles between the two sets of direction vectors in the spatial coordinate plane, selects vector pairs whose angle deviation values ​​are lower than the set threshold of the propagation angle, calculates the regional direction deviation index value, and uses the blocks with deviation values ​​lower than the propagation angle threshold as matching areas to generate a direction matching propagation distribution value set.

[0066] According to the main axis direction sequence B of the behavior, for example, including vector B1 = (0.65, 0.76), obtain the angle relationship between the main axis direction vector of the behavior and the theme direction segment marked in the core value trend value of cultural communication, such as the direction change gradient vector T1 = (0.8, 0.6) of the aforementioned theme from HD002 to HD003. Compare the angles between the two sets of direction vectors in the spatial coordinate plane in turn and calculate the angle between B1 and T1. Degrees are used to filter vector pairs whose angle deviation is lower than the propagation angle threshold. The propagation angle threshold is 15 degrees. This threshold is verified based on the following experiment: information display boards are set up in a simulation environment, with angles of 0, 5, 10, 15, 20, and 25 degrees offset from the main crowd direction. The effective information reach rate is observed. It is found that when the deviation is within 15 degrees, the reach rate remains above 80%, but exceeds 15 degrees and drops sharply to below 60%. Therefore, 15 degrees is selected. Since 11.2 degrees is less than 15 degrees, the vector pair (B1, T1) is filtered out. The direction deviation index value of this area is calculated to obtain 11.2 degrees. The block with a deviation value lower than the propagation angle threshold, that is, the behavior area corresponding to B1, is regarded as the matching area. Its matching status is recorded as "matched" and the direction deviation value is 11.2 degrees. A direction matching propagation distribution value set is generated, for example, {area 1: (match, 11.2 degrees), area 2: (unmatch, 25.3 degrees), ...}.

[0067] The spatial cross-construction submodule matches the propagation distribution value set according to the direction, retrieves the propagation path information within the time period, establishes a propagation vector projection layer in a unified spatial coordinate system, locates the projection boundary intersection block between the propagation path and the marked behavior area, extracts the boundary line index value of the overlapping block and generates a spatial geometric coverage group to establish the propagation efficiency distribution information;

[0068] According to the direction matching propagation distribution value set, for example, {region 1: (matching, 11.2 degrees)}, where the geographical range of region 1 is known, for example, by the coordinate point set {(x1, y1), (x2, y2), ..., (x n ,y n )}, retrieve the cultural communication path information within a time period, for example, on June 1, 2024, such as a preset AR guide path, which consists of a series of geographic coordinate points L = [(lx1,ly1), (lx2,ly2),…, (lx m ,ly m )] are connected sequentially, and a propagation vector projection layer is established in a unified spatial coordinate system, such as the WGS84 coordinate system. That is, the path L is drawn on a two-dimensional map, and the projection boundary intersection block between the propagation path L and the behavior area 1 marked as "matched" is located. Through geometric operations, it is determined whether the path segment intersects with the area 1 polygon or is contained in it. For example, it is determined whether a line segment on the path L intersects with the area 1 polygon or is contained in it. Whether it intersects with the boundary of region 1, or whether the line segment is completely inside region 1. If there is an intersection or inclusion, the partial path and the region form an intersection block, and the boundary line index values ​​of these overlapping blocks are extracted. For example, the boundary of region 1 is composed of vertices v1, v2, ..., v nThe intersection of path L and area 1 is cp1, cp2, and the boundary of the overlapping part may be composed of the path segment (cp1, cp2) and the part from cp2 to cp1 on the boundary of area 1. The geometric description of these overlapping blocks, such as their vertex coordinate sequence, is stored, and a spatial geometric coverage group is generated, which contains all the overlapping geometric objects of the identified paths and matching behavior areas, and establishes the communication efficiency distribution information. This information is essentially a map that marks the intersection of the communication path and the high-potential audience behavior area, where the intersection area is given a higher communication efficiency expectation.

[0069] See also Figure 4 , the resource adaptation and optimization module includes:

[0070] The resource arrangement extraction submodule obtains information on the distribution of communication efficiency, collects data on the allocation structure of cultural communication resources, extracts the two-dimensional spatial position index corresponding to the resource number grid, spatially locates the arrangement rows and numbering sequence of resources in the actual site, and generates a resource spatial position set based on the orientation angle value of the resource receiving surface.

[0071] Obtain the generated communication efficiency distribution information, which indicates which geographical areas have a high degree of match between communication activities and audience behavior. At the same time, collect cultural communication resource allocation structure data, such as obtaining data from a resource management system, as shown in Table 1:

[0072] Table 1: Cultural Communication Resource Information

[0073]

[0074] Table 1 lists the number, type, spatial location, and orientation information of some communication resources. The two-dimensional spatial position index corresponding to the resource number grid is extracted, that is, (position X coordinate, position Y coordinate). For example, the location of RES001 is (10.5, 25.3). The resources are spatially located by their arrangement and numbering sequence in the actual site. The coordinates here are the spatial positioning results. Combined with the orientation angle of the resource receiving surface, for example, the orientation angle of the RES001 electronic screen is 90 degrees, indicating that its display surface faces east (assuming the positive X-axis is east and the positive Y-axis is north), a resource spatial location set is generated. This is a data structure containing each resource ID, location coordinate, and orientation angle, for example, {RES001: ((10.5, 25.3), 90°), RES002: ((12.0, 30.1), 180°), …}.

[0075] The propagation path mapping submodule calls the resource spatial location set, collects the propagation altitude angle, propagation intensity measurement value, and resource surface orientation azimuth, calculates the propagation path direction corresponding to the resource receiving surface based on the propagation direction vector, projects the path direction to the plane coordinate, determines whether there is overlap with the coverage range of the propagation efficiency distribution information, obtains the spatial path segment where the propagation and resource areas coincide, and generates the path resource coincidence distribution data;

[0076] Call the resource space position set, for example, including the position (10.5, 25.3) and orientation of 90 degrees of the RES001 electronic screen, and collect the propagation altitude angle of the electronic screen. For example, if the center of the screen is 2 meters above the ground and the average visual height of the viewer is 1.5 meters, then the relative height is 0.5 meters. If the effective vertical propagation angle of the screen is 15 degrees up and down, then the propagation altitude angle range can be regarded as an interval. Collect the propagation intensity measurement value. For example, the light intensity of RES001 at a distance of 1 meter is 300 lux, and its effective propagation distance is set to 5 meters (the brightness is considered invalid when it decays below 50 lux). This effective distance is determined by experiment. The recognizability of screen content at different distances under different ambient light conditions is tested. The average maximum distance that 80% of viewers can clearly identify is selected as the effective propagation distance, for example, 5 meters, and the azimuth angle of the resource surface, that is, 90 degrees. Based on the propagation direction vector, which is defined by the resource position, orientation angle, effective propagation distance, and horizontal propagation sector (for example, the 120-degree sector area in front of the electronic screen), calculate the resource The receiving surface of RES001, i.e., the screen itself, corresponds to the propagation path direction. This is a 120-degree fan-shaped area starting from (10.5, 25.3), extending 5 meters in the 90-degree direction. This path direction area is projected onto plane coordinates to obtain a geometric description of the fan-shaped area. It is then determined whether this fan-shaped area spatially overlaps with the high-efficiency area (the intersection of the path and behavior matching area) marked in the aforementioned propagation efficiency distribution information. Geometric operations are performed, such as calculating the intersection area of ​​two polygons. If the intersection area is greater than a set minimum effective overlap area, such as 1 square meter (this value is set based on the resource influence range and site accuracy; for example, a small display board has a small influence range, the value can be smaller), then overlap is considered. The spatial path segment where the propagation and resource areas overlap is obtained, i.e., the intersection polygon of the fan-shaped area and the high-efficiency area. After performing this operation for all resources, path resource overlap distribution data is generated. This data identifies which resources' service ranges effectively cover areas with high propagation efficiency.

[0077] The intersection block identification submodule extracts the coordinate point set corresponding to the resource segment based on the path resource overlap distribution data, identifies the numbered area to which the coordinate point belongs in the resource arrangement space, cross-checks the area boundary with the resource path boundary, extracts the resource number index corresponding to the intersection point, determines the resource number and corresponding coordinate area of ​​the discontinuous propagation area, and establishes the resource adaptation optimization block coordinate set;

[0078] According to the path resource overlap distribution data, which indicates the overlap between the effective propagation range of each resource (such as RES001) and the high propagation efficiency area, the coordinate point set corresponding to these overlapping resource segments is extracted. For example, the intersection of the effective propagation sector of RES001 and a high efficiency area A is an irregular polygon S. RA , whose vertex coordinates are {(x ra1 ,y ra1 ),(x ra2 ,y ra2 ),…}, identify the numbered areas to which these coordinate points belong in the resource arrangement space. Since the resources here are point-shaped or have a fixed orientation, their "numbered areas" are their own identifiers, such as RES001. The area boundary, that is, the boundary of the high-efficiency area A, and the resource path boundary, that is, the boundary of the effective propagation fan of RES001, are cross-judged. This has been completed through the overlap calculation in the previous step. Now we focus on those resources that fail to effectively cover the high-efficiency area, or the parts of the high-efficiency area that are not effectively covered by any resources. The resource number index corresponding to the intersection point is extracted. Here we mainly focus on those areas with high propagation efficiency but insufficient resource coverage or improper resource allocation. For example, if the high-efficiency area If most of A is not covered by any resources or is covered by a resource RES00X with a poor orientation, then area A is marked and associated with the resource RES00X that may need to be adjusted, or marked as needing new resources. Identify the discontinuous propagation areas, that is, the "holes" or "weak areas" of resource coverage in the high-efficiency area, and the corresponding resource numbers (if adjusting existing resources) or blank states (if adding new resources) and coordinate areas of these areas. Establish a resource adaptation optimization block coordinate set. For example, in the block 1 coordinate set {(ax1,ay1),…}, it is recommended to adjust the orientation of the resource RES00X, or in the block 2 coordinate set {(bx1,by1),…}, it is recommended to add new resources.

[0079] See also Figure 5 , the benefit evaluation generation module includes:

[0080] The node signal acquisition submodule calls the resource adaptation optimization block coordinate set, extracts the corresponding resource number index at the propagation level, collects the unit resource output value and propagation count value recorded by resource monitoring, constructs the propagation status data group of each resource at the current moment, and generates a resource operation signal set;

[0081] Call the resource adaptation optimization block coordinate set, for example, block 1. The block recommends adjusting the resource RES00X, extracting the corresponding resource number index, i.e. RES00X, at the communication level, collecting the monitoring record data of the resource RES00X before and after the optimization adjustment (e.g., adjusting the orientation), including the unit resource output value. For example, if RES00X is an electronic screen, the effective number of viewers per unit time (e.g., per hour) before the optimization is 50, and the effective number of viewers after the optimization adjustment (e.g., adjusting its orientation from 30 degrees to 75 degrees to align with the flow of people) is 80, as well as the communication Count values, such as its energy consumption is 0.5 kWh per hour before and after optimization, construct a propagation status data group for each resource at the current moment. For example, for RES00X, the state before optimization is: {resource ID: RES00X, time: T1, effective number of viewers: 50, energy consumption: 0.5 kWh}, and the state after optimization is: {resource ID: RES00X, time: T2, effective number of viewers: 80, energy consumption: 0.5 kWh}. Similar data collection is performed on all resources involved in optimization and adjustment and newly added resources (if any) to generate a resource operation signal set.

[0082] The trend difference marking submodule obtains the benefit output values ​​of resources in the propagation area and resources in the adjacent non-propagation area at the same time point based on the resource operation signal set, compares the difference in change direction, identifies signal items with inconsistent directions, records the difference items in the form of symbolic values, and classifies and organizes them into a separate input factor sequence to obtain a benefit trend difference vector group;

[0083] According to the resource operation signal set, the resources in the propagation area are obtained, for example, after optimization, RES00X, whose benefit output value is 80 people / hour, and the resources in the adjacent non-propagation area (or the state before optimization), for example, RES00X before optimization, whose benefit output value is 50 people / hour. At the same time point, the benefit output values ​​are compared. After optimization, the benefit of RES00X changes from 50 people / hour to 80 people / hour, with a change of +30 people / hour, and a positive change direction. Assume that there is another similar resource RES00Y that has not been optimized. Its benefit output in the same period changes from 45 people / hour to 40 people / hour, with a change of -5 people / hour, and a negative change direction. Identify signal items with inconsistent directions. Here, RES00X (after optimization compared to before optimization) is positive, and RES00Y (comparison during the same period) is negative. There is a trend difference between the two. These difference items are recorded in the form of symbolic values. For example, define an optimization effect indicator E = (benefit 优化后 -benefit 优化前 ) / Before benefit optimization, RES00X’s E X =(80-50) / 50=0.6. If a benchmark improvement rate is set, for example, 0.1 (i.e. 10%), then EX >0.1, marked as +1, if the change of RES00Y is regarded as the reference, its E Y =(40-45) / 45≈-0.11, marked as -1, and classified into a separate input factor sequence. For example, the factor sequence contains one item {resource ID: RES00X, optimization effect mark: +1, specific benefit change rate: 0.6}, and the benefit trend difference vector group is obtained. This vector group summarizes the relative benefit changes brought about by each optimization measure.

[0084] The optimization information generation submodule calls the benefit trend difference vector group, inputs the neural network input sequence structure completed through historical cycle sample training, inputs the vector group and the constructed input factor items in parallel, and performs combined optimization deduction based on the input factors of communication intensity, resource density and audience behavior data. The optimized value distribution results at the corresponding time point are obtained, and the economic benefit evaluation information of the optimization impact is generated;

[0085] Call the benefit trend difference vector group, for example, including {resource ID: RES00X, optimization effect tag: +1, specific benefit change rate: 0.6}, input historical cycle samples, such as similar optimization cases and their effect data in the past 6 months, and the trained neural network input sequence structure. The sequence structure may contain the following fields: [timestamp, resource ID, optimization type (such as orientation adjustment, content update, location move), pre-optimization communication intensity (such as viewing distance), pre-optimization resource density (such as the number of similar resources within 10 meters), pre-optimization average audience stay time, benefit trend difference tag, actual benefit change rate]. The benefit trend difference vector group of the current optimization case, such as RES00X's optimization effect tag +1, and the constructed input factor items, such as RES00X's optimization type of "orientation adjustment", pre-adjustment communication intensity estimated as 2 (a comprehensive indicator based on distance and orientation, 1-5 levels), pre-adjustment resource density of 3, and pre-adjustment average audience stay time of 2 minutes, are input into the sequence structure in parallel to form New input data is combined and optimized based on input factors such as communication intensity, resource density, and audience behavior data. This process is executed by a trained computational model. Internally, the model maps input factors to expected economic impacts through a series of weighted summations and nonlinear transformations, yielding the distribution of optimized values ​​at the corresponding time point. For example, for RES00X's "direction adjustment," the model estimates the potential economic benefit increase to be 500 yuan / day (calculated by multiplying the increase in viewers by the average potential consumption conversion value per viewer). This value is calculated based on the average increase in advertising revenue or related product sales corresponding to similar increases in viewership in historical data. For example, historical data shows that every 10 viewers / hour can generate an average daily economic benefit of approximately 166 yuan. Therefore, an increase of 30 viewers / hour corresponds to 3×166≈500 yuan / day, generating the economic benefit assessment information for the optimization impact: {Resource ID: RES00X, Optimization measure: Direction adjustment, Estimated economic benefit increase: 500 yuan / day}.

[0086] See also Figure 6 , the credibility assignment module includes:

[0087] The deviation value calculation submodule calls the optimization impact economic benefit evaluation information, obtains the optimized output sequence and actual output sequence of each resource in the optimization time period, constructs a set of deviation differences between corresponding sequence indexes, and normalizes the deviation ratio based on the optimization output to obtain the normalized benefit deviation index.

[0088] The optimization impact economic benefit evaluation information is called, for example, the estimated economic benefit of RES00X is increased by 500 yuan / day, and the optimized output sequence of each resource within the optimization period, for example, one week (7 days) after the optimization is implemented, is obtained, that is, the estimated economic benefit improvement value is 500 yuan per day, totaling 500×7=3500 yuan. Compared with the actual output sequence, by tracking the actual income or conversion indicators related to the resource, such as the total sales increase of a cultural and creative product directly related to RES00X in the week after its optimization, or the economic value converted by the brand influence improvement evaluated by the survey questionnaire, assuming that the actual observed total economic benefit increase brought about by the RES00X optimization in one week is 3200 yuan, a set of deviation differences between the corresponding sequence indexes is constructed. Here, it is a comparison of a single total value, and the deviation is 3500-3200=300 yuan. The deviation ratio is normalized in combination with the optimization output to calculate the normalized benefit deviation index I B =(estimated value - actual value) / estimated value=(3500-3200) / 3500=300 / 3500≈0.0857. If calculated on a daily basis, the daily estimate is 500 yuan. Assuming that the actual daily benefits are 480, 510, 490, 470, 520, 460, and 470 yuan respectively, the daily deviations are 20, -10, 10, 30, -20, 40, and 30 yuan, and the daily normalized deviations are 0.04, -0.02, 0.02, 0.06, -0.04, 0.08, and 0.06.

[0089] The deviation segmentation submodule sets the confidence deviation threshold according to the normalized benefit deviation index, groups and classifies the index sequence according to the upper limit of the threshold, establishes index labels for the time periods corresponding to the groups, and numbers the deviation levels corresponding to the groups in sequence to obtain the deviation level mapping interval value;

[0090] Based on the normalized benefit deviation index, for example, the total deviation index of RES00X obtained in the previous period is 0.0857, or the daily deviation index sequence {0.04, -0.02, 0.02, 0.06, -0.04, 0.08, 0.06}, the confidence deviation threshold is called and set. The threshold is set according to the acceptable error range of the forecast model. For example, based on historical model performance and business needs, three levels of confidence deviation threshold upper limits are set: level 1 (high confidence) deviation absolute value ≤ 0.05, level 2 (medium confidence) deviation absolute value [0.05, 0.15], and level 3 (low confidence) deviation absolute value > 0.15. These thresholds are set based on retrospective analysis of historical forecast data and statistics on the actual application value and risk of forecast results within different deviation ranges. For example, forecasts with a deviation of less than or equal to 5% are almost risk-free in decision support, and forecasts with a deviation of 5%-15% are risk-free. Expert opinions are needed. For predictions greater than 15%, it is recommended to review the model or input data. The exponential series should be grouped and classified according to the upper threshold. For example, for the sequence {0.04, -0.02, 0.02, 0.06, -0.04, 0.08, 0.06}, its absolute value sequence is {0.04, 0.02, 0.02, 0.06, 0.04, 0.08, 0.06}, and the corresponding deviation levels are: level 1, level 1, level 1, level 2, level 1, level 2, level 2. An index label is established for the time period corresponding to the group. For example, the first to third days and the fifth day are level 1, and the fourth, sixth, and seventh days are level 2. The deviation levels corresponding to the groups are numbered in sequence to obtain the deviation level mapping interval value, for example, {RES00X, day 1: level 1, RES00X, day 2: level 1, …, RES00X, day 4: level 2, …}.

[0091] The trusted level labeling submodule calls the deviation level mapping interval value, establishes the trusted level interval corresponding benchmark based on the time mapping table of level interval and resource number, outputs the trusted label result set corresponding to the resource in each time period, and outputs the trusted level information of the optimized coverage area;

[0092] Call the deviation level mapping interval value, for example, {RES00X, day 1: level 1, RES00X, day 4: level 2}, and establish a trust level interval corresponding benchmark based on the time mapping table of level intervals and resource numbers. This table is the deviation level mapping interval value itself, or a more comprehensive table containing the rating results of multiple resources and multiple time periods. That is, level 1 represents high trust, level 2 represents medium trust, and level 3 represents low trust. Output the trust label result set corresponding to the resource in each time period. For example, the trustworthiness of the optimization benefit evaluation result of resource RES00X on the first day is The credibility level is "high" and the credibility level on the fourth day is "medium". This information is summarized and the credibility level information of all evaluated resources in the optimization coverage area at each evaluation time point is output. For example, a report is generated with the content: "In area Z01, the economic benefit evaluation results of resource RES00X from June 1 to June 3 and June 5, 2024 are high credibility, and the economic benefit evaluation results on June 4, June 6, and June 7 are medium credibility; resource RES00Y on...", and finally a credibility map or list of the benefit evaluation results of the entire optimization plan is formed.

[0093] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. An AI-based cultural communication economic benefit optimization system, characterized by: The system comprises: The cultural value extraction module obtains the cultural theme characteristics of each activity in the historical communication activity data, compares the theme popularity change value with the audience feedback vector, determines whether the change direction of the theme popularity has consistent characteristics, and generates the core value trend value of cultural communication; The communication efficiency analysis module calls the cultural communication core value trend value, combines it with the behavior heat map provided by the audience behavior monitoring device, and generates communication efficiency distribution information based on the cross-projection contour of the heat map coverage area and the communication path; The resource adaptation optimization module obtains the cultural communication resource allocation structure data based on the communication efficiency distribution information, performs spatial overlap judgment on the communication path and the resource allocation image, identifies the communication segments with uneven resource allocation, and generates a resource adaptation optimization block coordinate set; The benefit evaluation generation module calls the resource adaptation optimization block coordinate set, obtains the number index of the coverage node at the propagation level, marks the output change trend difference item, inputs the difference item as an external optimization feature into the optimization input sequence, and generates optimization impact economic benefit evaluation information.

2. The AI-based cultural communication economic benefit optimization system according to claim 1 is characterized by: The core value trend value of cultural communication includes the fluctuation amplitude of topic popularity, the density of value direction change, the consistency level of audience feedback and the identification mark of topic mutation; the communication efficiency distribution information includes the main axis angle distribution of overlapping areas, the communication coverage breadth index, the audience behavior classification level and the projection area number mapping table; the resource adaptation optimization block coordinate set is specifically the node projection intersection boundary point, the correspondence between the communication area and the resource arrangement, the path occlusion block index and the spatial overlap density factor; the optimization impact economic benefit evaluation information includes the external resource interference feature sequence, the node-level response trend symbol sequence, the neural network prediction value set and the evaluation segment index structure.

3. The AI-based cultural communication economic benefit optimization system according to claim 1 is characterized by: The cultural value extraction module includes: The topic feature extraction submodule obtains the cultural theme features of each activity in the historical communication activity data, extracts the behavioral boundary points in each activity whose attention is greater than the set topic popularity threshold, establishes a topic point set, and constructs a topic boundary line structure sequence based on the relative positions of the topic points; The heat change calculation submodule calls the theme boundary line structure sequence, extracts the coordinate values ​​of corresponding boundary point pairs in adjacent activity theme line sequences, calculates relative displacement vectors, constructs an angle sequence set based on the angles between points, performs ratio operations on the boundary angle sequences of continuous activities, calculates the mean value of the boundary change gradient, screens the sections where the angle change exceeds the change gradient threshold, and generates a theme direction change gradient value set; The feedback consistency judgment submodule calls the theme direction change gradient value set, retrieves the average feedback vector recorded by the audience behavior monitoring device, performs projection operation on the direction change gradient, compares the angle between it and the feedback direction, and screens the segments based on the set consistency threshold to obtain the proportion of the number of segments whose theme change direction is consistent with the feedback direction, and obtains the core value trend value of cultural communication.

4. The AI-based cultural communication economic benefit optimization system according to claim 1 is characterized by: The transmission efficiency analysis module includes: The behavior direction extraction submodule calls the core value trend value of cultural communication, obtains the behavior heat map recorded by the audience behavior monitoring device, identifies the boundary of the medium behavior area in the heat map, determines the behavior distribution contour based on the set of isovalue contour lines, extracts the direction distribution of the boundary line and performs vectorized fitting in the position order to generate a behavior main axis direction sequence; The angle matching calculation submodule obtains the angular relationship between the vector segment and the theme direction segment marked in the core value trend value of cultural communication based on the main axis direction sequence of the behavior, sequentially compares the angles between the two sets of direction vectors in the spatial coordinate plane, selects vector pairs whose angle deviation values ​​are lower than the set threshold of the propagation angle, calculates and obtains the regional direction deviation index value, and uses the blocks with deviation values ​​lower than the propagation angle threshold as matching areas to generate a direction matching propagation distribution value set; The spatial cross-construction submodule matches the propagation distribution value set according to the direction, retrieves the propagation path information within the time period, establishes a propagation vector projection layer in a unified spatial coordinate system, locates the projection boundary intersection block between the propagation path and the marked behavior area, extracts the boundary line index value of the overlapping block and generates a spatial geometric coverage group to establish the propagation efficiency distribution information.

5. The AI-based cultural communication economic benefit optimization system according to claim 1 is characterized by: The resource adaptation optimization module includes: The resource arrangement extraction submodule obtains the communication efficiency distribution information, collects the cultural communication resource allocation structure data, extracts the two-dimensional spatial position index corresponding to the resource number grid, spatially locates the arrangement rows and numbering sequence of the resources in the actual site, and generates a resource spatial position set based on the orientation angle value of the resource receiving surface; The propagation path mapping submodule calls the resource spatial location set, collects the propagation altitude angle, propagation intensity measurement value, and resource surface orientation azimuth, calculates the propagation path direction corresponding to the resource receiving surface based on the propagation direction vector, projects the path direction to the plane coordinate, determines whether there is overlap with the coverage range of the propagation efficiency distribution information, obtains the spatial path segment where the propagation and resource areas overlap, and generates path resource overlap distribution data; The intersection block identification submodule extracts the coordinate point set corresponding to the resource segment based on the path resource overlap distribution data, identifies the numbered area to which the coordinate point belongs in the resource arrangement space, cross-judges the area boundary with the resource path boundary, extracts the resource number index corresponding to the intersection point, determines the resource number of the discontinuous propagation area and the corresponding coordinate area, and establishes a resource adaptation optimization block coordinate set.

6. The AI-based cultural communication economic benefit optimization system according to claim 1 is characterized by: The benefit evaluation generation module includes: The node signal acquisition submodule calls the resource adaptation optimization block coordinate set, extracts the corresponding resource number index at the propagation level, collects the unit resource output value and propagation count value recorded by resource monitoring, constructs the propagation status data group of each resource at the current moment, and generates a resource operation signal set; The trend difference marking submodule obtains the benefit output values ​​of the resources in the propagation area and the resources in the adjacent non-propagation area at the same time point based on the resource operation signal set, compares the difference in change direction, identifies signal items with inconsistent directions, records the difference items in the form of symbolic values, classifies and organizes them into a separate input factor sequence, and obtains a benefit trend difference vector group; The optimization information generation submodule calls the benefit trend difference vector group, inputs the neural network input sequence structure completed by historical period sample training, inputs the vector group and the constructed input factor items in parallel, performs combined optimization deduction based on the input factors of communication intensity, resource density and audience behavior data, obtains the optimization value distribution results at the corresponding time point, and generates optimization impact economic benefit evaluation information.

7. The AI-based cultural communication economic benefit optimization system according to claim 1 is characterized by: The system further comprises: The credibility assignment module calls the optimization impact economic benefit evaluation information, obtains the optimization value and the actual output value of the resource at each optimization time point, compares the degree of deviation between the optimization value and the actual output value, and compares it with the set confidence deviation threshold, divides the deviation segment above the threshold into intervals according to the level division benchmark, marks each level division segment as a credibility level identifier, and generates credibility level information for the optimization coverage area; The trust level information of the optimized coverage area specifically refers to the deviation interval level index, the credibility label mapping result, the deviation fluctuation trend group and the credibility judgment result number table.

8. The AI-based cultural communication economic benefit optimization system according to claim 7 is characterized in that: The credibility assignment module includes: The deviation value calculation submodule calls the optimization impact economic benefit evaluation information, obtains the optimized output sequence and the actual output sequence of each resource in the optimization time period, constructs a deviation difference set between the corresponding sequence indexes, and normalizes the deviation ratio based on the optimization output to obtain a normalized benefit deviation index. The deviation segmentation submodule sets a confidence deviation threshold based on the normalized benefit deviation index, groups and classifies the index sequence according to the upper limit of the threshold, establishes index labels for time periods corresponding to the groups, and numbers the deviation levels corresponding to the groups in sequence to obtain deviation level mapping interval values; The trusted level labeling submodule calls the deviation level mapping interval value, establishes a trusted level interval corresponding benchmark based on the time mapping table of level interval and resource number, outputs the trusted label result set corresponding to the resource in each time period, and outputs the trusted level information of the optimized coverage area.

Images