System and method for intelligent operations and visitor experience optimization for museums and art exhibitions

Abstract

The application discloses a museum and art exhibition intelligent operation and visitor experience optimization system and method, relates to the technical field of intelligent management of museums and art exhibitions, and comprises the following steps: obtaining interactive behavior data and identity data of each target object; performing group projection on the interactive behavior data and the identity data, generating group structure data, and performing hierarchical attribution division on the group structure data from an identity dimension, to generate exhibition participation data of each group; and performing scheduling analysis on the correlation degree of the target objects contained in the corresponding group according to the exhibition participation data, to generate behavior-identity deviation data. Through the hierarchical allocation and dynamic reallocation mechanism, the scheme can accurately reflect the resource demand differences of different group levels, and solves the problem of insufficient display strategy adaptability caused by ignoring the behavior differences between group levels in the prior art.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent management of museums and art exhibitions, in particular to an intelligent operation and audience experience optimization system and method for museums and art exhibitions. BACKGROUND

[0002] In the field of modern culture and art, museums and various art exhibition venues bear the important functions of displaying cultural art, educating the public, and improving audience experience. With the development of technology, especially the widespread application of big data analysis, artificial intelligence, and real-time sensing technology, exhibition operation management has gradually shifted from traditional manual monitoring and static arrangement to an intelligent, data-driven, and refined management model. In this technical environment, operators not only focus on exhibition layout, exhibit protection, and visitor number statistics, but also need to understand audience behavior patterns, interest preferences, and group interaction characteristics in real time, so as to make scientific operation adjustments and personalized display strategies.

[0003] However, in the process of implementing the technical solutions of the embodiments of the present application, it is found that the above-mentioned technologies at least have the following technical problems:

[0004] In the prior art, the interactive behavior data and identity data of target objects are usually collected by cameras, sensors, or electronic ticketing systems, and statistical analysis or pattern recognition is performed. The above methods can provide overall trend reference when dealing with large-scale exhibition crowds, helping operators understand exhibition hotspots and peak periods, but they ignore the behavior differences within the group and the potential interaction feedback effect, making it difficult to accurately grasp the behavior deviation and mutual influence of each target object in the group, as well as to respond to the behavior deviation of the target object in a timely manner, ultimately resulting in insufficient adaptability of the generated hierarchical display scheme. SUMMARY

[0005] The purpose of the present application is to provide an intelligent operation and audience experience optimization system and method for museums and art exhibitions to solve the problems raised in the background art.

[0006] In order to achieve the above-mentioned purpose, the technical solutions of the present application are as follows:

[0007] In a first aspect, the present application discloses an intelligent operation and audience experience optimization method for museums and art exhibitions, comprising the following steps:

[0008] Obtaining interactive behavior data and identity data of each target object;

[0009] Projecting the interactive behavior data and the identity data into groups to generate group structure data, and performing hierarchical attribution division on the group structure data from the identity dimension to generate exhibit participation data of each group;

[0010] The behavior-identity deviation data and the interactive behavior data of different target objects are input into a pre-constructed behavior response model, and a behavior influence result of the corresponding target object is output, and the behavior influence result is compared and analyzed with the interactive behavior data;

[0011] The behavior-identity deviation data and the interactive behavior data of different target objects are input into a pre-constructed behavior response model, and a behavior influence result of the corresponding target object is output, and the behavior influence result is compared and analyzed with the interactive behavior data;

[0012] According to the comparison and analysis result, the exhibition participation data of the group to which the corresponding target object belongs is overall scheduled, and the overall scheduling result is matched with a preset display strategy library to generate a hierarchical display scheme.

[0013] In a second aspect, the present application discloses a museum and art exhibition intelligent operation and audience experience optimization system, comprising:

[0014] A data acquisition module is configured to acquire interactive behavior data and identity data of each target object.

[0015] A data processing module is configured to project the interactive behavior data and the identity data in groups to generate group structure data, and to divide the group structure data into layers according to identity dimensions to generate exhibition participation data of each group.

[0016] A behavior deviation analysis module is configured to schedule analysis of the correlation degree of target objects contained in the corresponding group according to the exhibition participation data, and to generate behavior-identity deviation data.

[0017] A behavior influence analysis module is configured to input the behavior-identity deviation data and the interactive behavior data of different target objects into a pre-constructed behavior response model, output a behavior influence result of the corresponding target object, and compare and analyze the behavior influence result with the interactive behavior data.

[0018] A hierarchical display scheme generation module is configured to overall schedule the exhibition participation data of the group to which the corresponding target object belongs according to the comparison and analysis result, and match the overall scheduling result with a preset display strategy library to generate a hierarchical display scheme.

[0019] Compared with the prior art, the present application has the following advantages:

[0020] 1. The present application can accurately reflect the resource demand differences of different group levels through hierarchical allocation and dynamic reallocation mechanism, solve the problem of insufficient display strategy adaptability caused by ignoring the behavior differences between group levels in the prior art, and realize real-time optimization of allocation strategy by combining the original attribution degree of the target object, thereby improving the matching accuracy of the hierarchical display scheme and the behavior characteristics of the target object.

[0021] 2、The present application introduces behavior-identity composite signature and stability score, combined with dynamic attribution threshold interval, can adaptively correct the group division result, solve the group division deviation problem caused by ignoring the dynamic correlation between behavior and identity in the prior art, and significantly improve the generation accuracy of exhibit participation data, provide reliable basis for the adaptability optimization of subsequent hierarchical display scheme. BRIEF DESCRIPTION OF DRAWINGS

[0022] The disclosure of the present application will be described with reference to the accompanying drawings. It should be understood that the drawings are only for illustrative purposes, and are not intended to limit the scope of protection of the present application. In the drawings, the same reference numerals are used to refer to the same parts. Among them:

[0023] Figure 1 The step flow chart of the museum and art exhibition intelligent operation and visitor experience optimization method of the present application;

[0024] Figure 2 The flowchart for generating behavior-identity composite signature provided by the present application;

[0025] Figure 3 The flowchart for generating behavior-identity deviation data provided by the present application;

[0026] Figure 4 The flowchart for generating overall scheduling results provided by the present application;

[0027] Figure 5 The module function diagram of the museum and art exhibition intelligent operation and visitor experience optimization system provided by the present application. DETAILED DESCRIPTION

[0028] It is easy to understand that according to the technical scheme of the present application, those skilled in the art can propose a plurality of structure modes and implementation modes which can be replaced with each other without changing the essential spirit of the present application. Therefore, the following specific embodiments and drawings are only exemplary description of the technical scheme of the present application, and should not be regarded as the whole or as the limitation or restriction of the technical scheme of the present application.

[0029] SUMMARY OF THE APPLICATION:

[0030] In the prior art, the museum and art exhibition venue collects target object behavior data and identity information through camera, sensor or electronic ticket system, generates operation strategy based on statistical analysis or pattern recognition. This kind of method can identify hot area and peak period, but does not consider the behavior difference and dynamic interaction effect within the group, which leads to the difficulty in accurately capturing the behavior deviation of individual in the group, and the difficulty in adjusting the display strategy in time.

[0031] In order to solve the above problems, a method capable of fine analysis of the behavior characteristics of the group and dynamic adjustment of the display scheme is needed. First, how to extract the group structure characteristics from massive data and divide the hierarchical attribution becomes the key. Second, the correlation degree calculation problem between individual behavior and group characteristics needs to be solved to identify behavior deviation. Finally, how to dynamically adjust the display strategy according to the real-time analysis result becomes a technical difficulty. By introducing the group projection technology to structure the mapping of behavior and identity data, combining the hierarchical attribution division to identify the differences of group characteristics, and constructing the behavior response model to predict the influence of individual behavior, a dynamic scheduling mechanism is formed, and finally the precise adaptation of the hierarchical display scheme is realized.

[0032] After introducing the basic concept of the application, the embodiments of the application will be specifically introduced with reference to the drawings.

[0033] Embodiment one:

[0034] Please refer to Figure 1 , the intelligent operation and visitor experience optimization method of museum and art exhibition, which comprises the following steps:

[0035] Obtain the interactive behavior data and identity data of each target object;

[0036] Perform group projection on the interactive behavior data and the identity data to generate group structure data, and perform hierarchical attribution division on the group structure data from the identity dimension to generate exhibit participation data of each group;

[0037] According to the exhibit participation data, the correlation degree of the target objects contained in the corresponding group is analyzed to generate behavior-identity deviation data; wherein the correlation degree is obtained by matching and analyzing the interactive behavior data and the identity data;

[0038] Input the behavior-identity deviation data and the interactive behavior data of different target objects into the pre-constructed behavior response model to output the behavior influence result of the corresponding target object, and compare and analyze the behavior influence result and the interactive behavior data;

[0039] According to the comparison and analysis result, the exhibit participation data of the group to which the corresponding target object belongs is overall scheduled, and the overall scheduling result is matched with the preset display strategy library to generate a hierarchical display scheme.

[0040] Among them, the interactive behavior data refers to the collection of quantifiable behavior information generated by each target object (single visitor) in the museum and art exhibition during the visit;

[0041] Identity data refers to the overall data set of unique identity information and static attributes of each target object in the museum and art exhibition environment;

[0042] Group projection refers to the process of mapping and integrating the interaction behavior data and identity data of each target object according to the group dimension to generate group structure data that can reflect the group behavior pattern and group structure characteristics;

[0043] Group structure data refers to the overall description data generated by group projection operation based on the interaction behavior data and identity data of all target objects;

[0044] Hierarchical attribution division refers to the process of hierarchical analysis of the identity characteristics of each target object in the group structure data;

[0045] Exhibition participation data refers to a structured data set of the participation degree, preference and association state of each exhibition by the target object, generated by analyzing its interaction behavior data and identity data;

[0046] Correlation degree refers to a quantitative indicator of the degree of mutual connection between target objects or between target objects and exhibitions in the dimensions of interaction behavior and identity characteristics;

[0047] Dispatch analysis refers to the process of systematic analysis of the association between target objects, participation intensity and behavior deviation within the group based on group exhibition participation data;

[0048] Behavior-identity deviation data refers to a data set used to represent the difference between the actual interaction behavior of each target object and the expected behavior of its identity characteristics;

[0049] Pre-built behavior response model refers to a mathematical / algorithms model based on historical data and group behavior rules;

[0050] Behavior impact result refers to a quantitative representation of the behavior of a target object that may be affected by the behavior of other target objects and its own characteristics under group interaction;

[0051] Overall scheduling refers to the process of systematic and overall allocation and optimization of the exhibition participation data of each target object in the group based on the existing group exhibition participation data and behavior-identity deviation analysis results;

[0052] Pre-set display strategy library refers to a database providing rules for guiding the hierarchical display of exhibitions and group management;

[0053] Hierarchical display scheme refers to a structured scheme for guiding the exhibition display order, priority and interaction form in museums or art exhibitions, generated by dispatch analysis and strategy matching based on the interaction behavior data, identity data and group structure analysis results of target objects.

[0054] The present scheme can accurately capture the behavior characteristic differences of different identity groups through group projection and hierarchical division technology; the interaction influence of individuals and groups is quantified by constructing a behavior response model to realize dynamic monitoring of behavior deviation; the matching degree of the group structure data at time t and the interactive behavior data at time t is adjusted based on the real-time comparative analysis results to significantly improve the adaptability of the scheme to the needs of target objects; through the above technical solutions, the present application can effectively solve the problem of insufficient recognition of internal behavior differences in groups, and by dynamically analyzing the correlation between individual behavior and group characteristics, the exhibit participation data distribution strategy is adjusted in real time, and at the same time, based on the behavior response model to predict the interaction influence of the group, the real-time and accuracy of the hierarchical display scheme is improved, and finally the individual optimization of target object experience is realized.

[0055] As introduced above is the complete scheme of the museum and art exhibition intelligent operation and visitor experience optimization method, the following introduces obtaining the interactive behavior data and identity data of each target object, specifically including:

[0056] Obtaining the interactive behavior data of each target object through an online interactive interface; the interactive behavior data includes but is not limited to access path, stay time, interactive operation, click, browse or participate in activities, etc.;

[0057] Obtaining the identity data of each target object through an exhibition management terminal; the identity data includes but is not limited to the unique identifier of the target object, session ID, membership level and historical access record, etc.;

[0058] The interactive behavior data is mapped as a whole semantic block to a space-time grid, and the identity data is taken as a whole context modulator to dynamically adjust the space-time grid mapping result to generate group structure data , and the specific calculation formula is as follows:

[0059]

[0060] In the formula, is the projection function parameter determined by the identity data , and is a nonlinear projection function, and the above data is normalized when calculating;

[0061] The matching degree of the group structure data at time t and the interactive behavior data at time t is calculated, and the specific calculation formula is as follows: In the formula,

[0062]

[0063] In the formula, ​​represents a kernel mapping function, represents an inner product operation, represents a vector norm, and the above data is normalized when performing calculation.

[0064] determining whether the matching degree is greater than a preset discrimination threshold, otherwise reacquiring the interactive behavior data and the identity data of each target object.

[0065] The scheme jointly maps the interactive behavior data and the identity data as a whole object, dynamically adjusts the space-time grid projection by using the identity data as a context modulator, so that the generated group structure data can truly reflect the overall coupling relationship between behavior and identity, calculates the matching degree of the interactive behavior data and the group structure data by using kernel mapping, and controls the data effectiveness by using a discrimination threshold, so as to realize automatic screening and resampling of abnormal or inconsistent data, thereby ensuring the accuracy and reliability of the subsequently generated data.

[0066] As introduced above, the interactive behavior data and the identity data of each target object are acquired, and the following describes that after the group projection of the interactive behavior data and the identity data is performed to generate group structure data, the following still includes generating a behavior-identity composite signature, specifically including:

[0067] integrally mapping the interactive behavior data according to the identity data, and offset correcting the integrally mapped result according to the group structure data;

[0068] performing iterative re-encoding on the offset corrected integrally mapped result until iterative convergence, to generate a behavior-identity composite signature.

[0069] The integrally mapping refers to a process of combining the identity data (as a whole) to perform gating change on the interactive behavior data as an indivisible whole object;

[0070] The offset correction refers to a process of differentially adjusting the integrally mapped result of a single target object by using the group structure data in the process of generating the behavior-identity composite signature;

[0071] The iterative re-encoding refers to a process of embedding and aligning the dynamic characteristics of the interactive behavior data and the stable characteristics of the identity data through repeated nonlinear mapping and feedback correction steps on the basis of the offset corrected integrally mapped result;

[0072] The behavior-identity composite signature refers to comprehensive representation information generated after group structure analysis based on the interactive behavior data and the identity data of a target object.

[0073] The above content will be described in detail as follows:

[0074] The identity data The whole as a gate, to the interactive behavior data Overall mapping, generating initial behavior-identity composite signature And it as a whole mapping results, its specific formula as follows:

[0075]

[0076] In the formula, Indicates a nonlinear gate function, Indicates the element-wise multiplication (gate amplification or inhibition operation), the above data in the calculation are normalized;

[0077] According to the group structure data Offset correction to the overall mapping results, generating intermediate behavior-identity composite signature And it as an offset correction after the whole mapping results, its specific formula as follows:

[0078]

[0079] In the formula, Indicates a group reflection function, Indicates the reflection coefficient, the above data in the calculation are normalized;

[0080] The offset correction after the overall mapping results are executed iteration re-encoding, and in each iteration re-encoding process, the change amplitude of the offset correction after the overall mapping results and the offset correction after the overall mapping results generated by the last iteration re-encoding is obtained by difference calculation;

[0081] Determine whether the change amplitude is greater than the preset convergence threshold, otherwise it is determined that the iteration converges, and the change amplitude generated by all iteration re-encoding processes is normalized to generate stability score Its specific formula as follows:

[0082]

[0083] In the formula, Indicates the total number of iteration re-encoding, Indicates the change amplitude generated by the Iteration re-encoding process, the above data in the calculation are normalized;

[0084] And the corresponding offset correction after the whole mapping results as behavior-identity composite signature.

[0085] The scheme can dynamically eliminate group interference and optimize data expression form through offset correction and iterative re-encoding, so that the behavior-identity composite signature more accurately reflects the behavior characteristics of individuals in the group; Through the above technical scheme, the application solves the individual behavior analysis deviation problem caused by ignoring group dynamics in the prior art, generates stable behavior-identity composite signature, provides reliable data basis for subsequent group hierarchical attribution division, and improves the adaptability of hierarchical display scheme to actual needs of target objects.

[0086] As introduced above, after the group projection of the interaction behavior data and the identity data and the generation of the group structure data, the behavior-identity composite signature is generated. The following introduces the execution of iterative re-encoding on the offset-corrected overall mapping result until the iterative convergence, and the generation of the behavior-identity composite signature. Please refer to Figure 2 , Figure 2 The flowchart for generating the behavior-identity composite signature provided by the embodiment of the application is as follows:

[0087] The iterative re-encoding is performed on the offset-corrected overall mapping result, and in each iterative re-encoding process, the change amplitude of the offset-corrected overall mapping result after the current offset correction and the offset-corrected overall mapping result generated by the last iterative re-encoding is calculated.

[0088] It is judged whether the change amplitude is greater than a preset convergence threshold, otherwise it is determined that the iteration converges. The change amplitudes generated in all iterative re-encoding processes are normalized, a stability score is generated, and the corresponding offset-corrected overall mapping result is taken as the behavior-identity composite signature.

[0089] The change amplitude refers to the difference quantization value between the offset-corrected overall mapping results of two adjacent iterative re-encodings.

[0090] The preset convergence threshold refers to the upper limit value for measuring the change amplitude between the offset-corrected overall mapping results of two consecutive iterative re-encodings. The average fluctuation value and the standard deviation of the change amplitudes generated in all iterative re-encoding processes are calculated, and the average fluctuation value and the standard deviation are weighted to generate the convergence threshold.

[0091] The normalization processing refers to the quantization and standardization processing of the change amplitude of the offset-corrected overall mapping result in each round of iterative re-encoding and the result of the previous round.

[0092] The stability score refers to an index for quantifying the convergence degree of the behavior-identity composite signature in the iterative re-encoding process.

[0093] The above content has been described in detail in the above part, and will not be described in detail here.

[0094] The present scheme realizes adaptive iteration control through real-time comparison of variation amplitude and convergence threshold, ensures the efficiency and accuracy of composite signature generation, and further introduces stability score to provide reliability verification basis for subsequent group stratification, so as to avoid error attribution caused by iteration fluctuation; through the above technical scheme, the present application effectively solves the problems of unclear convergence condition and insufficient result reliability in the iteration and re-encoding process, ensures the stability of composite signature generation through dynamic convergence judgment mechanism, and provides reliability verification index for group stratification attribution division through the quantitative output of stability score, so as to improve the accuracy of exhibit participation data generation and optimize the adaptability of subsequent hierarchical display scheme.

[0095] As introduced above, the iteration and re-encoding is performed on the overall mapping result after offset correction until the iteration converges, and the behavior-identity composite signature is generated, and the following introduces the stratification attribution division of the group structure data from the identity dimension to generate the exhibit participation data of each group, which specifically includes:

[0096] The group structure data is stratified and attributed from the identity dimension according to the hierarchical judgment result to generate the exhibit participation data of each group;

[0097] The hierarchical judgment result is obtained by matching the behavior-identity composite signature with the predefined group hierarchy set, and the highest original attribution degree is weighted according to the stability score, and the hierarchical judgment is performed on the weighted result according to the preset attribution threshold interval.

[0098] The hierarchical judgment result refers to the data set representing the attribution strength and hierarchical category of each target object in different group hierarchies obtained by matching and analyzing the behavior-identity composite signature with the predefined group hierarchy set;

[0099] The predefined group hierarchy set refers to the pre-defined stratification standard set based on the identity dimension, which can be obtained by classifying and training historical identity data using clustering algorithm;

[0100] The original attribution degree refers to the data index for measuring the attribution strength of the target object in different group hierarchies;

[0101] The preset attribution threshold interval refers to the continuous or discrete numerical segmentation set used for hierarchical judgment of the weighted original attribution degree of each target object;

[0102] Hierarchical determination refers to the process of using the behavior-identity composite signature of each target object as the main identification feature, matching it with a predefined set of group hierarchies, and determining the level of the target object in the group hierarchical system after calculating the degree of matching and weighted correction.

[0103] The above content will be described in detail below:

[0104] The behavior-identity composite signature is matched with a predefined set of group levels to obtain the original affiliation degree of each target object relative to each group level. The specific calculation formula is as follows:

[0105]

[0106] In the formula, Represents the target object Relative to group hierarchy The original degree of belonging, Represents the target object Behavior-identity composite signature, Indicates group hierarchy The hierarchical template vectors, and all the above data have been normalized during the calculation;

[0107] The highest original affiliation degree among the stability scores is weighted and processed according to the following formula:

[0108]

[0109] In the formula, Represents the target object The weighted processing result Represents the target object The highest original affiliation, The weighting coefficients for the stability score indicate that all the above data have been normalized during the calculation.

[0110] Based on the identity data, weights are assigned to the weighted processing results from a dimensional perspective:

[0111] For identity data Dimension value Calculate the corresponding first digit in the weighted processing result of the overall semantic mapping result. Dimension's contribution ratio This is used as the weight for the allocation of the corresponding dimension, and the specific calculation formula is as follows:

[0112]

[0113] In the formula, This represents the total number of dimensions in the identity data. identity data The values of the dimensions are normalized when calculating the above data;

[0114] The average value of the assigned weight of each dimension is solved, and the average value solving result is multiplied by the weighted processing result to obtain the final weighted processing result;

[0115] According to the preset attribution threshold interval, the final weighted processing result is executed to obtain:

[0116] Determine whether the final weighted processing result is greater than the upper threshold of the preset attribution threshold interval, and if so, determine that the identity level of the corresponding target object is level one, and take it as the hierarchical judgment result of the corresponding target object;

[0117] Determine whether the final weighted processing result is less than the lower threshold of the preset attribution threshold interval, and if so, determine that the identity level of the corresponding target object is level three, and take it as the hierarchical judgment result of the corresponding target object;

[0118] Determine whether the final weighted processing result is between the preset attribution threshold interval, and if so, determine that the identity level of the corresponding target object is level two, and take it as the hierarchical judgment result of the corresponding target object;

[0119] Among them, the upper threshold of the preset attribution threshold interval is to count all original attribution degrees and calculate the comprehensive mean and comprehensive standard deviation, and the sum of the comprehensive mean and the comprehensive standard deviation is obtained;

[0120] The lower threshold of the preset attribution threshold interval is to count all original attribution degrees and calculate the comprehensive mean and comprehensive standard deviation, and the difference between the comprehensive mean and the comprehensive standard deviation is obtained;

[0121] According to the hierarchical judgment result, the group structure data is hierarchically attributed from the identity dimension to generate the exhibit participation data of each group.

[0122] The above technical scheme can adaptively correct the group division result by introducing behavior-identity composite signature and stability score, combining dynamic attribution threshold interval, thereby improving the accuracy and robustness of hierarchical attribution; Through the above technical scheme, the present application can solve the group division deviation problem caused by ignoring the dynamic correlation between behavior and identity in the prior art, and significantly improve the generation accuracy of the exhibit participation data by fusing the composite signature stability evaluation and the dynamic threshold judgment mechanism, thereby providing a reliable basis for the adaptability optimization of the subsequent hierarchical display scheme.

[0123] As introduced above is the hierarchical attribution division of the population structure data from the identity dimension, to generate the exhibit participation data of each population, below introduces the scheduling analysis of the correlation degree of the target objects contained in the corresponding population according to the exhibit participation data, to generate the behavior-identity deviation data, please refer to Figure 3 , Figure 3 The flowchart for generating the behavior-identity deviation data provided by the embodiment of the present application, the behavior-identity deviation data, specifically includes:

[0124] The distribution weight and the correlation degree are weighted and matched according to the target objects, to generate the weighted correlation data; wherein the distribution weight is obtained by calculating the relative distribution of the exhibit participation data;

[0125] The scheduling data of each target object is generated by sorting according to the weighted correlation data and the exhibit participation data, and the scheduling data is matched with the interactive behavior data, to generate the expected behavior data;

[0126] The expected behavior data and the correlation degree of the corresponding target object are compared and analyzed, to generate the behavior-identity deviation data;

[0127] Wherein, the correlation degree is obtained by matching and analyzing the interactive behavior data and the identity data.

[0128] Wherein, the weighted correlation data refers to the data set for describing the comprehensive correlation strength between the exhibit participation behavior and the identity characteristics of different target objects in the population;

[0129] The distribution weight refers to the relative numerical value for representing the contribution degree of each target object to the population exhibit participation data within the population hierarchy;

[0130] The scheduling data refers to the numerical or structured representation of the priority, order and allocation state of each target object in the population hierarchy or the exhibit resources in a specific population, according to the exhibit participation, population structure and correlation degree analysis results;

[0131] The expected behavior data refers to the predicted data set of the possible interactive behavior mode of each target object, which is derived by weighted matching and sorting calculation based on the exhibit participation and population scheduling analysis results of each target object in the population;

[0132] The difference comparison refers to the numerical or structural comparison between the expected behavior data of each target object and its corresponding actual or original correlation degree data, so as to quantify the deviation degree between the target object behavior and the population participation characteristics.

[0133] The above contents are described in detail as follows:

[0134] by the exhibit participation data of the crowd to obtain the distribution weight of the crowd , and the specific calculation formula is as follows:

[0135]

[0136] In the formula, denotes the crowd set, denotes the exhibit participation data of the crowd ;

[0137] The distribution weight of the crowd is matched with the association degree of the target object according to the target objects contained in the corresponding crowd to generate the weighted association data of the target object relative to the crowd , and the specific calculation formula is as follows:

[0138]

[0139] In the formula, denotes the target object set contained in the crowd , and the above data is normalized when calculating;

[0140] The association degree of the target object is obtained by matching and analyzing the interaction behavior data and the identity data, and the specific calculation formula is as follows:

[0141]

[0142] In the formula, denotes the interaction behavior data of the target object , and the above data is normalized when calculating;

[0143] For each target object, the weighted association data and the exhibit participation data are arranged according to the exhibit dimension to form a pairing vector of the target object and the exhibit participation feature. On this basis, the participation order of the target object on each exhibit is sorted according to the size of the weighted association data in the pairing vector to generate a scheduling sorting list, and the scheduling sorting list is matrixed to obtain a scheduling data matrix, wherein each row of the scheduling data matrix corresponds to a target object, each column corresponds to an exhibit, and the matrix element value represents the scheduling data of the target object on the corresponding exhibit.

[0144] The target object The scheduling data With the target object The interactive behavior data Perform alignment matching to generate the target object. Expected behavioral data The specific calculation formula is as follows:

[0145]

[0146] In the formula, Indicates the weighting coefficient. This represents the alignment matching function; all the data above has been normalized during the calculation.

[0147] target object Expected behavioral data With the target object The degree of correlation Perform difference comparison to generate target object Behavioral-identity bias data The specific calculation formula is as follows:

[0148]

[0149] In the formula, The above data are normalized during calculations and represent positive numbers.

[0150] This solution dynamically adjusts the correlation by introducing distributed weights and generates scheduling data by combining exhibit participation data. This enables more accurate capture of behavioral differences within a group. Furthermore, by comparing the expected behavioral data with the correlation, the degree of deviation between behavior and identity attributes is directly quantified. Through this technical solution, the application can adjust the scheduling analysis process based on the dynamic distribution characteristics of group participation, solving the problem of inaccurate deviation identification caused by ignoring group differences in existing technologies. The comparison of the expected behavioral data with the correlation further improves the accuracy of generating behavior-identity deviation data, providing a reliable data foundation for the dynamic optimization of subsequent hierarchical display schemes.

[0151] The above describes the scheduling analysis of the correlation between target objects within the corresponding group based on the exhibit participation data, generating behavior-identity bias data. The following describes how to input the behavior-identity bias data and the interaction behavior data of different target objects into a pre-built behavior response model, outputting the behavioral impact results of the corresponding target objects, specifically including:

[0152] inputting the behavior-identity deviation data and the interaction behavior data of different target objects into the pre-constructed behavior response model one by one, and outputting behavior influence initial values representing the influence of behaviors of other target objects;

[0153] performing group-by-group weighted calculation on the behavior influence initial values according to the distribution weights, and generating behavior influence results of each target object.

[0154] The behavior influence initial value refers to quantitative data representing the influence of behaviors of other target objects on a target object, which is calculated by the model after inputting the behavior-identity deviation data and the interaction behavior data of each target object into the pre-constructed behavior response model.

[0155] The group-by-group weighted calculation refers to a process of performing weighted average or weighted accumulation calculation on the behavior influence initial value of each target object according to the distribution weight of the group to which the target object belongs, so as to generate the behavior influence result of each target object.

[0156] The above content will be described in detail as follows:

[0157] constructing a behavior response model; the behavior response model comprises a response input layer, a deviation mapping layer, and a response output layer;

[0158] The response input layer is configured to receive behavior-identity deviation data, weighted association data, and interaction behavior data of a different target object of a target object, and to perform standardization processing on the received data, and map the behavior-identity deviation data and the interaction behavior data to a unified numerical range.

[0159] The deviation mapping layer is configured to perform weighted mapping on the behavior-identity deviation data and the weighted association data after the standardization processing, and to perform linear combination on the weighted mapping result and the interaction behavior data, so as to obtain behavior influence initial values representing the influence of behaviors of a different target object on a target object, thereby capturing the potential interaction between behaviors of different target objects.

[0160] The response output layer is configured to output the behavior influence initial values.

[0161] inputting the behavior-identity deviation data and the interaction behavior data of different target objects into the pre-constructed behavior response model one by one, and outputting behavior influence initial values representing the influence of behaviors of other target objects;

[0162] multiplying the distribution weights and the behavior influence initial values of the target objects included in the corresponding groups one by one, accumulating and averaging the multiplication results, accumulating the average values corresponding to each group, and generating behavior influence results of each target object caused by behaviors of all other target objects.

[0163] The behavior influence result is subjected to difference analysis with the interaction behavior data item by item, and the difference analysis result of each item is normalized to obtain a comparative analysis result.

[0164] The scheme can effectively distinguish the behavior influence strength of different identity groups by introducing distribution weights and sub-group calculation mechanism, thereby more accurately reflecting the actual interaction state within the group; through the above technical scheme, the application can accurately quantify the behavior influence strength of target objects in different groups, solve the influence evaluation deviation problem caused by group feature confusion in the prior art, and make the behavior influence values of professional audience groups and ordinary audience groups with high exhibit participation differ from each other, thereby providing accurate data support for generating a hierarchical display scheme adapted to different group features.

[0165] The above introduction is that the behavior-identity deviation data and the interaction behavior data of different target objects are input into a pre-constructed behavior response model, and the behavior influence result of the corresponding target object is output. The following introduces the overall scheduling of the exhibit participation data of the group to which the corresponding target object belongs according to the comparative analysis result, please refer to Figure 4 , Figure 4 The flowchart for generating an overall scheduling result provided by the embodiment of the application is shown in the figure, and the overall scheduling result is generated, which specifically includes:

[0166] According to the comparative analysis result, the exhibit participation data of the group to which the corresponding target object belongs is weighted and sorted to generate initial allocation data, and the initial allocation data is classified and statistically segmented in the group hierarchy range to form hierarchical allocation data.

[0167] The hierarchical allocation data is dynamically redistributed according to the original attribution of the target object in the group hierarchy to generate an overall scheduling result of the target object.

[0168] The initial allocation data refers to a data set obtained by weighting and sorting the exhibit participation data of the group to which each target object belongs based on the comparative analysis result;

[0169] Data classification and statistical segmentation refer to a process of classifying and organizing data according to certain rules according to the group hierarchy, attributes or preset dimensions, and statistically counting key numerical indicators in each class to form a data structure that can be used for subsequent dynamic redistribution;

[0170] The hierarchical allocation data refers to a data set obtained by classifying, statistically counting and segmenting the exhibit participation data of the target object group according to the group hierarchy structure;

[0171] The dynamic re-allocation refers to a process of real-time or condition-triggered re-allocation of the exhibit participation data of each target object based on the original attribution degree and the hierarchical allocation data of the target object.

[0172] The above will be described in detail as follows:

[0173] The preliminary weight value of each target object on each exhibit is calculated by multiplying the exhibit participation data of each target object in the group with the comparative analysis result element by element, and the preliminary weight value of each target object is executed in descending order to generate an initial ranking list;

[0174] The initial ranking list of each target object is aggregated into the same group according to the hierarchical judgment result, and the initial ranking list of the target objects of the same identity level is aggregated into the same group, and the preliminary weight value is re-executed in descending order within the group to obtain the hierarchical allocation data of the target objects on each exhibit within the group;

[0175] The hierarchical allocation data of the target objects on each exhibit is weighted and adjusted according to the original attribution degree, and the weighted and adjusted hierarchical allocation data is re-grouped and processed in descending order within the identity level to generate the final overall scheduling result of each target object.

[0176] The scheme can accurately reflect the resource demand difference of different group levels through the hierarchical allocation and dynamic re-allocation mechanism, and realize real-time optimization of the allocation strategy in combination with the original attribution degree of the target object. Through the above technical scheme, the present application solves the problem of insufficient display strategy adaptability caused by ignoring the behavior difference between group levels in the prior art, and adjusts the allocation weight dynamically to make the scheduling result of the exhibit resource more suitable for the actual attribution state of the target object in the group, thereby improving the matching accuracy of the hierarchical display scheme and the behavior characteristics of the target object.

[0177] The above is to schedule the exhibit participation data of the group to which the corresponding target object belongs according to the comparative analysis result, and the following describes matching the overall scheduling result with a preset display strategy library to generate a hierarchical display scheme, which specifically includes:

[0178] The overall scheduling result of each target object is matched with the preset display strategy library in terms of similarity, and the first T display strategy templates are output; wherein the preset display strategy library refers to a database containing various display strategy templates of exhibits and their success rates;

[0179] It is judged whether the success rate is greater than a preset evaluation threshold, and if so, the display strategy template with the largest success rate is selected from all display strategy templates meeting the success rate requirement and is fine-tuned to generate a hierarchical display scheme.

[0180] The similarity matching refers to a process of comparing the overall scheduling result of each target object with each display strategy template in the preset display strategy library and calculating the similarity between them.

[0181] The display strategy template refers to a standardized data structure of an executable display scheme for a specific exhibit or exhibition scene.

[0182] The preset evaluation threshold refers to a numerical limit for judging whether the display strategy template has a sufficient success rate, which is obtained by calculating the average, quantile or weighted average of the success rates of the strategy templates in the preset display strategy library.

[0183] The above content will be described in detail as follows:

[0184] The overall scheduling result of each target object is matched with the display strategy templates in the preset display strategy library one by one, and the first T display strategy templates are selected for each target object, where T is a positive integer between 5 and 10.

[0185] The preset display strategy library refers to a database containing display strategy templates of various exhibits and their success rates.

[0186] If the success rate is greater than the preset evaluation threshold, the display strategy template with the largest success rate is selected from all display strategy templates meeting the success rate requirement. The preset evaluation threshold is independently calculated for the success rate of the display strategy template output for different target objects. Specifically, the success rates of the first T display strategy templates corresponding to the target object are counted and the average and variance are calculated, and the average and variance are calculated by a nonlinear calculation, and the specific calculation formula is as follows:

[0187]

[0188] In the formula, the evaluation threshold is represented by , the nonlinear gain parameter is represented by , and the nonlinear suppression parameter is represented by All the above data are normalized when calculating.

[0189] A fine-tuning model is constructed. The fine-tuning model includes an input layer, an embedding and encoding layer, a fusion layer, a fine-tuning layer and an output layer.

[0190] The input layer receives the display strategy template with the highest success rate, the behavior-identity composite signature, the behavior-identity deviation data, and the stability score, converts the text of the display strategy template with the highest success rate into a token sequence and maps it into an initial token embedding, and splices the behavior-identity composite signature and the behavior-identity deviation data into a behavior-identity vector input aligned with the token sequence;

[0191] The embedding and coding layer performs position coding on the initial token embedding and the behavior-identity vector input output by the input layer and context coding through a plurality of self-attention and feed-forward sub-layers (or lightweight coding blocks), and outputs a behavior-identity template feature representation;

[0192] The fusion layer performs weighted fusion processing on the behavior-identity template feature representation by taking the stability score as a weighting coefficient, and outputs a behavior-identity fusion feature;

[0193] The fine-tuning layer performs multi-layer linear mapping and nonlinear activation transformation (such as ReLU or GELU) on the behavior-identity fusion feature to generate an intermediate feature representation, and performs residual connection or weighted superposition on the intermediate feature representation and the behavior-identity composite signature and the behavior-identity deviation data to generate a fine-tuned feature representation;

[0194] The output layer maps the fine-tuned feature representation to a predefined hierarchical category vector space, each vector corresponding to a hierarchical display level; performs softmax normalization or similar probability distribution operation on the mapped hierarchical category vectors to obtain the hierarchical probability distribution of each hierarchical category vector, and determines the final mapped hierarchical category vector of each hierarchical display level for each target object according to the probability maximum value principle, and generates a hierarchical display scheme through structured combination;

[0195] The display strategy template with the highest success rate, the behavior-identity composite signature, and the behavior-identity deviation data are input as input features, and the stability score is input as a weighted loss basis into a pre-built fine-tuning model to generate a hierarchical display scheme.

[0196] The scheme generates a display scheme that not only conforms to historical successful experience, but also dynamically adapts to the interactive characteristics of the current audience group through similarity matching and success rate double screening mechanism combined with real-time behavior data. Through the above technical scheme, the present application can effectively solve the problem of insufficient adaptability of the hierarchical display scheme in the prior art, and generate a hierarchical display scheme that better fits the behavior characteristics of the target object group through dynamic matching and optimization mechanism, thereby improving the flexibility of exhibition operation and the individualization level of audience experience.

[0197] The above is to match the overall scheduling result with the preset display strategy library to generate a hierarchical display scheme. Next, the display strategy template with the highest success rate is selected and fine-tuned to generate a hierarchical display scheme, which specifically includes:

[0198] The display strategy template with the highest success rate, the behavior-identity composite signature, and the behavior-identity deviation data are input as input features, and the stability score is input as a weighted loss basis into a pre-built fine-tuning model to generate a hierarchical display scheme.

[0199] The pre-built fine-tuning model refers to a calculation model for optimizing the selected display strategy template.

[0200] The above content has been described in detail in the above content, and will not be described in detail here.

[0201] The above technical scheme introduces the behavior-identity composite signature and the behavior-identity deviation data, and combines the stability score to constrain the fine-tuning training process, so that the generated display scheme can dynamically adapt to the behavior characteristic changes of different groups, and through the data-driven optimization of the fine-tuning model, the matching accuracy of the display strategy and the real-time audience behavior is significantly improved. Based on the group behavior stability evaluation result, the present application can dynamically optimize the basic display strategy to generate a hierarchical display scheme that takes into account historical successful experience and current audience behavior characteristics, thereby solving the problem of insufficient strategy adaptability caused by ignoring the behavior differences within the group in the prior art, and realizing precise matching of the exhibit display strategy and the real-time interactive state of the audience.

[0202] Embodiment Two:

[0203] Please refer to Figure 5 , an intelligent operation and audience experience optimization system for museums and art exhibitions, comprising:

[0204] A data acquisition module is configured to acquire interactive behavior data and identity data of each target object.

[0205] A data processing module is configured to project the interactive behavior data and the identity data to generate group structure data, and to perform hierarchical attribution division on the group structure data from the identity dimension to generate exhibit participation data of each group.

[0206] A behavior deviation analysis module is configured to perform scheduling analysis on the correlation degree of target objects included in the corresponding group according to the exhibit participation data to generate behavior-identity deviation data.

[0207] a behavior influence analysis module, configured to input the behavior-identity deviation data and the interaction behavior data of different target objects into a pre-constructed behavior response model, output behavior influence results of the corresponding target objects, and perform comparative analysis on the behavior influence results and the interaction behavior data;

[0208] a hierarchical display scheme generation module, configured to perform overall scheduling on exhibit participation data of a group to which the corresponding target object belongs according to the comparative analysis result, match the overall scheduling result with a preset display strategy library, and generate a hierarchical display scheme.

[0209] The embodiment has the same technical effects as those of the first embodiment.

[0210] It should be noted that, in this document, the terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or sequence between the entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. The data mentioned in the present application is normalized and dimensionally unified before performing xx calculation.

[0211] Although the embodiments of the present application have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent operation and visitor experience optimization in museums and art exhibitions, characterized in that: Includes the following steps: Acquire interaction behavior data and identity data of each target object; The interactive behavior data and the identity data are projected into a group to generate group structure data. The group structure data is then hierarchically classified according to the identity dimension to generate exhibit participation data for each group. Based on the exhibit participation data, the correlation between the target objects within the corresponding group is analyzed to generate behavior-identity deviation data; wherein, the correlation is obtained by matching and analyzing the interaction behavior data and the identity data. The behavior-identity bias data and the interaction behavior data of different target objects are input into a pre-built behavior response model, the behavior impact results of the corresponding target objects are output, and the behavior impact results are compared and analyzed with the interaction behavior data. Based on the comparative analysis results, the exhibit participation data of the corresponding target group are coordinated and scheduled, and the coordination and scheduling results are matched with the preset display strategy library to generate a hierarchical display plan; The interaction behavior data is mapped as a whole based on the identity data, and the overall mapping result is offset corrected based on the group structure data. Perform iterative recoding on the overall mapping result after offset correction until the iteration converges, and generate a behavior-identity composite signature; In each iteration of recoding, the change in the overall mapping result after offset correction is calculated compared with the overall mapping result after offset correction generated in the previous iteration of recoding; Determine whether the change amplitude is greater than a preset convergence threshold; otherwise, it is determined to be iterative convergence. Normalize the change amplitude generated by all iterative recoding processes to generate a stability score, and use the overall mapping result after offset correction as the behavior-identity composite signature. The overall scheduling results of each target object are matched with the preset display strategy library for similarity, and the top T display strategy templates are output; whereby the preset display strategy library refers to a database containing display strategy templates for various exhibits and their success rates. If the success rate is greater than the preset evaluation threshold, then select the display strategy template with the highest success rate from all display strategy templates that meet the success rate requirements and fine-tune it to generate a tiered display scheme. The display strategy template with the highest success rate, the behavior-identity composite signature, and the behavior-identity deviation data are used as input features, and the stability score is used as the weighted loss basis. All of these are input into a pre-built fine-tuning model to generate a hierarchical display scheme.

2. The intelligent operation and visitor experience optimization method for museums and art exhibitions according to claim 1, characterized in that: The group structure data is hierarchically classified and categorized according to identity, and the exhibit participation data for each group is generated specifically including: The group structure data is hierarchically classified according to the identity dimension based on the classification judgment results, and the exhibit participation data of each group is generated. The classification determination result is obtained by matching the behavior-identity composite signature with a predefined set of group levels to obtain the original affiliation degree of each target object relative to each group level, weighting the highest original affiliation degree according to the stability score, and performing classification determination on the weighted result according to a preset affiliation threshold range.

3. The intelligent operation and visitor experience optimization method for museums and art exhibitions according to claim 1, characterized in that: Based on the exhibit participation data, the correlation between target objects within the corresponding group is analyzed to generate behavioral-identity bias data, specifically including: The distribution weights and the correlation degree are weighted and matched according to the target object to generate weighted correlation data; wherein, the distribution weights are obtained by calculating the relative distribution of the exhibit participation data; Based on the weighted correlation data and the exhibit participation data, scheduling data for each target object is generated by sorting, and the scheduling data is aligned and matched with the interactive behavior data to generate expected behavior data; The expected behavior data is compared with the correlation degree of the corresponding target object to generate behavior-identity deviation data.

4. The intelligent operation and visitor experience optimization method for museums and art exhibitions according to claim 3, characterized in that: The behavior-identity bias data and the interaction behavior data of different target objects are input into a pre-built behavior response model, and the output of the behavior impact results of the corresponding target objects specifically includes: The behavior-identity bias data and the interaction behavior data of different target objects are input into the pre-built behavior response model one by one, and the initial value of the behavior influence represented by the behavior of other target objects is output. The initial value of the behavioral impact is calculated by weighting the distribution weights for each group, and the behavioral impact result of each target object is generated.

5. The intelligent operation and visitor experience optimization method for museums and art exhibitions according to claim 2, characterized in that: Based on the comparative analysis results, the participation data of exhibits belonging to the corresponding target groups will be coordinated and scheduled in a coordinated manner, specifically including: Based on the comparative analysis results, the exhibit participation data of the corresponding target object group are weighted and sorted to generate initial allocation data. The initial allocation data is then classified and segmented within the group level to form hierarchical allocation data. The hierarchical allocation data is dynamically redistributed according to the original affiliation degree of the target object in the group hierarchy to generate the overall scheduling result of the target object.

6. A smart operation and visitor experience optimization system for museums and art exhibitions, characterized in that: include: The data acquisition module is used to acquire interaction behavior data and identity data of each target object; The data processing module is used to perform group projection on the interactive behavior data and the identity data to generate group structure data, and to classify the group structure data into hierarchical categories from the identity dimension to generate exhibit participation data for each group. The behavior deviation analysis module is used to schedule and analyze the correlation of target objects within the corresponding group based on the exhibit participation data, and generate behavior-identity deviation data. The behavior impact analysis module is used to input the behavior-identity deviation data and the interaction behavior data of different target objects into a pre-built behavior response model, output the behavior impact results of the corresponding target objects, and compare and analyze the behavior impact results with the interaction behavior data. The tiered display scheme generation module is used to coordinate and schedule the exhibit participation data of the corresponding target group based on the comparative analysis results, and match the coordination and scheduling results with the preset display strategy library to generate a tiered display scheme. The interaction behavior data is mapped as a whole based on the identity data, and the overall mapping result is offset corrected based on the group structure data. Perform iterative recoding on the overall mapping result after offset correction until the iteration converges, and generate a behavior-identity composite signature; In each iteration of recoding, the change in the overall mapping result after offset correction is calculated compared with the overall mapping result after offset correction generated in the previous iteration of recoding; Determine whether the change amplitude is greater than a preset convergence threshold; otherwise, it is determined to be iterative convergence. Normalize the change amplitude generated by all iterative recoding processes to generate a stability score, and use the overall mapping result after offset correction as the behavior-identity composite signature. The overall scheduling results of each target object are matched with the preset display strategy library for similarity, and the top T display strategy templates are output; whereby the preset display strategy library refers to a database containing display strategy templates for various exhibits and their success rates. If the success rate is greater than the preset evaluation threshold, then select the display strategy template with the highest success rate from all display strategy templates that meet the success rate requirements and fine-tune it to generate a tiered display scheme. The display strategy template with the highest success rate, the behavior-identity composite signature, and the behavior-identity deviation data are used as input features, and the stability score is used as the weighted loss basis. All of these are input into a pre-built fine-tuning model to generate a hierarchical display scheme.

Images