Exhibition item popularity analysis method

Abstract

The invention provides an exhibition exhibit popularity analysis method. The method comprises the following steps: step 1, collecting related data of an exhibition exhibit; step 2, constructing an exhibit popularity model for calculating the popularity of the exhibit; step 3, training and optimizing the exhibit popularity model; 4, constructing a popularity prediction model for predicting the popularity of the exhibit; step 5, training the popularity prediction model; and step 6, using the trained popularity prediction model to predict the popularity of the exhibit.

Description

Technical Field

[0001] The invention relates to a thermal analysis method, in particular to a thermal analysis method for exhibition exhibits. Background Art

[0002] This section merely provides background information related to the present disclosure and is not necessarily prior art.

[0003] As the popularity of cultural tourism continues to rise, the number of visitors to public cultural venues such as museums, art galleries, and art galleries has grown explosively. As exhibition organizers or managers, it is necessary to collect and analyze multi-dimensional data of exhibition halls to help exhibition curators better understand the characteristics of the audience group, visitor behavior habits, the popularity of exhibits, and the overall effect of the exhibition, so as to optimize the exhibition content, improve the audience experience, and adjust the marketing strategy in a targeted manner. Multi-dimensional data includes but is not limited to the number of visitors, the distribution of visitor traffic, the length of visitor visits, the length of stay of visitors, the frequency and method of audience interaction in the exhibition, etc.

[0004] Currently, existing technologies mainly collect and extract overall audience data through gate machines, ticketing systems, etc., and lack the means to collect and analyze fine-grained data on exhibition halls and exhibits.

[0005] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0006] Purpose of the invention: The technical problem to be solved by the present invention is to provide a method for analyzing the heat of exhibits in view of the deficiencies in the prior art.

[0007] In order to solve the above technical problems, the present invention discloses a method for analyzing the heat of exhibits, comprising the following steps:

[0008] Step 1, collecting relevant data of the exhibition exhibits;

[0009] Step 2, constructing an exhibit heat model for calculating the heat of the exhibit;

[0010] Step 3: training and optimizing the exhibit heat model;

[0011] Step 4: construct a popularity prediction model to predict the popularity of exhibits;

[0012] Step 5, training the heat prediction model;

[0013] Step 6: Use the trained popularity prediction model to predict the popularity of the exhibits.

[0014] Furthermore, the collection of relevant data of the exhibits in step 1 includes:

[0015] Collect relevant data through virtual exhibition halls and physical exhibition halls, including exhibit attribute data, on-site audience behavior data, and online audience behavior data;

[0016] The exhibit attribute data includes the exhibit name, the exhibit author, the exhibit year, the exhibit category and the exhibit background;

[0017] The on-site audience behavior data includes audience stay time and exhibit interaction data;

[0018] The online audience behavior data includes page views, number of clicks on exhibits and duration of stay on the scene.

[0019] Furthermore, the construction of the exhibit heat model described in step 2 includes:

[0020] Step 2-1, construct the exhibit feature vector x i , used to characterize the characteristics of the i-th exhibit;

[0021] Step 2-2: Construct a comprehensive heat model H for a single exhibit i , used to calculate the popularity of the i-th exhibit.

[0022] Furthermore, the exhibit feature vector x is constructed as described in step 2-1 i ,include:

[0023] x i =[V i , T i , I i , S i ]

[0024] Among them, V i represents the number of visitors to the i-th exhibit, T i represents the average stay time of the i-th exhibit, I i represents the number of interactions with the i-th exhibit, S i Represents the number of social media topics for the i-th exhibit.

[0025] Furthermore, the comprehensive heat model H of a single exhibit is constructed as described in step 2-2. i ,include:

[0026] H i =ω1*V i +ω2*T i +ω3*I i +ω4*S i

[0027] Among them, ω1, ω2, ω3 and ω4 are corresponding weights.

[0028] Furthermore, the training and optimization of the exhibit popularity model in step 3 includes:

[0029] Step 3-1, randomly initialize the weight vector ω = [ω1, ω2, ω3, ω4];

[0030] Step 3-2, collect m exhibit feature vectors and determine the heat value. Let the heat value of the determined i-th exhibit be

[0031] Step 3-3, construct the mean square error function MSE as follows:

[0032]

[0033] in, is the heat value of the i-th exhibit calculated using the current weight vector;

[0034] Step 3-4, for the weight ω in the weight vector ω j Make updates;

[0035] Step 3-5, for each weight in the weight vector ω, the method of step 3-4 is used to iterate cyclically to form the final weight vector ω, thereby completing the training and optimization of the exhibit heat model.

[0036] Further, the update described in steps 3-4 is as follows:

[0037]

[0038] in, is the jth weight after update, is the jth weight before update, α is the learning rate;

[0039] Furthermore, the construction of the heat prediction model described in step 4 includes:

[0040] Step 4-1, construct the exhibit attribute vector, as follows:

[0041] y i =[C i1 , C i2 , …, C ik ]

[0042] Among them, y i represents the attribute vector of the i-th exhibit, C ik represents the kth attribute of the i-th exhibit;

[0043] Step 4-2, construct a heat prediction model, that is, establish a multivariate linear regression model between the exhibit attribute vector and the heat value, as follows:

[0044] H i =β0+β1*C i1 +β2*C i2 +…+β n *C in +∈

[0045] Among them, H i represents the popularity of the i-th exhibit, β0 is the intercept term, β1, β2, …, β n is the regression coefficient, n is the number of exhibit attributes, ∈ is the error term, C in Represents the nth attribute of the i-th exhibit.

[0046] Furthermore, the training of the heat prediction model in step 5 includes:

[0047] Collect the heat values ​​and attributes of exhibits in historical exhibitions as training sets, train the heat prediction model, and obtain the intercept term β0, regression coefficients β1, β2, …, β n and error term ∈.

[0048] Furthermore, the heat value determined in step 3-2 is scored by experts.

[0049] Beneficial effects:

[0050] 1. The present invention proposes a quantifiable exhibition exhibit popularity evaluation method, which can be used to accurately evaluate the exhibition exhibit popularity.

[0051] 2. The present invention newly proposes a correlation model between exhibit attributes and exhibit popularity, and implements a method for predicting exhibit popularity through exhibit attributes. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments, and the above and / or other advantages of the present invention will become more clear.

[0053] Figure 1 It is a schematic diagram of the overall process of the present invention. DETAILED DESCRIPTION

[0054] The overall idea of ​​the present invention is as follows: obtain raw data through online virtual exhibition halls and offline physical exhibition halls, establish an exhibit popularity model, integrate and analyze the exhibit popularity, and provide data support for visitors to visit exhibitions and subsequent exhibition planning. Based on the above ideas, a method for analyzing the popularity of exhibition exhibits is proposed, such as Figure 1 As shown, the specific technical solution is as follows:

[0055] (I) The data collected include:

[0056] (1) Basic data of exhibits (different attributes of exhibits): exhibit name, author, year, category, exhibit background, etc.

[0057] (2) On-site audience behavior data: duration of stay and interaction with exhibits.

[0058] (3) Online (panoramic system, mini program) audience data: number of views, number of clicks on exhibits, and time spent on the scene.

[0059] 2. Establishing a collection popularity model

[0060] (1) Construct the exhibit feature vector x i , x i =[V i , T i , I i , S i ],

[0061] Construct the exhibit attribute vector y i ,y i =[C i1 , C i2 , C in ],

[0062] in:

[0063] V: number of visitors;

[0064] T: average residence time;

[0065] I: number of interactions;

[0066] S: number of social media topics;

[0067] C1 C2 C n : Different attributes of exhibits (such as grade, age, author, etc.)

[0068] (2) Constructing a comprehensive heat model for a single exhibit

[0069] H i =ω1*V i +ω2*T i +ω3*I i +ω4*S i

[0070] ω1~ω4 are corresponding weights.

[0071] (3) Weight coefficient optimization

[0072] Since the ω1~ω4 coefficients cannot be obtained directly, we can use experts to score the popularity of exhibits in completed exhibitions to form a basic data set, which can be used to iteratively optimize the ω1~ω4 coefficients.

[0073] ① Collect m exhibit feature vectors and determine the heat value through expert scoring and other methods, which will be used as samples for parameter optimization;

[0074] ②Construct the mean square error function, Where m is the number of samples, is the heat value calculated using the current weight coefficient, The heat value is determined based on the expert scoring.

[0075] ③ Randomly initialize parameter vector ω=[ω1,ω2...,ω n ], where n is the number of parameters.

[0076] ④For each parameter ω in the ω vector x The update formula is:

[0077]

[0078] Where α is the learning rate, which determines the step size of each update.

[0079] In this step, for each parameter ω x This process is repeated until all parameters are updated. After multiple iterations, the final ω1~ω4 are formed to solidify the thermal model.

[0080] (4) Establish the relationship between the exhibit attribute vector and the heat model

[0081] ① Since some exhibit attributes are text-based, the corresponding attributes can be digitized through one-hot encoding to form a digital exhibit attribute vector y i

[0082] ② Establish a multivariate linear regression model between attribute vector and heat,

[0083] H i =β0+β1*C i1 +β2*C i2 +β3*C i3 +β n *C in +∈

[0084] Where β0 is the intercept term, β1-β n is the regression coefficient, ∈ is the error term

[0085] ③ Based on the heat values ​​generated by historical layout as the training set, use the least squares method or other methods to optimize and determine β1-β n ,∈ and other parameters

[0086] (III) Heat prediction:

[0087] ① Numerical processing of exhibit attributes to form numerical vectors

[0088] ② Based on the correlation between the above exhibit attributes and the heat model, input new exhibit attributes to generate the corresponding heat of the exhibit.

[0089] (IV) Collection methods include:

[0090] (1) Exhibit attributes (C): Extract relevant field data through the data interface of the collection management system (or exhibition management system, etc.).

[0091] (2) Number of visitors (V): Based on the visual visitor flow camera statistics module, the number of visitors to the exhibits is counted, including the number of visitors to the on-site exhibitions; the number of visitors to the VR panoramic exhibition hall exhibits. The sum of the number of online and offline visitors.

[0092] (3) Average dwell time (T): Offline, the average dwell time is calculated based on the passenger flow camera behavior analysis module. Online, the dwell time is calculated based on the VR panoramic exhibition hall. The online and offline dwell time are averaged.

[0093] (4) Interaction times (I): The number of times the offline exhibit’s smart interactive QR code is scanned and opened. Online, the VR panoramic exhibition hall data statistics module is used to obtain the user’s click interaction data for this exhibit. The sum of online and offline interaction data.

[0094] (5) Social media mentions (S): The topic status of each exhibit is obtained through third-party data analysis tools.

[0095] Example:

[0096] The present invention illustrates the above technical solution through an actual exhibit heat prediction process. As shown in Table 1, it is an actual test data table:

[0097] Table 1 Test data table

[0098]

[0099] Optimizing the parameterization through the gradient descent algorithm yields:

[0100] ω1=1.6685712888310158

[0101] ω2=0.7425804926852648

[0102] ω3=0.928508747295292

[0103] ω4=0.09413980998567825

[0104] The exhibit attributes are brought into the multiple linear regression model between the attribute vector and the heat, and the regression coefficients are 0.00096154, -0.27403846, and the intercept is 8.271153846153851

[0105] Suppose we predict an exhibit: if the year is 1200 and the exhibit level is 2, the heat obtained by substituting the attributes and heat model is 8.87; if the year is 1200 and the exhibit level is 1, the heat obtained by substituting the attributes and heat model is 9.15.

[0106] In a specific implementation, the present application provides a computer storage medium and a corresponding data processing unit, wherein the computer storage medium can store a computer program, and when the computer program is executed by the data processing unit, the invention content of the exhibition exhibit heat analysis method provided by the present invention and some or all steps in each embodiment can be executed. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0107] Those skilled in the art can clearly understand that the technical solutions in the embodiments of the present invention can be implemented by means of computer programs and their corresponding general hardware platforms. Based on such an understanding, the technical solutions in the embodiments of the present invention can be essentially or partly contributed to the prior art in the form of computer programs, i.e., software products, which can be stored in a storage medium and include several instructions for enabling a device including a data processing unit (which can be a personal computer, a server, a single-chip microcomputer, an MCU or a network device, etc.) to execute the methods described in various embodiments of the present invention or certain parts of the embodiments.

[0108] The present invention provides a method and idea for analyzing the heat of exhibits. There are many methods and approaches to implement the technical solution. The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the protection scope of the present invention. All components not specified in this embodiment can be implemented by existing technologies.

Claims

1. A method for analyzing the heat of an exhibition exhibit, characterized in that: The following steps are involved: Step 1, collecting relevant data of the exhibition exhibits; Step 2, constructing an exhibit heat model for calculating the heat of the exhibit; Step 3: training and optimizing the exhibit heat model; Step 4: construct a popularity prediction model to predict the popularity of exhibits; Step 5, training the heat prediction model; Step 6: Use the trained popularity prediction model to predict the popularity of the exhibits.

2. The method for analyzing the heat of an exhibition item according to claim 1, characterized in that: The data collected in step 1 about the exhibits include: Collect relevant data through virtual exhibition halls and physical exhibition halls, including exhibit attribute data, on-site audience behavior data, and online audience behavior data; The exhibit attribute data includes the exhibit name, the exhibit author, the exhibit year, the exhibit category and the exhibit background; The on-site audience behavior data includes audience stay time and exhibit interaction data; The online audience behavior data includes page views, number of clicks on exhibits and duration of stay on the scene.

3. The method for analyzing the heat of an exhibition item according to claim 2, characterized in that: The construction of the exhibit heat model described in step 2 includes: Step 2-1, construct the exhibit feature vector x i , used to characterize the characteristics of the i-th exhibit; Step 2-2: Construct a comprehensive heat model H for a single exhibit i , used to calculate the popularity of the i-th exhibit.

4. The method for analyzing the thermal properties of exhibits according to claim 3, characterized in that: Construct the exhibit feature vector x as described in step 2-1 i ,include: x i =[V i ,T i ,I i ,S i ] Among them, V i represents the number of visitors to the i-th exhibit, T i represents the average stay time of the i-th exhibit, I i represents the number of interactions with the i-th exhibit, S i Represents the number of social media topics for the i-th exhibit.

5. The method for analyzing the thermal properties of exhibits according to claim 4, characterized in that: Construct a comprehensive thermal model H of a single exhibit as described in step 2-2 i ,include: H i =ω1*V i +ω2*T i +ω3*I i +ω4*S i Among them, ω1, ω2, ω3 and ω4 are corresponding weights.

6. The method for analyzing the thermal properties of exhibits according to claim 5, characterized in that: The training and optimization of the exhibit heat model described in step 3 includes: Step 3-1, randomly initialize the weight vector ω = [ω1, ω2, ω3, ω4]; Step 3-2, collect m exhibit feature vectors and determine the heat value. Let the heat value of the determined i-th exhibit be Step 3-3, construct the mean square error function MSE as follows: in, is the heat value of the i-th exhibit calculated using the current weight vector; Step 3-4, for the weight ω in the weight vector ω j Make updates; Step 3-5, for each weight in the weight vector ω, the method of step 3-4 is used to iterate cyclically to form the final weight vector ω, thereby completing the training and optimization of the exhibit heat model.

7. The method for analyzing the thermal properties of exhibits according to claim 6, characterized in that: Update as described in steps 3-4, as follows: in, is the jth weight after update, is the jth weight before update, and α is the learning rate.

8. The method for analyzing the thermal properties of exhibits according to claim 7, characterized in that: The heat prediction model constructed in step 4 includes: Step 4-1, construct the exhibit attribute vector, as follows: y i =[C i1 ,C i2 ,…,C ik ] Among them, y i represents the attribute vector of the i-th exhibit, C ik represents the kth attribute of the i-th exhibit; Step 4-2, construct a heat prediction model, that is, establish a multivariate linear regression model between the exhibit attribute vector and the heat value, as follows: H i =β0+β1*C i1 +β2*C i2 +…+c n *C in +∈ Among them, H i represents the popularity of the i-th exhibit, β0 is the intercept term, β1, β2, …, β n is the regression coefficient, n is the number of exhibit attributes, ∈ is the error term, C in Represents the bth attribute of the i-th exhibit.

9. The method for analyzing the thermal properties of exhibits according to claim 8, characterized in that: The training of the heat prediction model described in step 5 includes: Collect the heat values ​​and attributes of exhibits in historical exhibitions as training sets, train the heat prediction model, and obtain the intercept term β0, regression coefficients β1, β2, …, β n and error term ∈.

10. The method for analyzing the thermal properties of exhibits according to claim 9, characterized in that: The heat value determined in step 3-2 is scored by experts.

Images