Digital exhibition hall recommendation system and method based on artificial intelligence

Abstract

The invention discloses a digital exhibition hall recommendation system and method based on artificial intelligence. The system comprises an acquisition module, a processing module, an extraction module, a construction module and a recommendation module. The acquisition module is used for acquiring behavior data of visitors; the processing module is used for preprocessing the behavior data to obtain processed data; the extraction module is used for performing feature extraction on the processed data to obtain key features of behaviors of visitors; the construction module is used for constructing a recommendation model based on the key features; and the recommendation module is used for performing digital exhibition hall recommendation based on the recommendation model. According to the invention, challenges confronted when a traditional exhibition hall provides personalized guide recommendation are effectively solved. According to the method, accurate analysis and classification of behaviors of visitors are realized, and real-time feedback of the user is responded by dynamically adjusting the recommendation strategy, so that the experience of the visitors is greatly improved, and the digital exhibition hall can better meet the requirements and preferences of different visitors.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and in particular to a digital exhibition hall recommendation system and method based on artificial intelligence Background Art

[0002] In digital exhibition halls, how to use artificial intelligence technology to provide visitors with personalized guided tour recommendations is a challenging problem. First, it is necessary to collect and analyze various behavioral data of visitors in the exhibition hall, such as movement trajectory, dwell time, interactive behavior, etc. These data are usually unstructured, high-dimensional, and change in real time. Secondly, the interests and preferences of different visitors vary greatly. How to mine implicit personal preference characteristics from massive data and conduct user profiling is the key. In addition, the exhibition hall environment is complex and changeable, and there are a large number of interference factors, such as traffic flow, light changes, etc. How to achieve stable and reliable user behavior perception and recognition in a noisy environment is also a major difficulty. Finally, the recommendation system needs to balance between accuracy and real-time performance. It is necessary to give accurate recommendations based on the user's historical behavior and current status, and to quickly respond to the user's real-time feedback and dynamically adjust the recommendation strategy. Summary of the invention

[0003] In order to solve the technical problems in the above background, the present invention provides the following technical solutions:

[0004] A digital exhibition hall recommendation system based on artificial intelligence, comprising: a collection module, a processing module, an extraction module, a construction module and a recommendation module;

[0005] The collection module is used to collect the behavior data of visitors;

[0006] The processing module is used to pre-process the behavior data to obtain processed data;

[0007] The extraction module is used to extract features from the processed data to obtain key features of visitor behavior;

[0008] The building module is used to build a recommendation model based on the key features;

[0009] The recommendation module is used to make digital exhibition hall recommendations based on the recommendation model.

[0010] Preferably, the processing module uses a Gaussian filtering algorithm to remove noise data from the behavior data:

[0011]

[0012] Among them, G(x, y) represents the value of a point (x, y) in two-dimensional space; x and y represent the coordinates in two-dimensional space; σ represents the standard deviation, which controls the width of the kernel.

[0013] Preferably, after removing the noise, the processing module converts the data into a standard format in the interval [0,1] using a Min-Max normalization method:

[0014]

[0015] In the formula, min(z) represents the minimum value; max(z) represents the maximum value; z norm Indicates the normalized value.

[0016] Preferably, the extraction module is used to perform feature-level fusion on the processed data to extract key features of visitor behavior.

[0017] Preferably, the construction module uses a K-means clustering algorithm to cluster the key features to construct the recommendation model.

[0018] The present invention also provides a method for recommending digital exhibition halls based on artificial intelligence, which is applied to the above system and comprises the following steps:

[0019] Collect visitor behavior data;

[0020] Preprocessing the behavior data to obtain processed data;

[0021] Performing feature extraction on the processed data to obtain key features of visitor behavior;

[0022] Based on the key features, a recommendation model is constructed;

[0023] Based on the recommendation model, digital exhibition hall recommendations are made.

[0024] Preferably, the method for performing the preprocessing includes: using a Gaussian filtering algorithm to remove noise data in the behavior data:

[0025]

[0026] Among them, G(x, y) represents the value of a point (x, y) in two-dimensional space; x and y represent the coordinates in two-dimensional space; σ represents the standard deviation, which controls the width of the kernel.

[0027] Preferably, after removing the noise, the data is converted to a standard format in the interval [0,1] using the Min-Max normalization method:

[0028]

[0029] In the formula, min(z) represents the minimum value; max(z) represents the maximum value; z morm Indicates the normalized value.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] The present invention effectively solves the challenges faced by traditional exhibition halls in providing personalized guided tour recommendations. The present invention achieves accurate analysis and classification of visitor behavior, and greatly improves the visitor experience by dynamically adjusting the recommendation strategy in response to real-time feedback from users, so that the digital exhibition hall can better meet the needs and preferences of different visitors. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0033] Figure 1 Schematic diagram of the system structure of an embodiment of the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0036] Embodiment 1

[0037] like Figure 1 As shown, it is a schematic diagram of the system structure of this embodiment, including: an acquisition module, a processing module, an extraction module, a construction module and a recommendation module; the acquisition module is used to collect the behavior data of visitors; the processing module is used to pre-process the behavior data to obtain processed data; the extraction module is used to extract features from the processed data; the construction module is used to use the extracted features to build a recommendation model; the recommendation module is used to make digital exhibition hall recommendations based on the recommendation model.

[0038] The following will explain in detail how the present invention solves technical problems in real life in conjunction with this embodiment.

[0039] First, a collection module is deployed inside the exhibition hall. In this embodiment, the collection module specifically includes sensors and cameras to collect visitors' behavior data in real time, including: movement trajectory, residence time, interactive behavior and other data.

[0040] Specifically, according to the layout and needs of the exhibition hall, infrared sensors and high-definition cameras are deployed at key locations such as the entrance, main passages, and exhibit areas of the exhibition hall. The distance between sensors is no more than 5 meters, and the camera pixel is no less than 1080P, ensuring effective coverage of more than 90% of the exhibition hall area. The sensor acquisition frequency is set to once per second, and the movement trajectory of visitors is recorded in real time, including data such as the visiting route, movement speed, and residence time. At the same time, the interactive behavior data such as facial expressions and body movements of visitors are collected through the camera. The average amount of data collected for each visitor is no less than 100MB.

[0041] Afterwards, the behavior data is preprocessed using a processing module, and the process includes: using a Gaussian filtering algorithm to remove noise data and improve the signal-to-noise ratio of the data.

[0042] Specifically, the image is smoothed by using a linear filter with a Gaussian function as the weight coefficient. First, a Gaussian kernel is defined, and each element value of the kernel is calculated by the Gaussian distribution formula, as shown below:

[0043]

[0044] Among them, G(x, y) represents the value of a point (x, y) in two-dimensional space; x and y represent the coordinates in two-dimensional space; σ represents the standard deviation, which controls the width of the kernel.

[0045] Then, the kernel is convolved with the image, that is, at each pixel of the image, the kernel is covered with the pixels around it, and the weighted average is calculated to replace the central pixel value with this value, thereby achieving a smoothing effect. This process is repeated for each pixel of the image until the entire image is processed. Gaussian filtering can effectively remove Gaussian noise in an image while maintaining the clarity of the image edges.

[0046] Then, the Min-Max normalization method is used to convert the data into a standard format in the interval [0,1] to ensure the consistency of data from different sources. The specific process includes: The Min-Max normalization method scales the data features to the range of 0 to 1 by calculating the minimum and maximum values ​​in the data set. For each value z in the data set, first determine the minimum value min(z) and the maximum value max(z), and then apply the following formula for conversion:

[0047]

[0048] In the formula, z normRepresents the normalized value. This process makes all data points standardized to the same scale, which facilitates subsequent data processing and analysis.

[0049] The extraction module then uses the Bayesian network algorithm to perform feature-level fusion on the processed data to extract key features of visitor behavior, such as the time spent browsing exhibits, interaction frequency, emotional changes, etc. Specifically, the Bayesian network algorithm is a computational method based on a probabilistic graphical model that represents the dependencies between variables through a directed acyclic graph (DAG). In this model, each node represents a random variable, and the directed edges between nodes represent the conditional dependencies between variables. The core idea of ​​the Bayesian network is to use known variable values ​​to infer the probability distribution of other variables. The basic formula is as follows:

[0050] P(A|B)=P(A,B) / P(B)

[0051] That is, the probability of A given B is equal to the probability of A and B occurring simultaneously divided by the probability of B occurring. In this way, Bayesian networks can perform probabilistic reasoning and decision analysis and extract key features of visitor behavior.

[0052] After feature extraction is completed, the construction module uses the K-means clustering algorithm to perform cluster analysis on the user behavior feature vector and builds a recommendation model to classify tourists. This embodiment divides tourists into the following categories: art lovers, history explorers, parent-child interaction groups, technology enthusiasts, and passers-by.

[0053] Specifically, cluster analysis is performed on the key features of visitor behavior extracted by the extraction module, and the steps include:

[0054] The K-means clustering algorithm is used to analyze the above extracted features to obtain the clustering results of the parameter characteristics of visitor behavior, and the clustering results are optimized and improved by manual review. The steps include:

[0055] 1. Randomly select K cluster centers to initialize each cluster.

[0056] 2. Traverse all data points and assign each data point to the cluster represented by the cluster center closest to it.

[0057] 3. Recalculate the cluster center of each cluster.

[0058] 4. Repeat steps 2 to 3 until convergence is reached.

[0059] 5. Data analysis results: The final clustering result is K different cluster sets, each of which contains several similar data points. These clusters can be regarded as tourist preference groups.

[0060] Finally, the recommendation module uses the recommendation model to complete the recommendation of the digital exhibition hall.

[0061] Embodiment 2

[0062] This embodiment also provides a method for recommending a digital exhibition hall based on artificial intelligence, the steps comprising:

[0063] S1. Collect visitor behavior data.

[0064] First, sensors and cameras are deployed inside the exhibition hall to collect visitor behavior data in real time, including movement trajectory, residence time, and interactive behavior.

[0065] Specifically, according to the layout and needs of the exhibition hall, infrared sensors and high-definition cameras are deployed at key locations such as the entrance, main passages, and exhibit areas of the exhibition hall. The distance between sensors is no more than 5 meters, and the camera pixel is no less than 1080P, ensuring effective coverage of more than 90% of the exhibition hall area. The sensor acquisition frequency is set to once per second, and the movement trajectory of visitors is recorded in real time, including data such as the visiting route, movement speed, and residence time. At the same time, the interactive behavior data such as facial expressions and body movements of visitors are collected through the camera. The average amount of data collected for each visitor is no less than 100MB.

[0066] S2. Preprocess the behavioral data to obtain processed data.

[0067] Afterwards, the behavioral data is preprocessed, and the process includes: using Gaussian filtering algorithm to remove noise data and improve the signal-to-noise ratio of the data.

[0068] Specifically, the image is smoothed by using a linear filter with a Gaussian function as the weight coefficient. First, a Gaussian kernel is defined, and each element value of the kernel is calculated by the Gaussian distribution formula, as shown below:

[0069]

[0070] Among them, G(x, y) represents the value of a point (x, y) in two-dimensional space; x and y represent the coordinates in two-dimensional space; σ represents the standard deviation, which controls the width of the kernel.

[0071] Then, the kernel is convolved with the image, that is, at each pixel of the image, the kernel is covered with the pixels around it, and the weighted average is calculated to replace the central pixel value with this value, thereby achieving a smoothing effect. This process is repeated for each pixel of the image until the entire image is processed. Gaussian filtering can effectively remove Gaussian noise in an image while maintaining the clarity of the image edges.

[0072] Then, the Min-Max normalization method is used to convert the data into a standard format in the interval [0,1] to ensure the consistency of data from different sources. The specific process includes: The Min-Max normalization method scales the data features to the range of 0 to 1 by calculating the minimum and maximum values ​​in the data set. For each value z in the data set, first determine the minimum value min(z) and the maximum value max(z), and then apply the following formula for conversion:

[0073]

[0074] In the formula, z norm Represents the normalized value. This process makes all data points standardized to the same scale, which facilitates subsequent data processing and analysis.

[0075] S3. Perform feature extraction on the processed data to obtain the key features of visitor behavior.

[0076] The Bayesian network algorithm is used to perform feature-level fusion on the processed data to extract key features of visitor behavior, such as the time spent browsing exhibits, interaction frequency, emotional changes, etc. Specifically, the Bayesian network algorithm is a computational method based on a probabilistic graph model that represents the dependencies between variables through a directed acyclic graph (DAG). In this model, each node represents a random variable, and the directed edges between nodes represent the conditional dependencies between variables. The core idea of ​​the Bayesian network is to use known variable values ​​to infer the probability distribution of other variables. The basic formula is as follows:

[0077] P(A|B)=P(A,B) / P(B)

[0078] That is, the probability of A given B is equal to the probability of A and B occurring simultaneously divided by the probability of B occurring. In this way, Bayesian networks can perform probabilistic reasoning and decision analysis and extract key features of visitor behavior.

[0079] S4. Build a recommendation model based on key features.

[0080] After feature extraction is completed, the K-means clustering algorithm is used to perform cluster analysis on the user behavior feature vectors, and a recommendation model is constructed to classify tourists. In this embodiment, tourists are divided into the following categories: art lovers, history explorers, parent-child interaction groups, technology enthusiasts, and passers-by.

[0081] Specifically, cluster analysis is performed on the key features of visitor behavior extracted by the extraction module, and the steps include:

[0082] The K-means clustering algorithm is used to analyze the above extracted features to obtain the clustering results of the parameter characteristics of visitor behavior, and the clustering results are optimized and improved by manual review. The steps include:

[0083] 1. Randomly select K cluster centers to initialize each cluster.

[0084] 2. Traverse all data points and assign each data point to the cluster represented by the cluster center closest to it.

[0085] 3. Recalculate the cluster center of each cluster.

[0086] 4. Repeat steps 2 to 3 until convergence is reached.

[0087] 5. Data analysis results: The final clustering result is K different cluster sets, each of which contains several similar data points. These clusters can be regarded as tourist preference groups.

[0088] S5. Make digital exhibition hall recommendations based on the recommendation model.

[0089] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A digital exhibition hall recommendation system based on artificial intelligence, characterized in that: include: Acquisition module, processing module, extraction module, construction module and recommendation module; The collection module is used to collect the behavior data of visitors; The processing module is used to pre-process the behavior data to obtain processed data; The extraction module is used to extract features from the processed data to obtain key features of visitor behavior; The building module is used to build a recommendation model based on the key features; The recommendation module is used to make digital exhibition hall recommendations based on the recommendation model.

2. The digital exhibition hall recommendation system based on artificial intelligence according to claim 1 is characterized in that: The processing module uses a Gaussian filtering algorithm to remove noise data from the behavior data: Among them, G(x, y) represents the value of a point (x, y) in two-dimensional space; x and y represent the coordinates in two-dimensional space; σ represents the standard deviation, which controls the width of the kernel.

3. The digital exhibition hall recommendation system based on artificial intelligence according to claim 2 is characterized in that: After removing the noise, the processing module uses the Min-Max normalization method to convert the data into a standard format in the interval [0,1]: In the formula, min(z) represents the minimum value; max(z) represents the maximum value; z norm Indicates the normalized value.

4. The digital exhibition hall recommendation system based on artificial intelligence according to claim 1 is characterized in that: The extraction module is used to perform feature-level fusion on the processed data to extract key features of visitor behavior.

5. The digital exhibition hall recommendation system based on artificial intelligence according to claim 1 is characterized in that: The construction module clusters the key features using a K-means clustering algorithm to construct the recommendation model.

6. A method for recommending digital exhibition halls based on artificial intelligence, the method being applied to the system according to any one of claims 1 to 5, characterized in that the steps include: Collect visitor behavior data; Preprocessing the behavior data to obtain processed data; Performing feature extraction on the processed data to obtain key features of visitor behavior; Based on the key features, a recommendation model is constructed; Based on the recommendation model, digital exhibition hall recommendations are made.

7. The method for recommending digital exhibition halls based on artificial intelligence according to claim 6, characterized in that: The method for performing the preprocessing includes: using a Gaussian filtering algorithm to remove noise data in the behavior data: Among them, G(x, y) represents the value of a point (x, y) in two-dimensional space; x and y represent the coordinates in two-dimensional space; σ represents the standard deviation, which controls the width of the kernel.

8. The method for recommending digital exhibition halls based on artificial intelligence according to claim 7, characterized in that: After removing the noise, the Min-Max normalization method is used to convert the data into a standard format in the [0,1] interval: In the formula, min(z) represents the minimum value; max(z) represents the maximum value; z norm Indicates the normalized value.

Images