Country culture characteristic tourist attraction construction method and system based on big data fusion

Abstract

The invention discloses a rural culture characteristic tourist attraction construction method based on big data fusion, and the method comprises the following steps: constructing a five-dimensional data collection matrix to obtain original data, and generating a data fusion feature vector through the processing of a federated learning framework and the weighting of a dynamic weight distribution algorithm; performing cross-modal feature alignment on the vectors by using a dual-channel feature extraction network, and outputting culture gene codes; mapping the codes to a four-dimensional space-time coordinate system, and generating a knowledge graph with space-time attributes by means of a time sequence diagram neural network; generating a resource scheduling and tourist tourism scheme according to the knowledge graph, and displaying and collecting the tourist intention information optimization scheme through a virtual-real fusion guide device; through multi-dimensional and real-time data acquisition and efficient data analysis, a scientific basis is provided for tourist attraction planning, resource scheduling and market operation, so that making of a more reasonable development strategy is facilitated, and sustainable development of rural tourism is realized.

Description

Technical Field

[0001] The present invention relates to the field of rural cultural tourism planning, and in particular to a method and system for constructing rural cultural tourism attractions based on big data fusion. Background Art

[0002] As people's living standards improve, rural cultural tourism is becoming a popular form of tourism, attracting a growing number of tourists. This type of tourism not only allows visitors to appreciate the unique natural scenery of rural areas, but also allows them to deeply experience the rich history, culture, folk customs, and traditional ways of life. This type of tourism is of great significance in promoting rural economic development and preserving and promoting rural culture.

[0003] However, current rural cultural tourism attractions face numerous challenges in their development and operation. First, the exploration and integration of rural tourism resources are insufficient. Rural areas boast a rich and diverse array of cultural resources, such as ancient architecture, traditional crafts, and unique folk customs. However, these resources are often scattered and lack systematic organization and development, hindering the formation of an attractive tourism product system. For example, while some rural areas have ancient villages with a long history, they fail to attract large numbers of tourists due to a lack of in-depth exploration and effective display of their historical and cultural value. Second, tourism services are unable to precisely meet the individual needs of tourists. Different tourists have varying preferences and needs for travel, with some preferring historical and cultural exploration, while others are more interested in rural cuisine and leisure experiences. However, existing rural tourism attractions struggle to obtain comprehensive and accurate information on tourist needs during their planning and service offerings, resulting in a lack of targeted tourism products and services and a poor visitor experience. Summary of the Invention

[0004] The present invention overcomes the shortcomings of the existing technology and provides a method and system for constructing rural cultural characteristic tourist attractions based on big data fusion.

[0005] To achieve the above objectives, the present invention adopts a technical solution: a method and system for constructing rural cultural tourism attractions based on big data fusion, comprising the following steps: S1. Construct a five-dimensional data acquisition matrix for acquiring raw data, and use the federated learning framework to perform data interaction processing on the raw data. Then, use a dynamic weight allocation algorithm to assign real-time weights to the raw data after data interaction processing to generate a data fusion feature vector. S2. Through a dual-channel feature extraction network, cross-modal feature alignment is performed on the data fusion feature vector and the cultural gene code is output; S3. Map the cultural gene code to a four-dimensional space-time coordinate system, and dynamically infer the relationship between nodes through a temporal graph neural network to generate a knowledge graph with space-time attributes; S4. Generate resource scheduling plans and tourist travel plans based on the knowledge graph, display the output results through the virtual-reality fusion guide device, collect tourists' intention information, reversely update the node weights of the knowledge graph, and optimize the resource scheduling plans and tourist travel plans.

[0006] In a preferred embodiment of the present invention, the five-dimensional data acquisition matrix includes five data layers, specifically: a historical context data layer, a physical object perception layer, a social media public opinion layer, an economic operation data layer, and an ecological environment monitoring layer; the formula expression of the data fusion feature vector is: Where, Represents the weight coefficient based on the information entropy of the i-th layer; is the data feature quantity of the i-th layer; n is the total number of data layers.

[0007] In a preferred embodiment of the present invention, the dynamic weight allocation algorithm evaluates the importance of the original data based on the weight coefficient of information entropy, wherein the calculation formula of the weight coefficient is: Where, represents the information entropy of the zth original data; m represents the total amount of original data.

[0008] In a preferred embodiment of the present invention, the gradient aggregation formula of the federated learning framework is: Where, Indicates the result value after the fusion of multiple layers of data features; represents the data feature quantity of the i-th data layer; q represents the total data feature quantity of the five data layers; is the data feature of the i-th layer.

[0009] In a preferred embodiment of the present invention, the cross-modal feature alignment includes: the cross-modal feature alignment includes: constructing an attention alignment matrix , the correlation based on image-text features is calculated, and its formula is: Where, Represents the element value of the h-th row and l-th column in the attention alignment matrix; Represents the hth dimension value based on the image feature vector; Represents the lth dimension value based on the text feature vector; represents the natural exponential function; represents the sum of all dimension values ​​l of the text feature vector; d is the number of rows and columns of the attention alignment matrix; The joint representation of the data fusion feature vector is: ;in, Represents element-wise product.

[0010] In a preferred embodiment of the present invention, the method for dynamically updating the knowledge graph includes: receiving cultural gene encoding data in real time, extracting spatiotemporal events through a sliding time window mechanism, and updating the weight of the knowledge graph.

[0011] In a preferred embodiment of the present invention, the temporal graph neural network adopts a graph attention mechanism when dynamically reasoning about the association relationship between nodes, and highlights the influence of key nodes on the association relationship reasoning by calculating the attention weight of the nodes.

[0012] In a preferred embodiment of the present invention, the virtual-reality fusion tour guide system utilizes augmented reality and virtual reality technologies to combine virtual rural cultural elements with real tourist scenes to provide tourists with an immersive tour guide experience. During the display process, the display effect of the virtual elements is adjusted in real time by tracking the tourists' position, movements and line of sight. The display content includes but is not limited to the restoration of scenes of historical stories and the demonstration of the production process of traditional handicrafts.

[0013] In a preferred embodiment of the present invention, a rural cultural characteristic tourist attraction construction system based on big data fusion is proposed, which is based on the rural cultural characteristic tourist attraction construction method based on big data fusion, including: a data acquisition module, a data processing module, a knowledge graph construction module, a solution generation module and a virtual-reality fusion navigation module; The data acquisition module is used to construct a five-dimensional data acquisition matrix, collect data from the historical context data layer, the physical IoT perception layer, the social media public opinion layer, the economic operation data layer, and the ecological environment monitoring layer, and transmit the collected raw data to the data processing module; The data processing module uses a federated learning framework to perform data interaction processing on the raw data, adopts a dynamic weight allocation algorithm to perform real-time weighting on the raw data after data interaction processing, generates a data fusion feature vector, and uses a dual-channel feature extraction network to perform cross-modal feature alignment on the data fusion feature vector to output the cultural gene code; The knowledge graph construction module maps the meme code to a four-dimensional space-time coordinate system, dynamically infers the association between nodes through a time-series graph neural network, generates a knowledge graph with space-time attributes, and updates the weight of the knowledge graph through a sliding time window mechanism based on the meme code data received in real time. The solution generation module generates resource scheduling solutions and tourist travel solutions based on the knowledge graph; The virtual-reality fusion navigation module displays the output results of the plan generation module, collects tourists' intention information, and feeds the intention information back to the knowledge graph construction module, reversely updates the node weights of the knowledge graph, and optimizes the resource scheduling plan and tourist travel plan.

[0014] In a preferred embodiment of the present invention, the five data layers in the data acquisition module have different data acquisition frequencies. The historical context data layer is updated regularly, the physical object perception layer collects data in real time, the social media public opinion layer captures data at preset time intervals, the economic operation data layer collects data according to the business cycle, and the ecological environment monitoring layer sets the collection frequency based on the performance of the monitoring equipment and monitoring requirements.

[0015] The present invention solves the defects existing in the background technology and has the following beneficial effects: (1) Construct a five-dimensional data collection matrix, covering the historical context data layer, the physical IoT perception layer, the social media public opinion layer, the economic operation data layer, and the ecological environment monitoring layer, to comprehensively collect rural tourism-related data. Through the federated learning framework, data interaction processing is carried out, and a dynamic weight allocation algorithm is used for real-time weighting to generate data fusion feature vectors. In-depth mining and integration of rural tourism resources provide data support for the construction of characteristic tourist attractions. At the same time, the scattered resources are integrated into an organic whole to form a richer, more diverse and more attractive tourism product system, thereby enhancing the overall competitiveness of rural tourism.

[0016] (2) Based on the generated knowledge graph, the association relationship of rural tourism resources can be clearly presented. Based on this, tourist travel plans are generated. According to the intention information of tourists during the virtual and real integrated tour, the node weights of the knowledge graph are reversely updated, and the travel plans are continuously optimized to achieve a precise match between tourism products and services and tourists' personalized needs.

[0017] (3) Multi-dimensional, real-time data collection and efficient data analysis provide a scientific basis for the planning, resource allocation and market operation of tourist attractions, thereby helping to formulate more reasonable development strategies and achieve sustainable development of rural tourism. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments described in the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive efforts. Figure 1 It is a method flow chart of a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0020] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0021] like Figure 1 As shown, the method and system for constructing rural cultural tourism attractions based on big data fusion include the following steps: S1. Construct a five-dimensional data acquisition matrix for acquiring raw data, and use the federated learning framework to perform data interaction processing on the raw data. Then, use a dynamic weight allocation algorithm to assign real-time weights to the raw data after data interaction processing to generate a data fusion feature vector. Furthermore, the five-dimensional data collection matrix includes five data layers, namely: historical context data layer, physical IoT perception layer, social media public opinion layer, economic operation data layer, and ecological environment monitoring layer; the formula expression of the data fusion feature vector is: Where, Represents the weight coefficient based on the information entropy of the i-th layer; is the data feature quantity of the i-th layer; n is the total number of data layers.

[0022] Specifically, the historical context data layer primarily covers the historical evolution, cultural heritage, and folk customs of rural areas. This includes information on historical buildings in the village, such as their construction dates, architectural styles, and historical allusions; data on intangible cultural heritage, such as the lineage and production processes of traditional handicrafts (such as paper-cutting, embroidery, and pottery); and the origins, development, procedures, and legends of folk activities (such as temple fairs, sacrificial rituals, and special festivals). This data can be obtained through consulting local historical records, interviewing local cultural experts, and studying historical archives. Regular updates ensure its accuracy and completeness, providing a foundation for in-depth exploration of rural cultural heritage.

[0023] The physical IoT perception layer leverages IoT technology to provide real-time perception and data collection of various physical resources involved in rural tourism. For example, sensors installed within tourist attractions monitor the temperature, humidity, and structural safety of ancient buildings, promptly identifying potential damage risks. Ecological indicators of rural natural landscapes (such as rivers and forests) are monitored, including water quality, air quality, and vegetation growth. The use of tourist facilities (such as parking lots, trails, and rest areas) can also be monitored to compile statistics on visitor flow and facility utilization. Real-time data collection provides a basis for the protection and rational use of tourism resources.

[0024] The social media public opinion layer uses web crawler technology to capture information related to rural tourism from major social media platforms (such as WeChat, Weibo, and Douyin) at preset intervals. This includes pictures, text, and videos posted by tourists, as well as reviews, feedback, and suggestions about rural tourism. Natural language processing technology is used to analyze this public opinion data to understand tourists' preferences, satisfaction levels, and pain points regarding rural tourist attractions. For example, if a large number of tourists complain about poor dining service at a rural attraction on social media, targeted improvements can be implemented to enhance the quality of tourism services.

[0025] The economic operations data layer collects economic data on rural tourism based on the business cycle. This data covers direct economic indicators such as ticket revenue, accommodation and catering revenue, and tourism product sales, as well as indirect economic benefits such as local employment and income growth. It also includes investment costs and operating costs (such as labor costs, material procurement costs, and facility maintenance costs) of tourism projects. By analyzing this economic data, we assess the economic feasibility and development potential of rural tourism and provide support for the formulation of appropriate economic operations strategies.

[0026] The ecological and environmental monitoring layer sets data collection frequencies based on the performance of monitoring equipment and monitoring needs, providing comprehensive monitoring of the rural ecological environment. In addition to the common air and water quality monitoring, this layer also includes soil quality monitoring, assessing indicators such as soil fertility and pH, providing data support for rural agricultural tourism and ecological protection. Biodiversity monitoring counts the species, abundance, and distribution of flora and fauna in rural areas, assessing ecosystem stability. This data is crucial for achieving sustainable development of rural tourism.

[0027] Furthermore, the dynamic weight allocation algorithm evaluates the importance of the original data based on the weight coefficient of information entropy, where the calculation formula of the weight coefficient is: Where, represents the information entropy of the zth original data; m represents the total amount of original data.

[0028] Furthermore, the gradient aggregation formula of the federated learning framework is: Where, Indicates the result value after the fusion of multiple layers of data features; represents the data feature quantity of the i-th data layer; q represents the total data feature quantity of the five data layers; is the data feature of the i-th layer.

[0029] S2. Through a dual-channel feature extraction network, cross-modal feature alignment is performed on the data fusion feature vector and the cultural gene code is output; A dual-channel feature extraction network extracts features from the data fusion feature vectors, respectively, from the image and text channels. In the image channel, a convolutional neural network architecture employs a combination of convolutional and pooling layers to extract visual features from the image, such as the architectural appearance and natural landscape colors of rural tourist attractions. In the text channel, word embedding models from natural language processing techniques are used to convert text data into vector representations, extracting semantic features from the text, such as the sentiment and thematic information contained in tourists' descriptive words about the attractions. These two channels operate in parallel, each extracting features from a different modality, providing a foundation for subsequent cross-modal feature alignment.

[0030] Furthermore, cross-modal feature alignment includes: constructing an attention alignment matrix , the correlation based on image-text features is calculated, and its formula is: Where, Represents the element value of the h-th row and l-th column in the attention alignment matrix; Represents the hth dimension value based on the image feature vector; Represents the lth dimension value based on the text feature vector; represents the natural exponential function; represents the sum of all dimension values ​​l of the text feature vector; d is the number of rows and columns of the attention alignment matrix; The joint representation of the data fusion feature vector is: ;in, Represents element-wise product.

[0031] S3. Map the cultural gene code to a four-dimensional space-time coordinate system, and dynamically infer the relationship between nodes through a temporal graph neural network to generate a knowledge graph with space-time attributes; When processing meme-encoded data, the temporal graph neural network uses a graph attention mechanism to dynamically infer the relationships between nodes. Taking the rural tourism knowledge graph as an example, each node can represent a tourist attraction, a cultural activity, or a tourist facility. To calculate the node's attention weight, the node's feature vector is first input into the attention calculation module. A specific calculation method (such as a dot product combined with an activation function) is used to obtain the attention scores between nodes. These scores are then normalized to obtain the attention weight for each node. For example, a popular tourist attraction node may have a closer relationship with surrounding dining facilities and accommodation nodes. The graph attention mechanism assigns higher attention weights to these related nodes, highlighting their importance in reasoning about the relationships and thus more accurately constructing the knowledge graph structure.

[0032] Furthermore, the dynamic update method of the knowledge graph includes: receiving cultural gene encoding data in real time, extracting spatiotemporal events through a sliding time window mechanism, and updating the weight of the knowledge graph.

[0033] Furthermore, when dynamically reasoning about the relationships between nodes, the temporal graph neural network adopts a graph attention mechanism to highlight the influence of key nodes on the reasoning of the relationship by calculating the attention weights of the nodes.

[0034] S4. Generate resource scheduling plans and tourist travel plans based on the knowledge graph, display the output results through the virtual-reality fusion guide device, collect tourists' intention information, and reversely update the node weights of the knowledge graph to optimize the resource scheduling plans and tourist travel plans.

[0035] Furthermore, the virtual-reality integrated tour guide system uses augmented reality and virtual reality technology to combine virtual rural cultural elements with real tourist scenes to provide tourists with an immersive tour guide experience. During the display process, the display effect of virtual elements is adjusted in real time by tracking the tourists' position, movements and gaze. The display content includes but is not limited to the restoration of historical story scenes and the demonstration of the production process of traditional handicrafts.

[0036] A rural cultural characteristic tourist attraction construction system based on big data fusion, and a rural cultural characteristic tourist attraction construction method based on big data fusion, including: data collection module, data processing module, knowledge graph construction module, solution generation module and virtual-reality fusion guide module; The data acquisition module is used to construct a five-dimensional data acquisition matrix, collecting data from the historical context data layer, the physical IoT perception layer, the social media public opinion layer, the economic operation data layer, and the ecological environment monitoring layer, and transmits the collected raw data to the data processing module; The data processing module uses a federated learning framework to interactively process the raw data. It then employs a dynamic weight allocation algorithm to weight the processed raw data in real time, generating a data fusion feature vector. The module then uses a dual-channel feature extraction network to perform cross-modal feature alignment on the data fusion feature vector and output the meme code. The knowledge graph construction module maps the meme code to a four-dimensional space-time coordinate system. It dynamically infers the relationships between nodes through a time-series graph neural network to generate a knowledge graph with space-time attributes. It then updates the weight of the knowledge graph through a sliding time window mechanism based on the real-time meme code data received. The plan generation module generates resource scheduling plans and tourist travel plans based on the knowledge graph; The virtual-reality integrated navigation module displays the output results of the plan generation module, collects tourists' intention information, and feeds the intention information back to the knowledge graph construction module, reversely updates the node weights of the knowledge graph, and optimizes the resource scheduling plan and tourist travel plan.

[0037] Furthermore, the five data layers in the data collection module have different data collection frequencies. The historical context data layer is updated regularly, the physical IoT perception layer collects data in real time, the social media public opinion layer captures data at preset time intervals, the economic operation data layer collects data according to the business cycle, and the ecological environment monitoring layer sets the collection frequency based on the performance of the monitoring equipment and monitoring needs.

[0038] The above description is based on the ideal embodiment of the present invention. Based on the above description, relevant personnel can make various changes and modifications without departing from the technical scope of this invention. The technical scope of this invention is not limited to the content of the specification and must be determined according to the scope of the claims.

Claims

1. A method for constructing rural cultural tourism attractions based on big data fusion, characterized in that: The following steps are involved: S1. Construct a five-dimensional data acquisition matrix for acquiring raw data, and use the federated learning framework to perform data interaction processing on the raw data. Then, use a dynamic weight allocation algorithm to assign real-time weights to the raw data after data interaction processing to generate a data fusion feature vector. S2. Through a dual-channel feature extraction network, cross-modal feature alignment is performed on the data fusion feature vector and the cultural gene code is output; S3. Map the cultural gene code to a four-dimensional space-time coordinate system, and dynamically infer the relationship between nodes through a temporal graph neural network to generate a knowledge graph with space-time attributes; S4. Generate resource scheduling plans and tourist travel plans based on the knowledge graph, display the output results through the virtual-reality fusion guide device, collect tourists' intention information, reversely update the node weights of the knowledge graph, and optimize the resource scheduling plans and tourist travel plans.

2. The method for constructing rural cultural tourism attractions based on big data fusion according to claim 1 is characterized by: The five-dimensional data acquisition matrix includes five data layers, including: historical context data layer, physical object perception layer, social media public opinion layer, economic operation data layer, and ecological environment monitoring layer; the formula expression of the data fusion feature vector is: Where, Represents the weight coefficient based on the information entropy of the i-th layer; is the data feature quantity of the i-th layer; n is the total number of data layers.

3. The method for constructing rural cultural tourism attractions based on big data fusion according to claim 2 is characterized by: The dynamic weight allocation algorithm evaluates the importance of the original data based on the weight coefficient of information entropy, where the calculation formula of the weight coefficient is: Where, represents the information entropy of the zth original data; m represents the total amount of original data.

4. The method for constructing rural cultural tourism attractions based on big data fusion according to claim 2 is characterized by: The gradient aggregation formula of the federated learning framework is: Where, Indicates the result value after the fusion of multiple layers of data features; represents the data feature quantity of the i-th data layer; q represents the total data feature quantity of the five data layers; is the data feature of the i-th layer.

5. The method for constructing rural cultural tourism attractions based on big data fusion according to claim 1 is characterized by: The cross-modal feature alignment includes: constructing an attention alignment matrix , the correlation based on image-text features is calculated, and its formula is: Where, Represents the element value of the h-th row and l-th column in the attention alignment matrix; Represents the hth dimension value based on the image feature vector; Represents the lth dimension value based on the text feature vector; represents the natural exponential function; represents the sum of all dimension values ​​l of the text feature vector; d is the number of rows and columns of the attention alignment matrix; The joint representation of the data fusion feature vector is: ;in, Represents element-wise product.

6. The method for constructing rural cultural tourism attractions based on big data fusion according to claim 1 is characterized by: The dynamic updating method of the knowledge graph includes: receiving cultural gene coding data in real time, extracting spatiotemporal events through a sliding time window mechanism, and updating the weight of the knowledge graph.

7. The method for constructing rural cultural tourism attractions based on big data fusion according to claim 6 is characterized by: The temporal graph neural network adopts a graph attention mechanism when dynamically reasoning about the association relationship between nodes. By calculating the attention weight of the nodes, the influence of key nodes on the association relationship reasoning is highlighted.

8. The method for constructing rural cultural tourism attractions based on big data fusion according to claim 1 is characterized by: The virtual-reality fusion tour guide system utilizes augmented reality and virtual reality technologies to combine virtual rural cultural elements with real tourism scenes, providing tourists with an immersive tour guide experience. During the display process, the display effect of the virtual elements is adjusted in real time by tracking the tourists' position, movements and sight lines. The display content includes but is not limited to the restoration of scenes of historical stories and the demonstration of the production process of traditional handicrafts.

9. A system for constructing rural cultural tourist attractions based on big data fusion, based on the method for constructing rural cultural tourist attractions based on big data fusion according to any one of claims 1 to 8, characterized in that: include: Data collection module, data processing module, knowledge graph construction module, solution generation module and virtual-reality fusion navigation module; The data acquisition module is used to construct a five-dimensional data acquisition matrix, collect data from the historical context data layer, the physical IoT perception layer, the social media public opinion layer, the economic operation data layer, and the ecological environment monitoring layer, and transmit the collected raw data to the data processing module; The data processing module uses a federated learning framework to perform data interaction processing on the raw data, adopts a dynamic weight allocation algorithm to perform real-time weighting on the raw data after data interaction processing, generates a data fusion feature vector, and uses a dual-channel feature extraction network to perform cross-modal feature alignment on the data fusion feature vector to output the cultural gene code; The knowledge graph construction module maps the meme code to a four-dimensional space-time coordinate system, dynamically infers the association between nodes through a time-series graph neural network, generates a knowledge graph with space-time attributes, and updates the weight of the knowledge graph through a sliding time window mechanism based on the meme code data received in real time. The solution generation module generates resource scheduling solutions and tourist travel solutions based on the knowledge graph; The virtual-reality fusion navigation module displays the output results of the plan generation module, collects tourists' intention information, and feeds the intention information back to the knowledge graph construction module, reversely updates the node weights of the knowledge graph, and optimizes the resource scheduling plan and tourist travel plan.

10. The rural cultural tourism attraction construction system based on big data fusion according to claim 9 is characterized by: The five data layers in the data collection module have different data collection frequencies. The historical context data layer is updated regularly, the physical IoT perception layer collects data in real time, the social media public opinion layer captures data at preset time intervals, the economic operation data layer collects data according to the business cycle, and the ecological environment monitoring layer sets the collection frequency based on the performance of the monitoring equipment and monitoring needs.

Images