Yacht tourism recommendation method and system based on artificial intelligence and big data analysis

Abstract

The invention discloses a yacht tourism recommendation method and system based on artificial intelligence and big data analysis, and the method comprises the steps: quantifying the preference degree Pij, # imgabs0 # of a user i for a yacht type j according to a historical behavior data set Di of the user i, the historical behaviors comprise a reservation behavior, a browsing behavior and user evaluation of a user i for a yacht type j, and the historical behavior data set Di is a set of various historical behaviors; wik represents the weight of the yacht type j in the specific historical behavior k. The method is used for quantifying the preference of the user for different yacht types, so that a more accurate user portrait is constructed.

Description

Technical Field

[0001] The present invention relates to the field of tourism management and big data technology, and in particular to a yacht tourism recommendation method and system based on artificial intelligence and big data analysis. Background Art

[0002] With the development of tourism, yacht tourism has become popular as a unique way of vacation. However, the safety and service quality of yacht tourism have always been the focus of attention. Traditional management methods often cannot effectively solve problems such as equipment failure and user behavior. Therefore, a more intelligent and efficient management system is needed to improve the service quality and operational efficiency of yacht tourism. Summary of the invention

[0003] One of the purposes of the present invention is to propose a yacht tourism recommendation method based on artificial intelligence and big data analysis, which is used to quantify users' preferences for different yacht types, thereby constructing a more accurate user portrait.

[0004] The technical solution adopted by the present invention to achieve the above-mentioned purpose is as follows:

[0005] A yacht tourism recommendation method based on artificial intelligence and big data analysis, comprising the steps of:

[0006] According to the historical behavior dataset Di of user i, quantify the preference Pij of user i for yacht type j,

[0007]

[0008] Among them, rik represents the score of user i's specific historical behavior k, and the historical behavior includes user i's booking behavior, browsing behavior and user evaluation of yacht type j. The historical behavior data set Di is a collection of various historical behaviors; wik represents the weight of yacht type j in the specific historical behavior k.

[0009] As a feasible implementation of the present invention, when calculating the score rik:

[0010] Each booking of yacht type j by user i is scored as a fixed score of 10 points, so the score of booking behavior is the fixed score of each booking (10 points) × the number of bookings;

[0011] The rating of user i’s browsing behavior for yacht type j is calculated based on the browsing time, with each minute being scored as 1 point;

[0012] The rating of user i on yacht type j is calculated based on the user's subjective evaluation and is recorded as 1 to 5 points.

[0013] As a feasible implementation of the present invention, when calculating the weight wik:

[0014] User i’s booking behavior for yacht type j is weighted as 20 points;

[0015] User i’s browsing behavior on yacht type j is weighted as 10 points;

[0016] User i’s evaluation behavior on yacht type j has a weight of 15 points.

[0017] The second object of the present invention is to provide a system for the yacht tourism recommendation method based on artificial intelligence and big data analysis, comprising:

[0018] Data collection module: responsible for collecting yacht sensor data, user historical behavior data and camera monitoring data;

[0019] Data processing module: Use big data to clean, analyze and integrate the collected data;

[0020] Deep learning model: Analyzes user historical behavior, generates yacht guidance push information, and monitors the status of yacht equipment to provide aging and damage warnings;

[0021] Push module: pushes yacht guidance to users and sends warning information to yacht managers based on the output of the deep learning model;

[0022] Feedback module: collect user feedback data.

[0023] Preferably, the data processing module adopts two big data processing frameworks, Spark and Flink.

[0024] Preferably, the deep learning model is a model built using Keras API in TensorFlow.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] 1. The present invention can quantify users' preferences for different yacht types and travel routes, thereby building more accurate user portraits. These user portraits can be used in intelligent recommendation systems to improve user experience and satisfaction.

[0027] 2. The system of the present invention can be widely used in the yacht rental service industry to improve service quality and management efficiency and provide customers with a safer, more comfortable and personalized yacht experience. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical scheme and advantages of the present invention more obvious, the exemplary embodiments of the present invention will be described in detail below. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments of the present invention, and it should be understood that the present invention is not limited to the exemplary embodiments described herein. Based on the embodiments of the present invention described in the present invention, all other embodiments obtained by those skilled in the art without paying creative work should fall within the protection scope of the present invention.

[0029] In the following description, a large number of specific details are provided to provide a more thorough understanding of the present invention. However, it is apparent to those skilled in the art that the present invention can be implemented without one or more of these details. In other examples, in order to avoid confusion with the present invention, some technical features well known in the art are not described.

[0030] It should be understood that the present invention can be implemented in different forms and should not be interpreted as limited to the embodiments set forth herein. On the contrary, these embodiments are provided to make the disclosure thorough and complete and to fully convey the scope of the present invention to those skilled in the art.

[0031] The purpose of the terms used herein is only to describe specific embodiments and is not intended to be limiting of the present invention. When used herein, the singular forms "one", "an" and "said / the" are also intended to include plural forms, unless the context clearly indicates otherwise. It should also be understood that the terms "consisting of" and / or "comprising", when used in this specification, determine the presence of the features, integers, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, parts and / or groups. When used herein, the term "and / or" includes any and all combinations of the relevant listed items.

[0032] In order to fully understand the present invention, a detailed structure will be proposed in the following description to illustrate the technical solution proposed by the present invention. The optional embodiments of the present invention are described in detail as follows, but in addition to these detailed descriptions, the present invention may also have other implementations.

[0033] A yacht tourism recommendation method based on artificial intelligence and big data analysis, comprising the steps of:

[0034] According to the historical behavior dataset Di of user i, quantify the preference Pij of user i for yacht type j,

[0035]

[0036] Among them, rik represents the score of user i's specific historical behavior k, and the historical behavior includes user i's booking behavior, browsing behavior and user evaluation of yacht type j. The historical behavior data set Di is a collection of various historical behaviors; wik represents the weight of yacht type j in the specific historical behavior k.

[0037] As a feasible implementation of the present invention, when calculating the score rik:

[0038] Each booking of yacht type j by user i is scored as a fixed score of 10 points, so the score of booking behavior is the fixed score of each booking (10 points) × the number of bookings;

[0039] The rating of user i’s browsing behavior for yacht type j is calculated based on the browsing time, with each minute being scored as 1 point;

[0040] The rating of user i on yacht type j is calculated based on the user's subjective evaluation and is recorded as 1 to 5 points.

[0041] As a feasible implementation of the present invention, when calculating the weight wik:

[0042] User i’s booking behavior for yacht type j is weighted as 20 points;

[0043] User i’s browsing behavior on yacht type j is weighted as 10 points;

[0044] User i’s evaluation behavior on yacht type j has a weight of 15 points.

[0045] In one embodiment, it is assumed that the historical behavior dataset Di of user i includes the following behaviors:

[0046] Booked yacht type A, 2 bookings in total.

[0047] Browsed the page for yacht type B for 5 minutes.

[0048] Yacht type C was evaluated with a score of 4.

[0049] According to the scoring and weighting rules of the present invention, the preference Pij of user i for yacht type A can be calculated as follows:

[0050] The score for booking behavior is 2 bookings multiplied by 10, which equals 20 points.

[0051] The weight is the booking behavior, and the weight is 2 bookings, multiplied by 20 points, which gives 40 points.

[0052] The user has neither browsed nor evaluated yacht type A. Therefore, when calculating the preference for type A, only the booking behavior needs to be considered.

[0053] Substituting into the formula, we can get user i's preference Pij for yacht type A:

[0054] Similarly, the preference Pij of user i for yacht type B or C can be calculated to construct a user portrait as a basis for recommending yachts to users.

[0055] The user browsed the page for yacht type B for a total of 5 minutes.

[0056] The score of browsing behavior is 5 points, the weight is 10 points, and the preference Pij of user i for yacht type B is:

[0057] Users rated Yacht Type C with a score of 4.

[0058] The score of the user's evaluation behavior is 4 points, the weight is 15 points, and the preference Pij of user i for yacht type C is obtained:

[0059] When user i has multiple historical behaviors on yacht type j at the same time, comprehensive calculation is performed according to the above formula.

[0060] The second object of the present invention is to provide a system for the yacht tourism recommendation method based on artificial intelligence and big data analysis, comprising:

[0061] Data collection module: responsible for collecting yacht sensor data, user historical behavior data and camera monitoring data;

[0062] Data processing module: Use big data to clean, analyze and integrate the collected data;

[0063] Deep learning model: Analyzes user historical behavior, generates yacht guidance push information, and monitors the status of yacht equipment to provide aging and damage warnings;

[0064] Push module: pushes yacht guidance to users and sends warning information to yacht managers based on the output of the deep learning model;

[0065] Feedback module: collect user feedback data.

[0066] The data acquisition module is a key component of the intelligent yacht management system. It is responsible for collecting various sensor data on the yacht, including temperature, humidity, pressure and other parameters that need to be monitored. By connecting temperature sensors, humidity sensors, pressure sensors and other devices, the yacht's environmental information can be obtained in real time for further analysis and processing.

[0067] In the data acquisition module, you can choose to use Raspberry Pi single-board computers as data acquisition devices. These single-board computers have small size, low power consumption and rich GPIO interfaces, which are very suitable for connecting and controlling various sensors.

[0068] First, connect the sensor to the single-board computer. Use a breadboard or a dedicated sensor expansion board to connect devices such as temperature sensors, humidity sensors, and pressure sensors to the GPIO interface of the Raspberry Pi. Depending on the sensor, refer to the corresponding pin diagram and wiring instructions to ensure the correct connection.

[0069] Next, use Python programming language to read and process sensor data. In order to facilitate the control of GPIO interface, you can use RPi.GPIO library, which provides a series of functions to read and control GPIO pins. In addition, in one embodiment, the DHT series temperature and humidity sensor is used, and you can choose to use Adafruit_DHT library, which is specially used for communication with DHT series sensors.

[0070] When writing code, you first need to import the libraries you need to use and initialize the GPIO interface. Then use the corresponding functions to read the sensor values. For example, using the RPi.GPIO library, you can use the GPIO.setup function to set the input or output mode of the GPIO pin, and then use the GPIO.input function to read the level value of the pin. For the DHT series sensors, we can use the read_retry function of the Adafruit_DHT library, which can retry reading the sensor value to ensure accurate data. After reading the sensor data, the data can be further processed. For example, convert the temperature to Celsius or Fahrenheit, calculate the percentage of humidity, or convert the pressure to standard units. Use mathematical operations, conditional judgments, string processing, and other methods to process and convert the data according to specific needs.

[0071] In addition to basic data collection, in one embodiment, other devices such as LED lights, steering gears, etc. are controlled through the GPIO interface to achieve remote control and operation of yacht equipment.

[0072] In summary, the data acquisition module is a very important part of the intelligent yacht management system. By using single-board computers such as Raspberry Pi as data acquisition devices, connecting temperature, humidity, pressure and other sensors, and using Python's RPi.GPIO or Adafruit_DHT library to read and process sensor data, it is possible to efficiently and accurately collect yacht environmental information and provide support for subsequent data analysis and decision-making.

[0073] The data processing module is an important part of the intelligent yacht management system. It is responsible for cleaning, sorting and preprocessing the collected sensor data for subsequent analysis and application. In the data processing module, we can use Spark and Flink, two big data processing frameworks, to process data.

[0074] First, use the Spark library to clean and organize data. Spark is a fast, general-purpose distributed computing system with powerful data processing capabilities. By using Spark's DataFrame API, you can easily perform data cleaning and organization operations. Use Spark's data transformation functions, such as filter, select, and groupBy, to filter, select, and group data. In addition, Spark also provides a wealth of data transformation and operation functions, such as agg, join, and sort, which can meet data processing needs in different scenarios.

[0075] Next, use Flink to process the collected sensor data. Flink is a powerful streaming computing and batch processing framework that can achieve low-latency and high-throughput data processing. By using Flink's DataStream API, real-time streaming data can be cleaned and sorted. Use Flink's window operations and aggregation functions to process streaming data, such as calculating the average, maximum, and minimum values ​​within the sliding window. In addition, Flink also supports the processing of event time and processing time, which can meet the data processing requirements of different precision requirements.

[0076] After data cleaning, operations such as data transformation, feature extraction, and dimensionality reduction are performed according to specific needs. Data transformation can convert data from one form to another to meet specific analysis needs. For example, Spark and Flink APIs can be used to map, filter, and merge data to generate new data sets. Feature extraction is the conversion of raw data into more meaningful features for further analysis and modeling. In the yacht management system, some features related to the status and performance of the yacht are extracted based on sensor data. For example, the average, maximum, and minimum values ​​of temperature, the rate of change of humidity, and the fluctuation amplitude of pressure are calculated. These features can help us better understand and analyze the operation of the yacht. Dimensionality reduction is the process of converting high-dimensional data into low-dimensional data for easy visualization and analysis. In the yacht management system, data from multiple sensors may be collected, and these data may have high dimensions. By using dimensionality reduction techniques such as principal component analysis (PCA) or linear discriminant analysis (LDA), the data can be reduced to two-dimensional or three-dimensional space for easy visualization and better understanding of the data.

[0077] In summary, the main task of the data processing module is to clean, organize and preprocess the collected sensor data. By using the two big data processing frameworks, Spark and Flink, data can be cleaned and organized efficiently, and data conversion, feature extraction and dimensionality reduction can be performed according to needs. This will provide strong support for subsequent data analysis and decision-making.

[0078] The deep learning model is an important part of the intelligent yacht management system. It can be trained using the collected sensor data to predict and classify the equipment status. In the process of building the deep learning model, Python's TensorFlow library will be used.

[0079] TensorFlow is an open source deep learning framework that provides a rich set of tools and functions for building, training, and deploying deep learning models.

[0080] The general steps for building a deep learning model with TensorFlow are as follows:

[0081] First, determine the specific task requirements and select the appropriate model architecture. In a yacht management system, we may face different tasks, such as prediction of equipment status and classification of equipment status. For the prediction of equipment status, choose a model architecture such as a recurrent neural network (RNN) or a long short-term memory network (LSTM). These models are able to process sequence data and have memory capabilities to capture temporal dependencies. For the classification of equipment status, choose a model architecture such as a convolutional neural network (CNN) or a deep neural network (DNN). These models are able to extract features and perform classification, and are suitable for processing data such as images and sounds. Select the corresponding model architecture according to the specific task requirements.

[0082] Next, prepare the data set and perform data preprocessing. In the intelligent yacht management system, the collected sensor data is used as training data. Before deep learning, it is usually necessary to preprocess the data to adapt it to the requirements of the model. The preprocessing steps include data cleaning, data normalization, data partitioning, etc. The purpose of data cleaning is to remove noise and outliers to ensure the quality of the data. The purpose of data normalization is to scale the data to the same range to facilitate model training and convergence. The purpose of data partitioning is to divide the data set into training set, validation set, and test set to facilitate model evaluation and verification of generalization ability.

[0083] Then start building a deep learning model. In TensorFlow, use the Keras API to build the model. Keras is an advanced deep learning library that provides a concise and flexible interface for easily defining and training deep learning models. Use Keras's Sequential model or functional API to define the architecture of the model and add the corresponding layers and activation functions. For example, for a convolutional neural network (CNN), you can use the Conv2D layer and the MaxPooling2D layer to extract the features of the image, and then use the fully connected layer and activation function for classification. For a recurrent neural network (RNN), you can use the LSTM layer or the GRU layer to process sequence data and predict the state. According to the specific task requirements, select the corresponding layer and activation function, and set the appropriate hyperparameters.

[0084] After building the model, you need to select the appropriate loss function and optimizer, and compile and train the model. The loss function is used to evaluate the difference between the model's prediction results and the true value, and the optimizer is used to adjust the model's weights and biases to minimize the loss function. In TensorFlow, select common loss functions such as mean square error (MSE) or cross entropy, and common optimizers such as stochastic gradient descent (SGD) or Adam optimizer. When compiling the model, specify the loss function and optimizer, and optionally add some evaluation indicators such as accuracy. Then use the training data to train the model, and use the validation set to tune the model and select parameters. Set the appropriate batch size and number of iterations during training to control the training speed and effect of the model.

[0085] Finally, the test set is used to evaluate the performance and generalization ability of the model. During the test, we can calculate the loss value and evaluation index of the model and compare them with the true value to evaluate the prediction ability and accuracy of the model. If the model performs well, it is saved and deployed to the actual yacht management system to achieve the prediction and classification of equipment status.

[0086] In summary, the steps to build a deep learning model include determining task requirements and selecting a model architecture, preparing and preprocessing data sets, building a model and selecting a loss function and optimizer, compiling the model and training it, and evaluating it using a test set. By using Python's TensorFlow library, you can easily build, train, and deploy deep learning models to predict and classify device status.

[0087] The push module is an important part of the smart yacht management system. It can generate personalized yacht guidance messages based on user needs and equipment status, and send them to users through SMS, push notifications, etc. In the process of building the push module, the backend server will be built using Python's Flask or Django framework, and the deep learning model will be deployed on the server.

[0088] First, choose a suitable framework to build the backend server. Flask and Django are both popular Python web frameworks that provide rich features and tools to help us quickly build scalable web applications. Flask is a lightweight framework that is easy to learn and use, and is flexible and scalable. Django is a full-featured framework that provides many ready-made components and functions, such as authentication, database management, form processing, etc., which can speed up development. You can choose a suitable framework according to your specific needs and project size.

[0089] Then start building the backend server. In Flask, we can use the routes and view functions provided by the Flask framework to define the API interface. In Django, we can use the URL mapping and view functions provided by the Django framework to define the API interface. By defining appropriate routes and view functions, we can receive and process user requests. For example, we can define a POST request route to receive device status data sent by users. In the view function, we call the deep learning model deployed on the server to predict and classify the device status, and generate corresponding yacht guidance messages according to user needs.

[0090] Next, deploy the deep learning model to the server. During the deployment process, the trained model is loaded into memory and kept available. In the Flask or Django framework, the model loading and prediction functions are used to implement the model deployment. For example, the tf.keras.models.load_model function of TensorFlow can be used to load the model, and the model.predict function can be used to make predictions. By deploying the deep learning model to the server, real-time processing of device status and generation of personalized yacht guidance messages can be achieved.

[0091] When generating personalized yacht guidance messages, corresponding processing and selection are performed according to user needs and device status. For example, if the user needs to receive guidance messages about weather and sea conditions, the weather API and sea condition API can be called to obtain the latest weather and sea condition information, and analyze and recommend them in combination with the device status. Then, the corresponding guidance message is generated based on the analysis results and sent to the user through SMS, push notifications, etc. In the Flask or Django framework, the corresponding libraries and tools are used to implement the SMS and push notification functions. For example, use Twilio or Alibaba Cloud SMS service to send SMS notifications, and use Firebase Cloud Messaging or Jiguang Push to send push notifications.

[0092] Finally, test and tune. During the test, use the test data to send requests to the backend server and verify the server's response and the accuracy of the push message. If the push module needs to be tuned and optimized, you can make corresponding improvements based on the test results.

[0093] The feedback module is a key component of the intelligent yacht management system. It can help collect user feedback data on yacht guidance and optimize and improve the guidance strategy of the system by analyzing and organizing this data. In the process of building the feedback module, Vue3 will be used as the front-end framework, Spring Boot as the back-end framework, and Spark will be used for data analysis and organization.

[0094] First, choose a suitable front-end framework to build a user feedback system. Vue3 is a popular JavaScript framework that provides concise, flexible and efficient tools to help us build interactive user interfaces. By using Vue3 to design a user-friendly feedback form, we can collect user feedback data such as satisfaction and suggestions on Yacht Guide. Vue3 also provides a series of components and lifecycle hook functions that can help us achieve two-way binding and real-time updates of data.

[0095] Next, use Spring Boot to build the backend service. Spring Boot is a rapid development framework that provides a series of tools and modules to help us quickly build reliable backend services. By using SpringBoot, you can define a RESTful API interface to receive feedback data passed by the front end and store it in the database. Spring Boot also provides powerful database operation tools and security authentication mechanisms to help us manage user feedback data and ensure data security and consistency.

[0096] After collecting user feedback data, use Spark to analyze and organize the data. Spark is a fast, general, and scalable big data processing framework that provides rich data processing and analysis functions. By using Spark, you can import user feedback datasets and use Spark's APIs and functions to clean, transform, and analyze the data. For example, use Spark's MLlib library to perform sentiment analysis to understand user satisfaction with yacht guides. Use Spark's machine learning algorithms to build predictive models to predict user behavior and feedback trends.

[0097] By analyzing and organizing user feedback data, valuable insights can be obtained to optimize and improve the system's guidance strategies. For example, if it is found that users are less satisfied with certain specific guidance strategies, these strategies can be adjusted and improved to improve the user experience. The system's functions and performance can also be continuously optimized and improved based on user suggestions and feedback.

[0098] In short, using Vue3 and Spring Boot to build a user feedback system and using Spark for data analysis and collation can help us better understand user needs and feedback, thereby optimizing and improving the system's guidance strategy and enhancing user experience.

[0099] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A yacht tourism recommendation method based on artificial intelligence and big data analysis, characterized in that: Includes steps: According to the historical behavior dataset Di of user i, quantify the preference Pij of user i for yacht type j, Among them, rik represents the score of user i's specific historical behavior k, and the historical behavior includes user i's booking behavior, browsing behavior and user evaluation of yacht type j. The historical behavior data set Di is a collection of various historical behaviors; wik represents the weight of yacht type j in the specific historical behavior k.

2. The yacht tourism recommendation method based on artificial intelligence and big data analysis according to claim 1 is characterized in that: When calculating the score rik: Each booking of yacht type j by user i is scored as a fixed score of 10 points, so the score of booking behavior is the fixed score of each booking (10 points) × the number of bookings; The rating of user i’s browsing behavior for yacht type j is calculated based on the browsing time, with each minute being scored as 1 point; The rating of user i on yacht type j is calculated based on the user's subjective evaluation and is recorded as 1 to 5 points.

3. The yacht tourism recommendation method based on artificial intelligence and big data analysis according to claim 2 is characterized in that: When calculating weight wik: User i’s booking behavior for yacht type j is weighted as 20 points; User i’s browsing behavior on yacht type j is weighted as 10 points; User i’s evaluation behavior on yacht type j has a weight of 15 points.

4. A system using the yacht tourism recommendation method based on artificial intelligence and big data analysis as claimed in any one of claims 1 to 3, characterized in that: include: Data collection module: responsible for collecting yacht sensor data, user historical behavior data and camera monitoring data; Data processing module: Use big data to clean, analyze and integrate the collected data; Deep learning model: Analyzes user historical behavior, generates yacht guidance push information, and monitors the status of yacht equipment to provide aging and damage warnings; Push module: pushes yacht guidance to users and sends warning information to yacht managers based on the output of the deep learning model; Feedback module: collect user feedback data.

5. The system for yacht tourism recommendation method based on artificial intelligence and big data analysis according to claim 4 is characterized by: The data processing module adopts two big data processing frameworks, Spark and Flink.

6. The system for yacht tourism recommendation method based on artificial intelligence and big data analysis according to claim 4, characterized in that: The deep learning model is a model built using Keras API in TensorFlow.

Images