Data analysis and visualization chart recommendation method and system based on deep learning

Abstract

The application discloses a data analysis and visualization chart recommendation method and system based on deep learning, and belongs to the technical field of software development.The implementation of the method comprises the following steps: simulating the behavior of the human brain by constructing a deep neural network to obtain common characteristics of a series of data; recording part of the use habits of a user and associating the use habits with the data characteristics according to the calibration of the user, and storing the use habits as key-value pair data of the personal information of the user; constructing a dictionary library of the data characteristics, storing binary fields formed by referring to built-in index values of the data characteristics; constructing a dictionary library of user behaviors, storing binary fields formed by referring to built-in index values of the user habits; constructing a pure function to obtain a data visualization chart type conforming to the use habits and expectations of the user and conforming to a probability; and obtaining a data chart type most conforming to the requirements of the user.The application helps the user to sort out disordered data, obtain regular, reliable and stable data, and realize the visualization display of the data.

Description

Technical Field

[0001] This invention relates to the field of software development technology, specifically to a data analysis and visualization chart recommendation method and system based on deep learning. Background Technology

[0002] In the software development industry, the effectiveness of data visualization varies greatly depending on the data type. The variability of data and the multi-dimensional nature of data analysis lead to differing expectations from users regarding the same data, creating a one-to-many relationship between data and visualizations, resulting in a discrepancy between user needs and development outcomes. Currently, in the software development industry, data visualization development faces challenges due to the variability of monthly data and varying data analysis requirements. A single data visualization format is no longer sufficient to meet the complex business needs of users and fails to satisfy their demands for multi-dimensional data visualization. Summary of the Invention

[0003] The technical objective of this invention is to address the above-mentioned shortcomings by providing a data analysis and visualization chart recommendation method and system based on deep learning. This method can help users organize messy data to obtain regular, reliable, and stable data, as well as the regular characteristics of a set of data. It can also help users visualize the data, facilitating users' judgment, analysis, and secondary utilization of the data.

[0004] The technical solution adopted by this invention to solve its technical problem is:

[0005] A data analysis and visualization chart recommendation method based on deep learning, the implementation of which includes:

[0006] By constructing deep neural networks to simulate the behavior of the human brain, machines can recognize, interpret, and analyze data, and obtain a series of common features of the data.

[0007] Based on the user's calibration, some of the user's usage habits are recorded and associated with data characteristics, and stored as key-value pairs of user personal information;

[0008] Build a dictionary of data features, convert the data features into binary fields based on the built-in indicator values, and store them; build a dictionary of user behavior, convert user habits into binary fields based on the built-in indicator values, and store them.

[0009] Construct pure functions to obtain data visualization chart types and their probabilities that conform to user habits and expectations;

[0010] Obtain the data chart type that best meets the user's needs.

[0011] This method utilizes deep learning and large model techniques to learn the impact of user behavior on the visualization charts required by users. Based on factors such as user raw data, data expectations, and user behavior, it obtains a relationship indicator between the user raw data and the built-in icon library. Then, based on the magnitude of the relationship indicator, it uses pure utility functions to obtain the configuration file for the visualization chart, and further generates the visualization chart for users to display.

[0012] Furthermore, the specific implementation steps of this method are as follows:

[0013] 1) Implement deep neural networks using the open-source Keras library, train models using big data, and obtain data features that meet the requirements based on user data; Keras is an open-source artificial neural network library written in Python, which can be used as a high-level application programming interface for the design, debugging, evaluation, application, and visualization of deep learning models.

[0014] 2) Based on user input data expectations and personal behavioral habits, key-value pairs are formed and stored as the user's personal information attributes;

[0015] 3) Determine the weight values ​​for data features and user behaviors;

[0016] 4) Construct a dictionary of data features and define binary weight values ​​for each data feature;

[0017] 5) Build a user behavior dictionary and assign weight values ​​in binary form to user behaviors;

[0018] 6) Implement a pure function that derives the general data characteristics of a user's data based on the user's data, the user's expectations of the data, and the user's personal attributes;

[0019] 7) Obtain a visual chart that meets the user's expectations.

[0020] Furthermore, in step 1), the data features that meet the requirements refer to the data features that have regularity initially obtained by weighted calculation based on the user's expectations and personal information attributes in the dictionary.

[0021] Furthermore, in step 2), the data expectation refers to the user's assessment of the weight value of the data feature based on the weight requirements of the dictionary.

[0022] Furthermore, in step 3), the weight value refers to the relationship coefficient between data features obtained after training on a large amount of data within the system, which conforms to the behavior and expectations of most users. For example, if a user sorts complex data, we speculate that the user wants to obtain a comprehensive overview of the data and compare the overall numerical values. After verification with a large amount of data, we call the probability that the accurate value of the inferred result matches the true value the weight coefficient.

[0023] Furthermore, in steps 4) and 5), the dictionary refers to the system's internal database that identifies the relationship between data and user behavior and the behavior as pointing to a certain visual chart.

[0024] Binary refers to a counting method. Using binary can better perform weighted calculations.

[0025] Furthermore, in step 6), the pure function refers to a JS execution method whose output parameters depend only on the input parameters and does not produce any side effects during execution, nor does it modify or depend on any external state.

[0026] Furthermore, in step 7), "meeting user expectations" means the best result as perceived by the system under the influence of factors such as internal dictionary values, user behavior, and user data.

[0027] This invention also claims protection for a deep learning-based data analysis and visualization chart recommendation system, including:

[0028] Obtain common feature modules and use deep learning technologies such as Keras and large model techniques to obtain general features for a series of data.

[0029] A dictionary module is built, which is an internal dictionary for user data expectations and user behavior based on social surveys and user needs analysis.

[0030] The visualization chart module, under the influence of dictionary values, re-weights user behavior and data expectations, and combines them with the original data to obtain the data chart type that meets the user's expectations. It then uses a pure function tool to obtain the configuration file for the data chart, thereby further obtaining the visualization chart required by the user.

[0031] Specifically, the system obtains the final visualized chart through the aforementioned deep learning-based data analysis and visualization chart recommendation method.

[0032] The present invention also claims a deep learning-based data analysis and visualization chart recommendation device, comprising at least one memory and at least one processor;

[0033] The at least one memory is used to store a machine-readable program;

[0034] The at least one processor is used to call the machine-readable program to implement the above method.

[0035] Compared with existing technologies, the data analysis and visualization chart recommendation method and system based on deep learning of the present invention have the following advantages:

[0036] 1. This technical solution can meet the user's needs on the development side. On the basis of ensuring big data training, it can predict the types of data charts needed in the future based on some of the user's operational behaviors. It can also learn the user's thinking behavior through user intervention to realize a multi-dimensional and changeable visualization chart analysis tool.

[0037] 2. This technical solution can save code on the development side, eliminating the need to write multiple different data processing methods to obtain different data structures for the same data, and eliminating the need to write multiple data charts on the front end;

[0038] 3. This technical solution can, to some extent, replace low-code tools, enabling the display of different charts under user intervention and achieving system integration. Attached Figure Description

[0039] Figure 1 This is a flowchart illustrating the data analysis and visualization chart recommendation method based on deep learning provided in an embodiment of the present invention. Detailed Implementation

[0040] The present invention will be further described below with reference to specific embodiments.

[0041] This invention provides a data analysis and visualization chart recommendation method based on deep learning. The implementation of this method includes:

[0042] By constructing deep neural networks to simulate the behavior of the human brain, machines can recognize, interpret, and analyze data, and obtain a series of common features of the data.

[0043] Based on the user's calibration, some of the user's usage habits are recorded and associated with data characteristics, and stored as key-value pairs of user personal information;

[0044] Build a dictionary of data features, convert the data features into binary fields based on the built-in indicator values, and store them; build a dictionary of user behavior, convert user habits into binary fields based on the built-in indicator values, and store them.

[0045] Construct pure functions to obtain data visualization chart types and their probabilities that conform to user habits and expectations;

[0046] Obtain the data chart type that best meets the user's needs.

[0047] This method utilizes deep learning and large model techniques to learn the impact of user behavior on the visualization charts required by users. Based on factors such as user raw data, data expectations, and user behavior, it obtains a relationship indicator between the user raw data and the built-in icon library. Then, based on the magnitude of the relationship indicator, it uses pure utility functions to obtain the configuration file for the visualization chart, and further generates the visualization chart for users to display.

[0048] The specific implementation steps of this method are as follows:

[0049] 1. Implement a deep neural network using the open-source Keras, train the model using big data, and obtain data features that meet the requirements based on user data;

[0050] 2. Based on user input data expectations and personal behavioral habits, key-value pairs are formed and stored as the user's personal information attributes;

[0051] 3. Define the weight values ​​for data features and user behaviors;

[0052] 4. Construct a dictionary of data features and define binary weight values ​​for each data feature;

[0053] 5. Build a user behavior dictionary and assign weight values ​​in binary form to user behaviors;

[0054] 6. Implement a pure function that derives the general data characteristics of a user's data based on the user's data, the user's expectations of the data, and the user's personal attributes.

[0055] 7. Obtain visual charts that meet the user's expectations.

[0056] Keras is an open-source artificial neural network library written in Python that can be used as a high-level application programming interface for the design, debugging, evaluation, application, and visualization of deep learning models.

[0057] In step 1, the data features that meet the requirements refer to the regular data features that are initially obtained by weighted calculation based on the user's expectations and personal information attributes in the dictionary.

[0058] In step 2, the data expectation refers to the user's assessment of the weight value of the data features based on the weight requirements of the dictionary.

[0059] In step 3, the weight value refers to the relationship coefficient between data features obtained after training on a large amount of data within the system, which conforms to the behavior and expectations of most users. For example, if a user sorts complex data, we speculate that the user wants to obtain an overview of the data and compare the overall numerical values. After verification with a large amount of data, we call the probability that the predicted result matches the actual value the weight coefficient.

[0060] In steps 4 and 5, the dictionary refers to the system's internal system for identifying the relationship between data and user behavior based on data characteristics and user behavior, and the relationship between behavior and a certain visual chart; binary refers to the counting method, which allows for better weighted calculation.

[0061] In step 6, the pure function refers to a JavaScript execution method whose output parameters depend only on the input parameters and which does not produce any side effects during execution, nor does it modify or depend on any external state.

[0062] In step 7, "meeting user expectations" means the best result that the system considers under the influence of factors such as internal dictionary values, user behavior, and user data.

[0063] The following are specific applications of this method:

[0064] ① In the early stage, deep learning models are used to train large models, so that the large models can obtain general data features from a series of data;

[0065] ② Design user attributes and generate user attribute storage based on users' prior data expectations and user behavior;

[0066] ③ Conduct market research and extensive user research, and formulate weight values ​​corresponding to different data characteristics and different user behaviors based on the research results;

[0067] ④ Based on the results obtained in ③, construct an internal dictionary for the system and obtain weight values ​​in a hybrid binary form based on data characteristics and user behavior;

[0068] ⑤ Construct pure functions within the system, with input parameters including user-provided data, user's initial data expectations, and user-specific attributes obtained from ②③④. Under the action of a series of functions, a unique return value is obtained—general data characteristics of the original data.

[0069] ⑥ Construct pure functions for visual chart types and raw data, enabling the generation of configuration files for corresponding visual charts based on chart type and raw data.

[0070] This invention also provides a data analysis and visualization chart recommendation system based on deep learning, comprising:

[0071] Obtain common feature modules and use deep learning technologies such as Keras and large model techniques to obtain general features for a series of data.

[0072] A dictionary module is built, which is an internal dictionary for user data expectations and user behavior based on social surveys and user needs analysis.

[0073] The visualization chart module, under the influence of dictionary values, re-weights user behavior and data expectations, and combines them with the original data to obtain the data chart type that meets the user's expectations. It then uses a pure function tool to obtain the configuration file for the data chart, thereby further obtaining the visualization chart required by the user.

[0074] Specifically, the system obtains the final visualized chart through the deep learning-based data analysis and visualization chart recommendation method described in the above embodiments.

[0075] Module for obtaining public features:

[0076] We use the open-source Keras to implement deep neural networks, use big data for model training, and obtain data features that meet the requirements based on user data.

[0077] Build the dictionary library module:

[0078] Based on user input data expectations and personal behavioral habits, key-value pairs are formed and stored as the user's personal information attributes;

[0079] Define the weight values ​​for data features and user behaviors;

[0080] Build a dictionary of data features and define binary weight values ​​for each data feature;

[0081] Build a user behavior dictionary and assign weight values ​​in binary form to user behaviors;

[0082] Visualization chart module:

[0083] Implement a pure function that derives the general data characteristics of a user's data based on the user's data, the user's expectations of the data, and the user's personal attributes.

[0084] Obtain visual charts that meet the user's expectations.

[0085] The present invention also claims a deep learning-based data analysis and visualization chart recommendation device, comprising at least one memory and at least one processor;

[0086] The at least one memory is used to store a machine-readable program;

[0087] The at least one processor is used to call the machine-readable program to implement the above method.

[0088] Through the specific embodiments described above, those skilled in the art can easily implement the present invention. However, it should be understood that the present invention is not limited to the specific embodiments described above. Based on the disclosed embodiments, those skilled in the art can arbitrarily combine different technical features to achieve different technical solutions.

[0089] Except for the technical features described in the specification, all other technologies are known to those skilled in the art.

Claims

1. A data analysis and visualization chart recommendation method based on deep learning, characterized in that, The implementation of this method includes: By constructing deep neural networks to simulate the behavior of the human brain, machines can recognize, interpret, and analyze data and obtain a series of common features of the data. Based on the user's calibration, some of the user's usage habits are recorded and associated with data characteristics, and stored as key-value pairs of user personal information; Build a dictionary of data features, convert the data features into binary fields based on the built-in indicator values, and store them; build a dictionary of user behavior, convert user habits into binary fields based on the built-in indicator values ​​and store them. Construct pure functions to obtain data visualization chart types and their probabilities that conform to user habits and expectations; Obtain the data chart type that best meets the user's needs; The specific implementation steps of this method are as follows: 1) Implement deep neural networks using the open-source Keras framework, train models using big data, and obtain data features that meet the requirements based on user data; 2) Based on user input data expectations and personal behavioral habits, key-value pairs are formed and stored as the user's personal information attributes; 3) Determine the weight values ​​for data features and user behaviors; 4) Construct a dictionary of data features and define binary weight values ​​for each data feature; 5) Build a user behavior dictionary and assign weight values ​​in binary form to user behaviors; 6) Implement a pure function that derives the general data characteristics of a user's data based on the user's data, the user's expectations of the data, and the user's personal attributes formed for that user; 7) Obtain visual charts that meet the user's expectations; In step 3), the weight value refers to the relationship coefficient between data features obtained after training on a large amount of data within the system, which conforms to the behavior and expectations of most users. In steps 4) and 5), the dictionary refers to the system's internal database that identifies the relationship between data and user behavior based on data characteristics and user behavior, pointing to a certain visual chart; binary refers to the counting method. In step 6), the pure function refers to a JS execution method whose output parameters depend only on the input parameters and does not produce any side effects during execution, nor does it modify or depend on any external state.

2. The data analysis and visualization chart recommendation method based on deep learning according to claim 1, characterized in that, In step 1), the data features that meet the requirements refer to the regular data features that are initially obtained by weighted calculation based on the user's expectations and personal information attributes in the dictionary.

3. The data analysis and visualization chart recommendation method based on deep learning according to claim 1, characterized in that, In step 2), the data expectation refers to the user's assessment of the weight value of the data feature based on the weight requirements of the dictionary.

4. The data analysis and visualization chart recommendation method based on deep learning according to claim 1, characterized in that, In step 7), "meeting user expectations" means the best result that the system considers under the influence of internal dictionary values, user behavior, and user data conditions.

5. A data analysis and visualization chart recommendation system based on deep learning, characterized in that, include: Obtain common feature modules and use deep learning and large model techniques to obtain general features for a series of data. A dictionary module is built, which is an internal dictionary for user data expectations and user behavior based on social surveys and user needs analysis. The visualization chart module, under the influence of dictionary values, re-weights and calculates user behavior and data expectations, and combines them with the original data to obtain the data chart type that meets the user's expectations. It then uses a pure function tool to obtain the configuration file for the data chart, thereby further obtaining the visualization chart required by the user. Specifically, the system obtains the final visualized chart through the deep learning-based data analysis and visualization chart recommendation method described in any one of claims 1 to 4.

6. A data analysis and visualization chart recommendation device based on deep learning, characterized in that, Includes at least one memory and at least one processor; The at least one memory is used to store a machine-readable program; The at least one processor is configured to invoke the machine-readable program to implement the deep learning-based data analysis and visualization chart recommendation method as described in any one of claims 1 to 4.

Images