Disclosure of Invention
In order to solve the technical problem, the invention provides a data visualization processing method which is used for solving the problem that a visualization model is built again because a user puts forward different business requirements.
The embodiment of the invention provides a data visualization processing method, which comprises the following steps:
collecting a plurality of groups of visual data;
carrying out normalization processing on the multiple groups of visual data;
optimizing a plurality of submodels of a preset visualization model according to the plurality of groups of visualization data after normalization processing to obtain a plurality of optimized submodels;
performing integrated processing on the obtained multiple optimization sub-models to obtain a visual model after optimization processing;
and outputting the visual information by the optimized visual model according to the information input by the user.
In one embodiment, the steps of: carrying out normalization processing on the multiple groups of visual data; previously, comprising:
acquiring data type information of a plurality of groups of visual data;
partitioning the multiple groups of visual data according to the data type information, and respectively storing the multiple groups of visual data in the partitions of the corresponding data type information;
acquiring data format information of a plurality of groups of visual data in the subarea;
acquiring the most quantitative data format information in the data format information of the multiple groups of visual data in the partition, and taking the data format information as standard format information in the partition;
and performing format conversion processing on the visual data of which the data format information of the multiple groups of visual data in the subarea is not the standard format information, and converting the visual data of which the data format information of the multiple groups of visual data in the subarea is not the standard format information into the visual data of which the data format information is the standard format information.
In one embodiment, the partitions include an image partition, a text partition, a video partition, and an audio partition;
the data format information in the image partition comprises jpg format information, png format information, tif format information, gif format information, pcx format information and psd format information;
the data format information in the text partition comprises txt format information, rtf format information, doc format information, xls format information and pdf format information;
the data format information in the video partition comprises ASF format information, WMV format information and rmvb format information;
and the data format information in the audio partition comprises APE format, WAV format and OGG format information.
In one embodiment, the steps of: carrying out normalization processing on the multiple groups of visual data; before, still include:
calculating data quantity information of a plurality of groups of the visualization data in the subarea;
acquiring visual data of which the data volume information exceeds preset data volume threshold information;
and extracting the characteristic information of the visual data of which the data volume information exceeds the preset data volume threshold information, and replacing the visual data with the extracted characteristic information of the visual data in the partition.
In one embodiment, the steps of: carrying out normalization processing on the multiple groups of visual data; the method specifically comprises the following steps:
acquiring data information of the visual data in the subarea;
obtaining maximum value information in the data information of the ith visualization data in the partition
Acquiring minimum information in the data information of the ith visual data in the partition
(ii) a The original value information in the data information of the ith visualization data in the partition is
(ii) a N visualization data in the partition;
original value information in data information of the ith visualization data in the partition
Normalization processing is carried out to obtain the visual data of the ith normalization processing
(ii) a The normalization processing formula is as follows:
i=1,2,3,……,n;
normalizing all the visual data in the subarea
Carrying out averaging processing to obtain the visual data X after normalization processing corresponding to the partitions; the averaging processing formula is as follows:
i=1,2,3,……,n。
in one embodiment, the submodels of the visualization model include an input submodel, a data transformation submodel, a visualization mapping submodel, and an output submodel.
In one embodiment, the steps of: optimizing a plurality of submodels of a preset visualization model according to the plurality of groups of visualization data after normalization processing to obtain a plurality of optimized submodels; the method specifically comprises the following steps:
acquiring initial value information and target value information of the visual data after normalization processing;
taking the initial value information as the input of the submodel, taking the target value information as the output of the submodel, and acquiring the optimization weight of the submodel corresponding to the visual data;
respectively repeating the steps on the multiple groups of visual data after normalization processing to obtain multiple optimization weights of the submodels corresponding to the multiple groups of visual data after normalization processing;
acquiring multiple groups of data characteristic information of multiple groups of visual data after normalization processing;
carrying out weighted analysis on a plurality of optimization weights of the submodel corresponding to the plurality of groups of visual data after normalization processing according to the plurality of groups of data characteristic information of the plurality of groups of visual data after normalization processing, and obtaining the optimization values of the plurality of groups of visual data after normalization processing on the submodel;
and optimizing the submodel according to the optimized value to obtain the optimized submodel.
In one embodiment, the steps of: optimizing the submodel according to the optimized value to obtain the optimized submodel; then, the method further comprises the following steps:
taking the initial value information of the visual data after normalization processing as the input of the optimization submodel, and outputting pre-evaluation value information by the optimization submodel;
comparing the target value information of the visual data after normalization processing with the pre-estimated value information output by the optimization submodel to obtain error information;
acquiring a plurality of groups of error information;
adjusting the optimized value of the submodel according to the multiple groups of data characteristic information and the multiple groups of error information of the multiple groups of visual data after normalization processing;
and optimizing the submodel by adopting the adjusted optimization value, and updating the optimized submodel.
In one embodiment, the steps of: performing integrated processing on the obtained multiple optimization sub-models to obtain a visual model after optimization processing; the method specifically comprises the following steps:
acquiring the connection relation of a plurality of sub models of the visual model;
according to the connection relation, performing integrated processing on the optimization sub-model to obtain the optimized visualization model;
detecting the connection interfaces of the optimization submodels, and judging whether the connection interfaces of the optimization submodels can normally transmit data;
when the connection interface of the optimized submodel is judged to be incapable of normally transmitting data, acquiring interface information of the submodel corresponding to the optimized submodel; adjusting and setting the connection interface of the optimization submodel according to the interface information;
the steps are as follows: performing integrated processing on the obtained multiple optimization sub-models to obtain a visual model after optimization processing; then, the method further comprises the following steps:
acquiring test information;
taking the test information as the input of the visual model, and outputting a first test result by the visual model;
taking the test information as the input of the optimized visual model, and outputting a second test result by the optimized visual model;
and comparing and analyzing the first test result output by the visual model and the second test result output by the optimized visual model, acquiring the optimized information of the optimized visual model, and displaying the optimized information of the optimized visual model to a user.
In one embodiment, the steps of: in the process of outputting the visualization information according to the information input by the user, the optimized visualization model further comprises:
acquiring information input by a user, and extracting keywords of the input information;
establishing a first association table based on the keywords based on a data source library;
meanwhile, screening N second association tables related to the first association table on the basis of the database;
acquiring first vertex data and first bottom point data of the first association table, meanwhile, performing coloring processing on the first vertex data and the first bottom point data based on a coloring database, and configuring the colored first vertex data and the colored first bottom point data into a standard three-dimensional coordinate system to obtain first visual color data;
acquiring second vertex data and second bottom point data of N second association tables, and meanwhile, coloring the second vertex data and the second bottom point data based on the coloring database, and acquiring corresponding N second visual color data;
and inputting the first visual color data and the N second visual color data into the visual model after the optimization processing, and outputting visual information by combining the input information.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Detailed Description
The preferred embodiments of the present invention will be described in conjunction with the accompanying drawings, and it will be understood that they are described herein for the purpose of illustration and explanation and not limitation.
An embodiment of the present invention provides a data visualization processing method, as shown in fig. 1, the method includes:
collecting a plurality of groups of visual data;
carrying out normalization processing on the multiple groups of visual data;
optimizing a plurality of submodels of a preset visualization model according to the plurality of groups of visualization data after normalization processing to obtain a plurality of optimized submodels;
performing integrated processing on the obtained multiple optimization sub-models to obtain a visual model after optimization processing;
and outputting the visual information by the optimized visual model according to the information input by the user.
The working principle of the method is as follows: collecting a plurality of groups of visual data, and carrying out normalization processing on the plurality of groups of visual data; optimizing a plurality of submodels of a preset visualization model according to the plurality of groups of visualization data after normalization processing to obtain a plurality of optimized submodels; integrating the plurality of optimization sub-models to obtain a visual model after optimization; and the user transmits the information to the optimized visual model, and the optimized visual model processes the information transmitted by the user and outputs the visual information.
The method has the beneficial effects that: normalization processing is carried out on the collected multiple groups of visual data, and the multiple groups of visual data after normalization processing are obtained; optimizing a plurality of submodels of a preset visualization model according to the plurality of groups of visualization data after normalization processing, so that a plurality of optimized submodels are obtained; the acquired multiple optimization sub-models are subjected to integrated processing, so that the visualization model after optimization processing is acquired; the user transmits the information to the optimized visual model, and the optimized visual model processes the information transmitted by the user, so that the output of the visual information is realized; according to the method, multiple submodels of the visual model are optimized through multiple groups of visual data after normalization processing, and the obtained multiple optimized submodels are subjected to integrated processing, so that the visual model after optimization processing is obtained; compared with the prior art, the method has the advantages that through normalization processing of multiple groups of visual data and integration processing of multiple optimization sub-models, the optimized visual models can meet different business requirements of users; when a user puts forward different business requirements, the visualization model is optimized by collecting visualization data related to the business requirements, the optimized visualization model can realize the processing of the business requirements, and corresponding visualization information is output; the problem that the visualization model is rebuilt due to the fact that the user puts forward different business requirements in the traditional technology is solved, and the data processing efficiency is further improved.
In one embodiment, the steps of: carrying out normalization processing on the multiple groups of visual data; previously, comprising:
acquiring data type information of a plurality of groups of visual data;
partitioning the multiple groups of visual data according to the data type information, and respectively storing the multiple groups of visual data in the partitions of the corresponding data type information;
acquiring data format information of a plurality of groups of visual data in the subarea;
acquiring the most quantitative data format information in the data format information of a plurality of groups of visual data in the subarea, and taking the data format information as standard format information in the subarea;
and carrying out format conversion processing on the visual data of which the data format information of the multiple groups of visual data in the subarea is not standard format information, and converting the visual data of which the data format information of the multiple groups of visual data in the subarea is not standard format information into the visual data of which the data format information is standard format information. According to the technical scheme, the data type information of the multiple groups of visual data is acquired, and the multiple groups of visual data are respectively stored in the partitions of the corresponding data type information, so that the multiple groups of visual data are partitioned according to the data types of the visual data, and the same type of visual data can be conveniently processed in the subsequent steps; acquiring data format information of a plurality of groups of visual data in the partition, and taking the most quantitative data format information in the data format information of the plurality of groups of visual data in the partition as standard format information in the partition; the data format information of the multiple groups of visual data in the subarea is not the visual data of the standard format information, and the visual data of the multiple groups of visual data in the subarea is converted into the visual data of which the data format information is the standard format information, so that the visual data in the subarea is converted into the same data format, and the visual data in the subarea can be conveniently processed in the subsequent steps; and the most data format information in the data format information of the multiple groups of visual data in the partition is used as the standard format information in the partition for conversion, so that the time required for converting the format of the visual data in the partition is effectively shortened, and the format conversion efficiency of the visual data is improved.
In one embodiment, the partitions include an image partition, a text partition, a video partition, and an audio partition; in the technical scheme, the visual data with the data type information of the images in the multiple groups of visual data are transmitted and stored to the image subareas; transmitting and storing the visual data with the data type information as texts in the multiple groups of visual data to the text partitions; transmitting and storing the visual data with the data type information of the video in the multiple groups of visual data to the video subareas; transmitting and storing the visual data with the data type information being audio frequency in the multiple groups of visual data to a video partition; therefore, the partitioning of the multiple groups of visual data according to the data type information of the visual data is realized.
The data format information in the image partition comprises jpg format information, png format information, tif format information, gif format information, pcx format information and psd format information; the technical scheme realizes the acquisition of the data format information of the visual data in the image partition.
The data format information in the text partition comprises txt format information, rtf format information, doc format information, xls format information and pdf format information; by the technical scheme, the data format information of the visual data in the text partition is acquired.
The data format information in the video partition comprises ASF format information, WMV format information and rmvb format information; the technical scheme realizes the acquisition of the data format information of the visual data in the video frequency area.
And data format information in the audio partition comprises APE format, WAV format and OGG format information. By the technical scheme, the data format information of the visual data in the audio partition is acquired.
In one embodiment, the steps of: carrying out normalization processing on the multiple groups of visual data; before, still include:
calculating data quantity information of a plurality of groups of visual data in the subarea;
acquiring visual data of which the data volume information exceeds preset data volume threshold information;
and extracting the characteristic information of the visual data of which the data volume information exceeds the preset data volume threshold information, and replacing the visual data with the extracted characteristic information of the visual data in the subarea. According to the technical scheme, before normalization processing is carried out on multiple groups of visual data, data volume information of the multiple groups of visual data in the subarea is calculated, characteristic information of the visual data, of which the data volume information exceeds preset data volume threshold information, is extracted, and the visual data is replaced by the extracted characteristic information of the visual data in the subarea, so that the data volume information of the multiple visual data is acquired, the characteristic information of the visual data, of which the data volume information exceeds the preset data volume threshold information, is extracted, the visual data is replaced, the data volume of the visual data with overlarge data volume information is effectively reduced, and normalization processing of the visual data in subsequent steps is facilitated.
In one embodiment, the steps of: carrying out normalization processing on the multiple groups of visual data; the method specifically comprises the following steps:
acquiring data information of visual data in a partition;
obtaining maximum value information in data information of ith visual data in subarea
Acquiring minimum information in the data information of the ith visual data in the partition
(ii) a The original value information in the data information of the ith visual data in the partition is
(ii) a The number of the visual data in the subarea is n;
original value information in data information for ith visualization data within a partition
Normalization processing is carried out to obtain visual data of ith normalization processing
(ii) a The normalization processing formula is as follows:
i=1,2,3,……,n;
in the technical scheme, the maximum value information in the data information of the ith visual data in the partition is obtained
Minimum value information
And original value information of
The visual data of the ith normalization processing in the subarea is calculated by a normalization processing formula
And (4) obtaining.
Normalizing all visual data in the subarea
Carrying out averaging processing to obtain visual data X after normalization processing corresponding to the partitions; the averaging processing formula is as follows:
i=1,2,3,……,n。
the technical scheme is that all visual data after normalization processing in the subareas are subjected to
And carrying out averaging processing to realize the acquisition of the visual data X after the normalization processing corresponding to the partition.
In one embodiment, the submodels of the visualization model include an input submodel, a data transformation submodel, a visualization mapping submodel, and an output submodel. According to the technical scheme, the preset visual model is constructed through the input sub-model, the data transformation sub-model, the visual mapping sub-model and the output sub-model.
In one embodiment, the steps of: optimizing a plurality of submodels of a preset visualization model according to the plurality of groups of visualization data after normalization processing to obtain a plurality of optimized submodels; the method specifically comprises the following steps:
acquiring initial value information and target value information of the visual data after normalization processing;
taking initial value information as the input of the submodel, taking target value information as the output of the submodel, and obtaining the optimization weight of the submodel corresponding to the visual data;
respectively repeating the steps on the multiple groups of visual data after normalization processing to obtain multiple optimization weights of the submodels corresponding to the multiple groups of visual data after normalization processing;
acquiring multiple groups of data characteristic information of multiple groups of visual data after normalization processing;
carrying out weighted analysis on a plurality of optimized weights of the submodels corresponding to the plurality of groups of normalized visual data according to the plurality of groups of data characteristic information of the plurality of groups of normalized visual data to obtain optimized values of the plurality of groups of normalized visual data to the submodels;
and optimizing the submodel according to the optimized value to obtain an optimized submodel. In the technical scheme, the initial value information of the visual data after normalization processing is used as the input of the submodel, and the target value information of the visual data after normalization processing is used as the output of the submodel, so that the optimal weight of the submodel corresponding to the visual data is obtained; repeating the steps on the multiple groups of visual data after normalization processing to obtain multiple optimization weights; and carrying out weighted analysis on a plurality of optimized weights of the submodels corresponding to the plurality of groups of visual data after normalization processing according to the plurality of groups of data characteristic information of the plurality of groups of visual data after normalization processing, so as to realize the acquisition of the optimized values of the plurality of groups of visual data after normalization processing on the submodels, and optimizing the submodels by adopting the acquired optimized values, thereby realizing the acquisition of the optimized submodels.
In one embodiment, the steps of: optimizing the submodel according to the optimized value to obtain an optimized submodel; then, the method further comprises the following steps:
taking initial value information of the visual data after normalization processing as input of an optimization submodel, and outputting pre-evaluation value information by the optimization submodel;
comparing the target value information of the visual data after normalization processing with the pre-estimated value information output by the optimization sub-model to obtain error information;
respectively repeating the steps on the multiple groups of visual data after normalization processing to obtain multiple groups of error information;
adjusting the optimized value of the sub-model according to the multiple groups of data characteristic information and the multiple groups of error information of the multiple groups of visual data after normalization processing;
and optimizing the submodel by adopting the adjusted optimization value, and updating the optimized submodel. In the technical scheme, the target value information of the visual data after normalization processing is compared with the pre-estimated value information output by the optimization submodel according to the input target value information of the visual data, so that the error information is acquired; the steps are repeated on the multiple groups of visual data after normalization processing, so that multiple groups of error information are obtained; and adjusting the optimized value of the submodel according to the multiple groups of data characteristic information and the multiple groups of error information of the multiple groups of visual data after normalization processing, optimizing the submodel by adopting the adjusted optimized value, and updating the optimized submodel to enable the predicted value information output by the optimized submodel to be more fit to the target value information.
In one embodiment, the steps of: performing integrated processing on the obtained multiple optimization sub-models to obtain a visual model after optimization processing; the method specifically comprises the following steps:
acquiring the connection relation of a plurality of sub-models of the visual model;
according to the connection relation, performing integrated processing on the optimization sub-model to obtain a visual model after optimization processing; according to the technical scheme, the connection relation of the plurality of submodels of the visual model is obtained, so that the integration processing of the optimized submodels is realized, and the visual model after the optimization processing is obtained.
Detecting the connection interfaces of the optimization submodels, and judging whether the connection interfaces of the optimization submodels can normally transmit data;
when judging that the connection interface of the optimized submodel can not normally transmit data, acquiring interface information of the submodel corresponding to the optimized submodel; adjusting and setting the connection interface of the optimization submodel according to the interface information; in the technical scheme, the connection interfaces of a plurality of optimized submodels of the optimized visual model are detected, whether the connection interfaces of the optimized submodels can normally transmit data is judged, and when the connection interfaces of the optimized submodels cannot normally transmit data is judged, interface information of the submodels corresponding to the optimized submodels is obtained; and adjusting and setting the connection interface of the optimized submodel according to the interface information, so that the optimized visual submodel can normally transmit data, and the normal operation of the optimized visual model is ensured.
The method comprises the following steps: performing integrated processing on the obtained multiple optimization sub-models to obtain a visual model after optimization processing; then, the method further comprises the following steps:
acquiring test information;
the test information is used as the input of a visual model, and the visual model outputs a first test result;
the test information is used as the input of the optimized visual model, and the optimized visual model outputs a second test result;
and comparing and analyzing the first test result output by the visual model and the second test result output by the optimized visual model, acquiring the optimization information of the optimized visual model, and displaying the optimization information of the optimized visual model to a user. In the technical scheme, the visual model takes test information as input to obtain a first test result; the optimized visual model takes the test information as input to obtain a second test result; and comparing and analyzing the first test result and the second test result, so that the optimization information of the optimized visual model is obtained, and the optimization information is displayed to the user, thereby obtaining the optimization performance of the optimized visual model by the user according to the optimization information.
In one embodiment, the steps of: in the process of outputting the visualization information according to the information input by the user, the optimized visualization model further includes, as shown in fig. 2:
step 21: acquiring information input by a user, and extracting keywords of the input information;
step 22: establishing a first association table based on keywords based on a data source library;
meanwhile, screening N second association tables related to the first association table based on the database;
step 23: acquiring first vertex data and first bottom point data of a first association table, meanwhile, coloring the first vertex data and the first bottom point data based on a coloring database, and configuring the colored first vertex data and the colored first bottom point data into a standard three-dimensional coordinate system to obtain first visual color data;
step 24: acquiring second vertex data and second bottom point data of the N second association tables, and meanwhile, coloring the second vertex data and the second bottom point data based on a coloring database, and acquiring corresponding N second visual color data;
step 25: and inputting the first visual color data and the N second visual color data into the visual model after optimization processing, and outputting visual information by combining the input information.
In this embodiment, for example, the information input by the user is fruits in the high season, and at this time, the keyword is fruits in the high season;
the corresponding first association table may be a fruit association table related to the current season, and a plurality of second association tables are filtered based on the fruit association table, and the second association tables may be related to attributes (such as moisture, sugar, etc.) of a certain fruit in the fruit association tables.
The top point data and the bottom point data of the association table are obtained for coloring convenience, for example, the obtained top point data is the fruit color of a certain fruit, and the obtained bottom point data is the fruit shape of a certain fruit.
In this embodiment, the standard three-dimensional coordinates are set so as to enable three-dimensional display and further improve the visualization thereof.
The beneficial effects of the above technical scheme are: the method comprises the steps of extracting keywords of information input by a user, obtaining a first association table based on the keywords, further obtaining a plurality of second association tables according to the first association table, providing a basis for coloring data, improving the reliability of obtaining visual colors by obtaining vertex data and bottom point data, combining the reliability with input information, and improving the visualization of visual information output.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.