CN111767317A - Big data decision analysis method based on data joint service - Google Patents

Abstract

A big data decision analysis method based on data joint service relates to the technical field of data analysis and intelligent decision. The method comprises the following steps: 1) and (3) metadata planning: planning indexes and grouping metadata information; 2) preparing data; 3) special subject configuration: one is to complete the configuration for the distribution user after creating the special topic; the other is to finish the configuration of column creation, column index selection and analysis example configuration after creating a special subject; 4) and (3) special subject release: 5) and (3) data authority configuration: sequentially configuring a user strategy and a data authority strategy; 6) and (5) displaying and analyzing the special subject. Compared with the prior art, the invention provides a universal data analysis decision support mode, which is an open, dynamic and omnibearing auxiliary decision work platform for deep fusion of data along with the process of analyzing thinking by a decision maker.

Description

Big data decision analysis method based on data joint service

Technical Field

The invention relates to the technical field of data analysis and intelligent decision making, in particular to a method for enabling a user to realize autonomous and interactive exploration analysis and infinite mesh data roaming on an analysis platform, aiming at deeply mining the association relation and the potential value among big data as a core and providing perfect support and experience for the macroscopic decision making requirements of managers.

Background

Currently, humans are moving from the IT era to the DT era. The DT era is a data-driven era, data becomes the leading edge of national competition and the source of enterprise innovation, and the industrial value chain with big data and big data as the core is influencing and leading new economic paradigm and national strategy, and becomes a tool for leading social revolution, promoting government function transformation and exciting enterprise technical innovation.

With the development and popularization of computer technology and the wide application of various information systems, a large amount of original data are accumulated in various systems, the rules contained in the data are analyzed, and the operation trend of related systems is predicted, so that the urgent needs of various industries in the modern world are provided.

At present, the popular data analysis application in the market carries out analysis on prepared data through analysis means, methods and skills, finds causal relationships, internal relations and business rules, and provides decision reference for users. And presenting the result of the data analysis in a mode of a graph and a table by means of a data presentation means. Making the data analyst more intuitive in expressing the information, perspectives, and suggestions that are desired to be presented.

The use of tools and methods is involved in the need to be able to manipulate data and perform data analysis using data analysis tools that are popular in the market. One of them needs to be familiar with the conventional data analysis method, and the most basic needs to know the principle, the application range, the advantages and the disadvantages and the explanation of the results of the multivariate and data analysis methods such as variance, regression, factors, clustering, classification, time series and the like; secondly, the method is familiar with various data analysis tools, Excel is the most common, general data analysis can be completed through Excel, and more professional data analysis needs to master a professional analysis software, such as data analysis tools SPSS/SAS/R/Matlab and the like, so that professional statistical analysis, data modeling and the like are facilitated.

The existing data analysis technology has the following defects: such data analysis tools have a high requirement on the expertise of the user, i.e., the analysis tools, data analysis algorithms (e.g., variance, regression, factors, clustering, classification, time series, etc.), algorithm parameters, analysis steps, etc. need to be known in depth before the analysis tools can be used to develop the analysis. The analysis required to be performed by these tools is associated with a high learning cost for the average person. Meanwhile, the tools cannot organically integrate data, and the correlation is limited in the industry (profession), namely a solidified state. The data analysis process can not be infinitely walked and roamed by utilizing the internal relation among the data. The analysis range is limited to the small range data specified by the user. And no recommendations and associations can be made for other data that may be associated with such data.

Disclosure of Invention

In view of the above-mentioned shortcomings in the prior art, the present invention provides a big data decision analysis method based on data networking. The universal data analysis decision support mode is an open, dynamic and omnibearing auxiliary decision work platform for data deep fusion along with the analysis thinking process of a decision maker.

In order to achieve the above object, the technical solution of the present invention is implemented as follows:

a big data decision analysis method based on data joint service comprises the following steps:

1) and (3) metadata planning: planning indexes and grouping metadata information;

2) preparing data:

a) if no import template exists, the frequency and the organization template are selected for data import; if the import template is directly selected, importing the data;

b) the import mode comprises the steps of adding newly-added part data and updated data and covering the original data to finish data import;

3) special subject configuration: one is to complete the configuration for the distribution user after creating the special topic; the other is to finish the configuration of column creation, column index selection and analysis example configuration after creating a special subject;

4) and (3) special subject release:

a) checking whether the configuration is completed;

b) if the configuration is not completed, continuing to check the patent configuration; after the configuration is completed, the special subject is issued;

5) and (3) data authority configuration: sequentially configuring a user strategy and a data authority strategy;

6) and (3) thematic display analysis:

a) checking the viewing authority;

b) obtaining data after the verification is qualified;

c) if the data authority needs to be configured, filtering the data and then analyzing thematic construction; if the data authority does not need to be configured, directly performing analysis topic construction;

d) after the analysis topic display is completed, simultaneously carrying out analysis conversion, analysis roaming and example curing;

e) judging whether the result meets the analysis requirement, and returning to the step a) to restart if the result does not meet the analysis requirement; and finishing if the analysis requirement is met.

In the big data decision analysis method based on data networking, the method for configuring the analysis instance includes:

1) creating an instance and selecting an analysis template;

2) respectively finishing configuration display indexes and configuration space dimensions;

3) completing configuration time dimension after data validity verification is carried out;

4) and respectively carrying out personalized analysis configuration and display content configuration.

In the big data decision analysis method based on data networking, the user policy configuration method includes:

1) creating a user group;

2) and respectively finishing the group allocation strategy and the group designated user adding.

In the big data decision analysis method based on data networking, the method for configuring the data authority policy includes:

1) creating a data permission policy;

2) respectively finishing the distribution of the authority of the structured data and the authority of the unstructured data;

3) the user groups are matched.

The decision analysis method is adopted, so that a universal decision support mode is provided, various decision resources such as indexes, algorithms, models, data, knowledge and the like are provided in a proper mode at a proper time along with the analysis thinking process of a decision maker for the decision maker to select, and the decision maker is helped to realize data-driven scientific decision to the greatest extent. Meanwhile, the method realizes the business and convenience of the analysis platform, and enables a user to quickly complete the design and configuration of various special subjects by utilizing a concise configuration function based on various prefabricated common analysis templates. In addition, the method of the invention is based on the business encapsulation of the product to the algorithm, thereby reducing the analysis threshold, and common personnel can also rapidly utilize the product to complete the required analysis; the data is subjected to multivariate data fusion by depending on a two-stage data architecture, so that data joint work is completed, data resources in a social management form are formed, and the data are in place at any time; during the analysis process, interactive exploration and analysis and infinite data roaming of data are performed based on a data relation mechanism inherent in the product.

The invention is further described with reference to the following figures and detailed description.

Drawings

FIG. 1 is a flow chart of a method of the present invention;

FIG. 2 is a flow chart of data preparation in the method of the present invention;

FIG. 3 is a flow chart of topic allocation in the method of the present invention;

FIG. 4 is a sub-flowchart of the analysis example configuration of FIG. 3;

FIG. 5 is a topic publishing flowchart of the present invention;

FIG. 6 is a flow chart of data right configuration in the present invention;

FIG. 7 is a user policy configuration sub-flow diagram of FIG. 6;

FIG. 8 is a sub-flowchart of the data permission policy configuration of FIG. 6;

FIG. 9 is a flowchart of the topic display analysis of the present invention.

Detailed Description

Referring to fig. 1 to 9, the big data decision analysis method based on data networking of the present invention includes the following steps:

1) and (3) metadata planning: the metrics are planned and metadata information is grouped.

2) Preparing data:

a) if no import template exists, the frequency and the organization template are selected for data import; if the import template is directly selected, importing the data;

b) the import mode is divided into adding new part data and updating data and covering the original data to finish data import.

3) Special subject configuration: one is to complete the configuration for the distribution user after creating the special topic; the other is to finish the configuration of column creation, column index selection and analysis example configuration after creating a special subject;

the method for analyzing the configuration of the example comprises the following steps:

creating an instance and selecting an analysis template;

respectively finishing configuration display indexes and configuration space dimensions;

completing configuration time dimension after data validity verification is carried out;

and respectively carrying out personalized analysis configuration and display content configuration.

4) And (3) special subject release:

a) checking whether the configuration is completed;

b) if the configuration is not completed, continuing to check the patent configuration; after the configuration is completed, the special subject is issued;

5) and (3) data authority configuration: sequentially configuring a user strategy and a data authority strategy;

the user strategy configuration method comprises the following steps:

creating a user group;

and respectively finishing the group allocation strategy and the group designated user adding.

The method for configuring the data authority strategy comprises the following steps:

creating a data permission policy;

respectively finishing the distribution of the authority of the structured data and the authority of the unstructured data;

the user groups are matched.

6) And (3) thematic display analysis:

a) checking the viewing authority;

b) obtaining data after the verification is qualified;

c) if the data authority needs to be configured, filtering the data and then analyzing thematic construction; if the data authority does not need to be configured, directly performing analysis topic construction;

d) after the analysis topic display is completed, simultaneously carrying out analysis conversion, analysis roaming and example curing;

e) judging whether the result meets the analysis requirement, and returning to the step a) to restart if the result does not meet the analysis requirement; and finishing if the analysis requirement is met.

The big data decision analysis system based on data joint service covers the whole process services of metadata planning, data access, data integration, authority management, resource overview, special topic configuration, special topic display and the like.

Metadata planning is the starting point of the intelligent decision support service and is used for planning indexes and grouping metadata information. The metadata is the core for constructing the data resources of the social management form, is also a link for performing data joint work, is a bridge for realizing autonomous roaming of data, and is particularly the centralized embodiment of the decision support analysis thought on the requirements of data services.

With the basis of metadata, the next step is to access external data into the system. The data used in the final decision support stage is concentrated comprehensive data, so that the system starts with the access of the comprehensive data, the access of the comprehensive data is divided into two steps, a comprehensive data import template is established in the first step, a common comprehensive data meta template is defaulted by the system, the meta template is selected according to the data content, the meta data information in the data is designated in a dragging mode and is stored as a template of a specific comprehensive data type, and the comprehensive data is imported into the system through the template in a batch and staged mode. Besides importing the comprehensive data, the system also supports access to the unstructured document in an uploading mode. The metadata description of the unstructured document is system preset. And completing the metadata process of the unstructured data through the access process.

In the process of accessing data, the relation between the data and the metadata is preliminarily established through the import template, the primary form of the data resource is formed, and in order to better process the data resource, form social resources in social management form and support subsequent data joint work, the system provides data integration means in various forms, including related index setting, space dimension management, layer management and the like. And the data joint support user performs autonomous roaming analysis according to the data characteristics and the data service relationship in the data analysis process. The data characteristics depend on the relation of metadata, and the service characteristics need to be specified by a user at a service level, for example, relevant index rules of the service level are set through relevant index setting. Different fields and industries have different distinguishing rules on the spatial dimension, and in order to better establish the relation of data on the spatial dimension and improve the friendliness of spatial dimension display, a user needs to participate in constructing information of other spatial dimensions except for administrative divisions. In addition, the system is preset with some common service index processing algorithms, so that when a user checks data, the user can check processed results such as ratio, ring ratio, sorting, mean value and the like through the algorithms at any time. The fusion process of the social management form data resources of the system covers the part of the foreground needing the participation of the user and also covers the part of the background system automatically processing according to the rules and the preset information.

The control of the data authority is used for limiting whether a user is authorized to view certain data resources on one hand, and is used for conveniently managing the data resource range concerned by the user on the other hand, so that the data concerned by the user can be seen more accurately and timely. The latter is mainly embodied on a comprehensive billboard, and is used for recommending the most concerned field comprehensive data for the user through the setting of the authority. The former is mainly embodied in applications to data resources, such as configuration and viewing thematic and autonomous roaming analysis processes. The authority control strategy is different for different scenes, and the authority control level of the authority control strategy is a domain level, namely, a user has the right to check which domains, and all index data in the domain can be checked. The former combines with the user group, and controls the metadata range accessible by the user group through the metadata system, and also controls the data resource range accessible by the user group, and then establishes the affiliation relationship between the user and the user group, and realizes the control of the data resource range accessed by the user. The end user accessible data resource scope is the union of the two.

Social management form the social resource is a complete data resource system with a network relation and service attributes, and is convenient for data networking. However, in actual services, data is divided into fields, and each field is related to key indexes, so that in order to facilitate a user to quickly check the key indexes according to the fields in service application and further find problems, the user carries the problems to analyze and make decisions, the system provides a comprehensive panel function for intensively displaying the key indexes in the divided fields, and the user can quickly see the data of the field concerned by the user through data authority. Configuration of the domain and selection of the index under the domain are accomplished with the participation of an administrator. This is also a function of the composite billboard portion.

The special subject is to carry out deep observation and analysis on a certain field or key and hot problems of economic and social development, fuse and organize the problems with related policies, data and analysis reports together, display the problems through rich visual means of graphs, tables, information graphs and GIS, and a user can carry out multi-angle and multi-level exploratory analysis according to decision-making requirements. For a specific field and focus, the administrator is required to configure the topic in advance. According to the structure of the special topic, a special topic framework, namely, special topic basic information, special topic column information and special topic example information, is maintained. And secondly, setting data resource information required by thematic display, wherein examples under thematic columns are generated according to the thematic templates, and after the generation, setting index data related to the examples and related apertures thereof, such as time range, area range and other grouping apertures. Or selecting a document template and associating the related unstructured documents. And after each instance under the theme is configured, checking the configured theme in a preview mode. And issuing after checking, giving the user the authority of browsing the special topic, and finally finishing the configuration of the special topic.

The above processes are all functions used by a configuration administrator, and the functions used by an end user mainly comprise three functions, namely resource overview, comprehensive watching board and thematic browsing.

The resource overview column shows statistical information of index conditions, data conditions and unstructured document storage conditions in the current system, and a user can conveniently know the overall condition of data from the whole. Wherein: the index resource overview realizes the summary statistics of the number of indexes, the number of subdivided caliber indexes and the index classification information in the system, and a user can check the detailed information of the indexes and add the indexes to my attention; the data resource overview realizes the summary statistics of the total data amount, the time span of the data and the area coverage condition, and a user can check the index data condition and can also directly convert the index data condition into index time trend analysis; the unstructured document overview realizes summary statistics of the number of documents, document classification and total clicking amount of the documents in the system, and a user can inquire and check the documents in detail.

The comprehensive billboard is a personalized information integration display and processing window customized for the leadership according to the leader management field, focus of attention and behavior habits. The urban big data resource based on deep fusion helps leaders to quickly master the overall situation of the field by providing value-added services such as data classification viewing, quick positioning, related index/related subject roaming and the like, and timely discovers and solves various daily decision problems.

Thematic analysis is a core service of a big data decision analysis system based on data joint service, and the multi-target decision service process is comprehensively responded. The method has the advantages that the fields are taken as units, the related indexes and data of multiple sources of cross departments, cross industries and cross regions are fused together and are organized in advance according to the special topics, an integrated display view with vivid theme, clear layers and prominent key points is formed, and the integrated display view is visually and three-dimensionally presented to a user. The method supports that a user can dynamically bring related data into a decision topic at any time in the process of developing data exploration around decision problems, and can also perform necessary processing and recombination on the data by means of business data calculation, structure conversion and the like according to needs to fuse into management form data resources meeting the leader decision requirements. And the decision making process is independent, and the data is intelligently expanded. Around decision problems, by adopting conventional means such as data comparison and recombination, statistical analysis and the like or simulation inference based on a model, dynamic 'sand table' rehearsal in a system is realized, a leader is helped to quickly evaluate the influence of various decision schemes on related fields, an optimal scheme is selected, and the purposes of 'sand table' dynamic rehearsal and scheme scientific evaluation are achieved. Through the flexible assembly of data, services, algorithms, models and the like, diversified accurate services which can comprehensively reflect macroscopic overall trends and give consideration to microscopic individual behavior patterns are built and provided, and accurate services are provided through diversified demand matrixes. When a user analyzes a decision under a set thematic frame, the user also supports independently developing targeted exploration analysis such as progressive data association, drilling, comparison and the like according to the decision habit of the user. Meanwhile, the system actively provides related data and exploration paths by combining with the analysis scenes and behavior habits of the user, and supports the user to perform further analysis. In the process, the system comprehensively records the thinking track of the leader decision analysis, more closely combines the human brain with the computer, supports the leader to visually reply the own analysis process at any time, finds more analysis clues, accumulates and summarizes decision analysis experience, and continuously improves the scientific decision level. In the market for viewing specific index data, the product can be disassembled and analyzed according to the characteristics of indexes, and related indexes are intelligently recommended for viewing by combining the internal relation among the indexes. The data resources are intelligently sorted according to the use scenes when being displayed, for example, when the trend is viewed, the data are sorted according to the time sequence, and when the region is viewed, the data are sorted according to the zone codes. The dynamic measurement and calculation of the analysis data are supported, the updated data are automatically brought into the analysis, and the data are more real-time. Aiming at different analysis template characteristics, a plurality of data analysis mining algorithms such as regression analysis, association analysis, cluster analysis, spatial feature analysis and the like and visualization means such as reports, graphs, information graphs and the like are utilized to find and confirm potential association relations among indexes, help leaders to qualitatively or quantitatively deduce the future trend of data, and provide a more efficient solution for decision making or prejudge decision implementation results.

The method realizes a universal and cross-industry scientific decision support mode through a set of universal data service and a set of universal thematic organization architecture. The acquisition of the analysis data can be accomplished by a general data service. In both a big data platform and a traditional database, as long as data are organized according to required data specifications, the data service can complete the acquisition of analysis data in the form of interface call.

The general topic organization structure can analyze the configured topics, the columns, the examples and the configuration information according to a set structure, and the construction work of analyzing the topics is completed. The topic structure is that an analysis topic comprises a plurality of columns, one column comprises a plurality of analysis examples, each analysis example corresponds to an analysis template, and each analysis template corresponds to a set of self-owned analysis methods. The existing system provides a trend analysis template, two time comparison templates, two region comparison templates, a multi-region analysis template based on a GIS, a monitoring evaluation template, a twenty-eight distribution analysis template, a correlation analysis template and an unstructured document template, and gives different analysis contents to the analysis templates through indexes, time dimensions, space dimensions and other information configured by analysis examples, so that cross-industry analysis support is realized, meanwhile, the analysis templates support is expanded, and the continuously increased analysis requirements can be flexibly met.

The data is the basis of all analysis, and the primary work of the invention in use is to complete the data access work. In the access aspect, as shown in fig. 2, an original table fragment import function is designed, various structures and forms of a data table are analyzed, 7 sets of import templates are finally provided, an original storage structure of data is restored to the maximum extent, data access work of a user is greatly simplified, imported information every time can be recorded, and data tracking and updating are facilitated.

After the data is accessed, the configuration work of the special topic can be carried out, as shown in fig. 3, the analysis special topic, the column and the example are respectively and sequentially created, the analysis example is configured, and meanwhile, the checking user of the special topic can be selected, and the special topic authority can be controlled to ensure the data security. The configuration of the analysis example is the core of the analysis topic, and the configuration of the analysis example firstly needs to select an analysis template, and then definitely analyzes information such as a time dimension, a space dimension, a data dimension, a display dimension and the like used by the template, specifically referring to fig. 4.

After the topic configuration is completed, the topics need to be issued, and as shown in fig. 5, the issued topics can be analyzed and used. The authority security of the special topic can be further controlled through the release of the special topic, and only the user distributed with the special topic authority can check the special topic after the release.

In order to further guarantee the security of the data, the authority control of the specific data can also be completed through a data authority configuration function, as shown in fig. 6. The control of the data authority is divided into two parts, namely user policy configuration and data authority policy configuration, wherein the user policy configuration mainly classifies users according to certain conditions to form individual user groups, and can also supplement the conditions of the user groups through an independent user adding function, as shown in fig. 7. The configuration of the data authority strategy is to make different data authority schemes, wherein the authority schemes comprise two parts of structured data and unstructured data, the structured data can control the dimensionality of data such as time, space, indexes, caliber and the like, the unstructured data can be specifically controlled to each file, and finally, a user group is associated with the data authority strategy to complete the configuration work of the user and the data authority, as shown in fig. 8.

After the special topic is released, the user can use the special topic to carry out analysis. As shown in fig. 9, when a user checks a topic, it is necessary to check whether there is a topic checking authority and a data use authority, and after the authority check is passed, the present invention completes construction of the topic according to the analysis topic structure information and the configuration information of the analysis template, and displays the analysis topic to the user. The user can look over each analysis template in the analysis topic, can use the function of template self to carry out conversion and adjustment to template show content, can utilize the associativity between the data to carry out the roaming analysis to can solidify the roaming result, look over when convenient next analysis.

The method has the main innovation points of multivariate data fusion, infinite data roaming and intelligent analysis support.

The multi-metadata fusion is that the traditional relatively solidified data resources are subjected to flexible transformation, dynamic fusion and data joint work to form the social management form data resources which are fused across industries, and finally the multi-objective decision is supported. The flexible transformation is to realize flexible topological transformation, on-demand processing and deep fusion of data by various means such as derivative calculation, summary processing and index feature extraction, generate brand-new management form data resources which are combined in time and space and meet the requirements of follow-up decision support services, and provide guarantee for extending channels and modes of data analysis and fully mining huge values contained in data resources. The dynamic fusion is to fuse the multi-source scattered data through the functional relation definition, provide the data for the user to display, and support the user to add new indexes or take some indexes according to own habits and experiences. Data association refers to the process of tightly grasping the correlation between data and keeping the correlation of the data in a state of being in place all the time and roaming along the data correlation.

The infinite data roaming is a novel data value mining mode which abandons the limitation of traditional data value mining, does not need to preset and analyze roaming paths, establishes free exploration through discovering potential relations among data, roams at any time, switches at any time and positions at any time.

The intelligent analysis support is from the analysis and mining to the understanding of data and then to the value and intelligence of data resources, and the product provides intelligent support for a user in the whole process, so that the speed of converting the data value of the user is greatly increased.

When the method is adopted to carry out big data decision analysis, taking the system supporting industrial energy conservation and consumption reduction analysis as an example, an index system and related grouping information are established and constructed around industry and energy consumption. Index data related to industrial energy consumption is processed according to business rules and algorithms, key business indexes interested by users are configured in the comprehensive billboard, and analysis topics are constructed around industrial energy saving and consumption reduction conditions.

The latest data and the historical data of the key indexes in the field are concerned through the comprehensive billboard, and meanwhile, the related indexes and the related subject contents of the key indexes are checked at any time. For regions and indicators of comparative interest, an attention state may be set, and quick review from my attention.

All index data, index grouping data and unstructured document data of the current system can be quickly checked through resource overview, relevant topics of index analysis can be quickly quoted in progress while indexes are checked, and data of the indexes or document contents can be checked.

The thematic browsing firstly sees a configured thematic frame and an analysis result display of thematic data. When a user views the thematic frame content configured by an analyst, the related indexes can be added in any analysis instance module for extended analysis. For index data containing space dimensionality, other dimensionalities can be switched at any time to check corresponding data. The method supports setting of the attention indexes anytime and anywhere, and index data can be conveniently and quickly checked in the attention module. And in the process of checking the index data, the index measurement units are supported to be switched randomly according to the dimension. According to the specific of the example template, when the example indexes are browsed, the data of the relevant indexes are checked at any time, and according to the analysis purpose, the analysis result of the relevant indexes is better displayed by expanding other analysis templates. The graphical result of each analysis instance can be configured according to the template, other graphical forms can be displayed in an adaptive mode, and the analysis result can be better expressed. And a general calculation function is provided, other calculation index results are rapidly processed according to example indexes, and the difference or relationship among data is better understood. Each time an instance of analysis is performed, the analysis steps are recorded, the full flow of analysis is viewed through a path overview, and quick location and backtracking are achieved. In addition, the information of the color, the shape and the like of the graph can be adjusted according to needs. According to different analysis purposes, the system presets more than ten analysis templates for thematic configuration and use, and free exploratory analysis process calling in the thematic browsing process. Based on a good framework, the thematic analysis template can be customized and expanded according to needs.

Claims (4)

1. A big data decision analysis method based on data joint service comprises the following steps:

1) and (3) metadata planning: planning indexes and grouping metadata information;

2) preparing data:

a) if no import template exists, the frequency and the organization template are selected for data import; if the import template is directly selected, importing the data;

b) the import mode comprises the steps of adding newly-added part data and updated data and covering the original data to finish data import;

3) special subject configuration: one is to complete the configuration for the distribution user after creating the special topic; the other is to finish the configuration of column creation, column index selection and analysis example configuration after creating a special subject;

4) and (3) special subject release:

a) checking whether the configuration is completed;

b) if the configuration is not completed, continuing to check the patent configuration; after the configuration is completed, the special subject is issued;

5) and (3) data authority configuration: sequentially configuring a user strategy and a data authority strategy;

6) and (3) thematic display analysis:

a) checking the viewing authority;

b) obtaining data after the verification is qualified;

c) if the data authority needs to be configured, filtering the data and then analyzing thematic construction; if the data authority does not need to be configured, directly performing analysis topic construction;

d) after the analysis topic display is completed, simultaneously carrying out analysis conversion, analysis roaming and example curing;

e) judging whether the result meets the analysis requirement, and returning to the step a) to restart if the result does not meet the analysis requirement; and finishing if the analysis requirement is met.

2. The big data decision analysis method based on data networking according to claim 1, wherein the method for analyzing the configuration of the instance comprises the following steps:

1) creating an instance and selecting an analysis template;

2) respectively finishing configuration display indexes and configuration space dimensions;

3) completing configuration time dimension after data validity verification is carried out;

4) and respectively carrying out personalized analysis configuration and display content configuration.

3. The big data decision analysis method based on data networking according to claim 1 or 2, wherein the user policy configuration method comprises the following steps:

1) creating a user group;

2) and respectively finishing the group allocation strategy and the group designated user adding.

4. The big data decision analysis method based on data networking according to claim 3, wherein the data authority policy configuration method comprises the following steps:

1) creating a data permission policy;

2) respectively finishing the distribution of the authority of the structured data and the authority of the unstructured data;

3) the user groups are matched.

Images