CN113094621A

CN113094621A - Network public opinion cloud platform

Info

Publication number: CN113094621A
Application number: CN202110444555.5A
Authority: CN
Inventors: 李芳芳; 刘志; 宁肯
Original assignee: Central South University
Current assignee: Central South University
Priority date: 2021-04-23
Filing date: 2021-04-23
Publication date: 2021-07-09
Anticipated expiration: 2041-04-23
Also published as: CN113094621B

Abstract

The invention provides an online public opinion cloud platform, which comprises a key phrase, a combined database management subsystem and a data analysis model exchange subsystem; the key phrase and joint database management subsystem comprises a key phrase collection module, a key phrase distribution module, a public opinion cloud data storage module and a public opinion cloud data distribution module; the data analysis model exchange subsystem comprises a standardized network public opinion data analysis model operation environment module and a network public opinion data analysis model transaction module. The network public opinion data is unified, integrated, exchanged and shared through the key phrase and the combined database management subsystem, so that comprehensive application and scheduling of a network access public opinion system are formed, and the data utilization rate is improved. The data analysis model exchange subsystem effectively improves the types and the number of the network public opinion data analysis models which can be provided by each access public opinion system, so that the final network public opinion users can flexibly configure the required network public opinion analysis models according to actual needs.

Description

Network public opinion cloud platform

Technical Field

The invention relates to the field of network public opinion information or data management, in particular to a network public opinion cloud platform.

Background

At present, the technical means of the public opinion monitoring system mainly focuses on the collection of internet public opinion data, the technical threshold is low, and many software companies, college institutions, traditional media, even public affair companies and the like are launched into the industry in a dispute to provide various public opinion products (hereinafter, public opinion product providers are referred to as "public opinion manufacturers"). However, most public opinion manufacturers have low independent innovation capability and serious homogenization caused by mutual plagiarism. The knowledge of many user units on network public opinion technical services is also in a wrong area, wherein a typical example is server 'stacking theory', and it is considered that sufficient acquisition and storage capacity can be formed through investment of a large amount of equipment and bandwidth, so that so-called 'whole network monitoring' is realized. The phenomena of 'respective administration' and 'small and bad' of a plurality of network public opinion systems are serious, the waste of bandwidth and computing resources is serious, and the repeated collection and storage of the same public opinion data source are serious. The public opinion service industry and the search engine industry have the scale effect of 'natural monopoly', and more data accumulation can be obtained only by continuously inputting a large amount of resources to form a forward feedback cycle. However, the current low-level market competition mode enables public opinion enterprises to have no time and fine tillage, the price war is more and more intense, and the market popularization cost extrudes the investment in the aspects of technical research and development, equipment upgrading and replacement and the like.

At present, the following problems generally exist in the public opinion monitoring industry.

Firstly, public opinion manufacturers are numerous, and at present, the public opinion manufacturers lack standard supervision, and lack of standard analysis indexes as a quantization basis, so that internal consumption of public opinion management is serious.

Secondly, an effective public opinion data collection system is lacked. Public opinion monitoring's course of work mainly relies on public opinion firm to obtain relevant data through all kinds of collection techniques and handles, and the data acquisition result is mastered inside public opinion firm mostly to the public opinion firm of difference has very big possibility to gather same source data simultaneously, leads to redundant collection, has caused the waste of resource.

And thirdly, a data processing and analysis overall coordination mechanism is lacked. The single public opinion manufacturer is limited by factors such as company scale, capital scale, geographical position, customer demand and the like, the data acquisition range is different, each public opinion manufacturer only considers the factor of the manufacturer to acquire data, the data acquisition range is narrow, and the data processing capability is limited by the public opinion manufacturers who provide services. Public opinion supervision and use departments can only use data provided by manufacturers providing public opinion services, lack of data sharing, and the public opinion data requirements cannot be met to the maximum extent.

Disclosure of Invention

The invention aims to provide an Internet public opinion cloud platform which can realize effective information sharing after effective information collection and overall planning processing.

The invention provides a network public opinion cloud platform, which comprises a key phrase and joint database management subsystem and a data analysis model exchange subsystem;

the key phrase and joint database management subsystem comprises a key phrase collection module, a key phrase distribution module, a public opinion cloud data storage module and a public opinion cloud data distribution module;

the key phrase collecting module is used for collecting the collected key phrases and constructing a public opinion cloud platform unified key phrase; the key phrase distribution module is used for distributing a uniform key phrase to each access public sentiment system; the public opinion cloud data storage module is used for collecting and sorting effective network public opinion data to form a public opinion cloud combined database; the public opinion cloud data distribution module is used for distributing corresponding effective network public opinion data to the public opinion system according to the unified key phrase corresponding to each access public opinion system;

the data analysis model exchange subsystem comprises a standardized network public opinion data analysis model operation environment module and a network public opinion data analysis model transaction module;

the standardized network public opinion data analysis model operation environment module is used for constructing a standardized network public opinion data analysis model operation environment; the online public opinion data analysis model trading module is used for constructing online public opinion data analysis model trading market so as to be accessed into a public opinion system to autonomously mark prices of the provided data analysis model, and the accessed public opinion system or a user unit can select any data analysis module.

Further, the keyword group and joint database management subsystem further comprises a specific emergency network public opinion event keyword group module, a specific emergency network public opinion event keyword group of a specific emergency network public opinion event is constructed according to actual requirements, the specific emergency network public opinion event keyword group is pushed to part or all of the access public opinion systems, and the access public opinion systems receive the specific emergency network public opinion event keyword group and then carry out matching and package matching data to share the matching data to the public opinion cloud platform; the priority of the key phrase of the specific emergency network public opinion event is higher than that of the conventional cloud platform unified key phrase, the access public opinion system carries out priority matching after receiving the key phrase of the specific emergency network public opinion event and packs and shares related data to the public opinion cloud platform at the first time.

Furthermore, the keyword group and joint database management subsystem further comprises a public opinion cloud data management module which comprises

Data resource integration, namely managing the cleaning and conversion process of the data set of the public opinion cloud platform, providing a standardized flow and specification for the process, and ensuring that the cleaned and converted data are available;

the public opinion cloud platform comprises a data resource catalog, a query module and a service module, wherein the data resource catalog is used for carrying out data classification on data of the public opinion cloud platform and forming the data resource catalog so as to facilitate data management and query;

data standard management, namely collecting and arranging standard information required by data resource management to form a public opinion cloud combined database into a standard database;

and (3) data security access management, namely establishing hierarchical and weighted management on the process of accessing the public opinion cloud combined database, and ensuring the security of the data in the access process.

When the network public opinion cloud platform is specifically implemented, the network public opinion cloud platform further comprises a standard library subsystem, and the standard library subsystem comprises a key phrase standard module, a public opinion data format standard module, a public opinion data exchange standard module and a network public opinion data analysis model standardized operation environment module.

Specifically, the keyword group standard module restricts the keyword data format uploaded to the keyword group collection module by public sentiment manufacturers; the public opinion data format standard module is used for constraining the effective network public opinion data format collected by the public opinion cloud data storage module; the public opinion data exchange standard module restricts the data exchange process in the data analysis model exchange subsystem; the standardized operation environment module of the network public opinion data analysis model sets up standardized input and output standards.

When the network public opinion cloud platform is specifically implemented, the network public opinion cloud platform further comprises an optimization subsystem for resource allocation of an accessed public opinion system, and the subsystem comprises

The network public opinion data acquisition and recommendation module is used for pushing configuration optimization suggestions of acquisition websites to each accessed public opinion system;

and the network public opinion data analysis model recommendation module is used for pushing optimization suggestions of network public opinion data analysis model research and development investment to each access public opinion system.

When the system is specifically implemented, the network public opinion cloud platform further comprises a subsystem for changing the configuration of a user unit, and the subsystem comprises

The online public opinion technical service self-selection combined supermarket module is used for constructing an online public opinion technical service self-selection combined supermarket, and the online public opinion data acquisition module and the online public opinion data analysis module which are connected into a public opinion system are displayed in a unified way and are used for user units to select and combine;

the network public opinion cloud platform provides a system interface for setting keyword groups for a user unit through a 'cloud public opinion system', so that the keyword groups required by the user unit are obtained;

the network public opinion cloud platform distributes keyword groups required by the cloud public opinion system user units to the related access public opinion system;

the charging module charges effective network public opinion information provided by the relevant access public opinion system to the user unit;

and the network public opinion data acquisition and analysis module is used for charging the network public opinion data analysis module used by the user unit for related access public opinion system.

When the system is specifically implemented, the network public opinion cloud platform further comprises an access operation efficiency evaluation subsystem, and the subsystem comprises

The data acquisition and exchange efficiency module is used for evaluating the data acquisition and exchange efficiency of each access public opinion system, and the data acquisition and exchange efficiency is determined according to the shared data and the received data of the access public opinion system;

the data analysis model efficiency module is used for evaluating the efficiency of the data analysis model of each access public opinion system, and the efficiency of the data analysis model is determined according to the time length and the times of using the data analysis model of the access public opinion system;

the network public opinion cloud platform is accessed to the operation efficiency module, and the access operation efficiency is determined according to the data acquisition exchange efficiency evaluation result and/or the data analysis model efficiency evaluation result of each accessed public opinion system.

Furthermore, the access operation efficiency evaluation subsystem further comprises an elimination module, and the elimination module sets the last elimination rate S1 of data acquisition and exchange efficiency and the last elimination rate S2 of data analysis model efficiency for an access public sentiment system in the public sentiment cloud platform; and calculating a to-be-eliminated access public opinion system set T1 according to S1, calculating a to-be-eliminated access public opinion system set T2 according to S2, and determining an intersection T3 of T1 and T2 according to the time period to determine an eliminated access public opinion system.

Furthermore, the access operation performance evaluation subsystem further comprises a performance monitoring module and an alarm module; the efficiency monitoring module monitors the shared or received data of each accessed public opinion system through the public opinion cloud platform and the data analysis model; the warning module carries out early warning on the public opinion cloud platform and the related access public opinion system according to the abnormal monitoring result of the efficiency monitoring module, finds the access public opinion system which cannot normally share or receive data in time, finds the public opinion data analysis model which cannot normally run on the cloud platform in time, and informs related access public opinion enterprises to modify in time.

The network public opinion data is unified, integrated, exchanged and shared through the key phrase and the combined database management subsystem, so that comprehensive application and scheduling of a network access public opinion system are formed, and the data utilization rate is improved. The data analysis model exchange subsystem can effectively improve the types and the number of the network public opinion analysis models which can be provided by each access public opinion system, and can ensure that the final network public opinion users are not limited to the data analysis model types owned by a certain public opinion company according to the actual requirements, thereby flexibly configuring the required network public opinion analysis models. By constructing the network public opinion data analysis model exchange platform, the utilization rate of analysis models of all access network public opinion companies can be improved.

Drawings

Fig. 1 is an architecture diagram of a network public opinion cloud platform according to a preferred embodiment of the present invention.

FIG. 2 is a flow chart of the key phrase and federated database management subsystem management in the first preferred embodiment.

FIG. 3 is a flow chart of the data analysis model exchange subsystem in accordance with one embodiment.

Fig. 4 is an architecture diagram of a network public opinion cloud platform according to a second preferred embodiment of the present invention.

Fig. 5 is a block diagram of a network public opinion data sharing exchange mechanism in the second preferred embodiment.

Fig. 6 is a flow chart of the operation of the subsystem for optimizing the configuration of the access public opinion system according to the second preferred embodiment of the present invention.

Fig. 7 is an architecture diagram of a network public opinion cloud platform according to a second preferred embodiment of the present invention.

FIG. 8 is a flow chart of the subsystem operation for changing the configuration of the subscriber unit in the third preferred embodiment.

Fig. 9 is a block chain charging flow chart in the third preferred embodiment.

Fig. 10 is an architecture diagram of a network public opinion cloud platform according to a fourth preferred embodiment of the present invention.

Fig. 11 is a flowchart illustrating the operation of the access operation performance evaluation subsystem according to the fourth preferred embodiment.

Detailed Description

Preferred embodiment 1

As shown in fig. 1, the network public opinion cloud platform provided in this embodiment includes two subsystems, namely a keyword group and joint database management subsystem and a data analysis model exchange subsystem.

The key phrase and joint database management subsystem comprises

The public opinion system comprises a public opinion cloud platform, a keyword group collection module, a keyword group identification module and a keyword group identification module, wherein the public opinion cloud platform comprises a plurality of public opinion systems;

the key phrase distribution module is used for distributing cloud platforms to all the access public opinion systems to unify key phrases;

the public opinion cloud database module collects and arranges the effective network public opinion data to form a public opinion cloud combined database; and

and the public opinion cloud data distribution module is used for distributing corresponding effective network public opinion data to the public opinion cloud data according to the cloud platform unified key phrase corresponding to each access public opinion system.

Each individual linkage public opinion cloud platform's individual opinion system (individual opinion company) is on the one hand the sharing of the cloud platform unified key phrase in the conventional or emergency state is realized through the public opinion cloud platform, and on the other hand the direct exchange sharing of the collected effective network public opinion data among the individual linkage individual opinion systems (individual opinion companies) is realized through the public opinion cloud platform. When the public opinion cloud platform collects data submitted by each party, the data needs to be cleaned and converted in an early period, collected data needs to be subjected to processes of data rearrangement, data cleaning and the like, and the collected data are accumulated in the collecting and distributing processes to form a public opinion cloud combined database. The data collection process comprises the following steps:

a. each accessed public opinion access system independently collects network data of each Internet platform;

b. matching network data by each accessed public opinion access system according to the cloud platform unified keyword group;

c. each accessed public opinion system uploads effective network public opinion data which is not needed by itself but is matched to a public opinion cloud platform;

d. public opinion cloud platform distributes uploaded effective network public opinion data to each access public opinion system according to 'the local opinion system (the local opinion company) + the user unit' marked by matching key word group corresponding to the effective network public opinion data submitted by each original access public opinion system;

e. after receiving public opinion data distributed by cloud, each accessed public opinion system analyzes and displays the accessed public opinion data of the relevant user units served by the accessed public opinion system according to the identification of the user unit in the 'local opinion system (the local opinion company) + the user unit';

f. and the public opinion cloud platform is used for carrying out centralized storage on all the forwarded effective network public opinion data to form a data source of the public opinion cloud combined database.

Meanwhile, in the embodiment, the keyword group and the joint database management subsystem are used for packing data required by each profile company in a targeted manner in the forms of keywords, unique identifiers, labels and the like, and the network public opinion cloud data distribution module is used for distributing each access public opinion system, and the data exchange platform is provided with a safety exchange authority to ensure that the data required by each access public opinion system can only be data within the authority range.

Each individual profile system (each profile company) carries out authority hierarchical distribution management to each manufacturer's exchange data process when pushing data through network public profile cloud data distribution module, and each individual profile system (each profile company) sends a data request, and public profile cloud platform audits the data request, and only after the audit is passed can data containing key word groups be packaged, each profile system (each profile company) can only exchange key word group related data containing the authority from public profile cloud platform data exchange platform promptly.

The keyword group and combined database management subsystem further comprises a specific emergency network public opinion event keyword group module, a specific emergency network public opinion event keyword group of a specific emergency network public opinion event is constructed according to actual requirements, the specific emergency network public opinion event keyword group is pushed to part or all of the access public opinion systems, and the access public opinion systems receive the specific emergency network public opinion event keyword group and then carry out matching and package and share matching data to the public opinion cloud platform. The priority of the key phrase of the specific emergency network public opinion event is higher than that of the conventional cloud platform unified key phrase, the access public opinion system carries out priority matching after receiving the key phrase of the specific emergency network public opinion event and packs and shares related data to the public opinion cloud platform at the first time.

The key phrase and combined database management subsystem enables all access public opinion systems to be fully utilized through the key phrase collection module, the key phrase distribution module and the database storage module, network public opinion data are uniformly integrated, exchanged and shared through the public opinion cloud platform, comprehensive application and scheduling of the network access public opinion systems are formed, and the data utilization rate is improved. Meanwhile, the system can realize high timeliness cooperative monitoring on emergency or major emergency.

Furthermore, the keyword group and joint database management subsystem also comprises a public opinion cloud data management module which comprises data resource integration, manages the cleaning and conversion process of the data collection of the public opinion cloud platform, provides a standardized flow and a standard for the process and ensures that the cleaned and converted data are available; the public opinion cloud platform comprises a data resource catalog, a query module and a service module, wherein the data resource catalog is used for carrying out data classification on data of the public opinion cloud platform and forming the data resource catalog so as to facilitate data management and query; data standard management, namely collecting and arranging standard information required by data resource management to form a public opinion cloud combined database into a standard database; data security access management, namely establishing hierarchical and weighted management on the process of accessing the public opinion cloud combined database to ensure the security of the data in the access process; and data label management, namely, a public opinion cloud platform unified marking system is built aiming at key phrases of all access public opinion systems to form unique labels of all access public opinion systems, so that the public opinion cloud platform can conveniently distribute data of all access public opinion systems.

Considering that mass public opinion data such as forums, websites, microblogs, WeChats, traditional media information and the like are gathered by data of a public opinion cloud platform, the data structure is complex, the data types are various, and the data maintenance work will become complex and heavy, so a set of convenient, efficient and safe data management system is established through a public opinion cloud data management module. The public opinion cloud platform provides data management and monitoring functions from different levels of data resource directory service, data resource integration process, data standard management, data security access management, public opinion system data keyword management of each access, data tag library management and the like.

As shown in fig. 2, when the subsystem performs data management, the accessed key phrases of each access public sentiment system are collected, and a cloud platform unified key phrase is constructed; the public opinion cloud platform distributes cloud platform unified key word groups to each access public opinion system, and each access public opinion system matches the data acquired by the access public opinion system according to the cloud platform unified key word groups; each access public opinion system shares effective network public opinion data which is not required by itself and is matched with the cloud platform unified key phrase to the public opinion cloud platform; and the public opinion cloud platform collects effective network public opinion data provided by each access public opinion system, forms a public opinion cloud combined database, and distributes the corresponding public opinion data to the cloud platform unified key word group corresponding to each access public opinion system according to the cloud platform unified key word group corresponding to each access public opinion system.

In the prior art, at present, user units served by each network access public opinion system belong to independent purchasing, and the access public opinion system built in a self-built self-used mode is subjected to a monitoring process of four stages of 'acquisition-comparison-display-analysis'. Each user unit needs to collect the content of the basic similar main websites, so that from the global aspect, most of the technical resources accessed to the public sentiment system are dissipated in the repeated redundant collection process, and the collected pages only configure the key phrases concerned by the unit each time, so that only the interesting useful information of the unit is extracted. A large amount of effective information which is not the unit is directly discarded because the keyword configuration is not carried out; from the global level, due to the lack of a unified integration scheduling platform covering a heterogeneous access public opinion system network operated by a plurality of public opinion companies, technical resources such as computing storage, network bandwidth and the like of the access public opinion system, which are invested and constructed by user units of all levels of all the departments, are in a large internal consumption state.

The public opinion cloud platform collects key word groups required by or provided by each accessed public opinion system, and constructs the cloud platform unified key word group according to a certain rule, and then sends the cloud platform unified key word group to each accessed public opinion system, so that each accessed public opinion system not only matches the key words concerned by itself but also matches the cloud platform unified key word group on the basis of the original data acquisition, thereby simultaneously extracting the data information corresponding to the key words of the cloud platform and the data information corresponding to the cloud platform unified key word group, and sharing the data information corresponding to the cloud platform unified key word group to the public opinion cloud platform, the cloud platform submits effective network public opinion data for each accessed public opinion system, and distributes the corresponding public opinion data to the accessed public opinion systems according to the cloud platform unified key word groups corresponding to the other accessed public opinion systems at the first time, therefore, other public opinion accessing systems can receive required data information from the public opinion cloud platform conveniently. The network public opinion data is uniformly integrated, exchanged and shared through a public opinion cloud platform and a key phrase management method, so that the comprehensive application and scheduling of the network access public opinion system are formed, and the data utilization rate is improved. Meanwhile, a public opinion cloud combined database is formed according to effective network public opinion data provided by each access public opinion system, and all data stored in the public opinion cloud combined database are effective or useful data, so that the public opinion value is achieved.

Each accessed public opinion system determines a required key phrase and marks the key phrase and then shares the key phrase to a public opinion cloud platform, and the public opinion cloud platform performs compliance check on the key phrase; if the rule is combined, the key phrases are collected, duplication removal and integration are carried out on the key phrases, and a cloud platform unified key phrase is constructed; and if the rule is not matched, feeding back the rule to a corresponding access public opinion system and requiring the rule to be adjusted or modified to be matched.

The keyword group and combined database management subsystem carries out compliance check on the keyword groups provided by the accessed public opinion systems, and the different key group forms of the heterogeneous accessed public opinion systems can be effectively avoided through the compliance check on the accessed public opinion systems. And the key phrases are deduplicated and integrated, and the unified key phrase of the cloud platform is determined, so that all the access public opinion systems are relatively unified, and the key phrases of all the clients served by all the access public opinion companies are included, so that the comprehensive coverage on data matching is ensured, and the running accuracy and the running efficiency of the public opinion cloud platform are ensured.

The marking mode of each key phrase required by accessing the public sentiment system is as follows: the marked content comprises the company name + the user name of the public opinion access system, and the marked form is a unique identifier. For example, the labeling rule may be preferably XnYm, where n in Xn is a unique code of the public opinion cloud platform for accessing the public opinion system, and m in Ym is a unique code of the public opinion cloud platform for accessing the user, such as X0001Y001 represents a 001-th user of a 0001-th access public opinion system for accessing the public opinion cloud platform.

The access public opinion system can also send data requests according to self requirements, and after the key phrase and the combined database management subsystem check the data requests, the public opinion data containing the keywords required by the access public opinion system are packaged and sent. The access public opinion system can also actively send related data requests to the key word group and the combined database management subsystem according to the change of the self demand and obtain related public opinion data after the public opinion cloud platform is audited, so that the access public opinion system can quickly and conveniently obtain various public opinion data, and the service quality and the service efficiency of the access public opinion system are favorably provided.

The data analysis model exchange subsystem in the embodiment comprises a standardized network public opinion data analysis model operation environment module, a standardized network public opinion data analysis model operation environment is constructed, and standardized operation environments are provided for various network public opinion data analysis models of various access public opinion systems accessed to a public opinion cloud platform; each access public opinion system standardized data analysis model ensures that the provided data analysis model can normally operate in the standardized operation environment of the public opinion cloud platform; the network public opinion data analysis model trading module is used for constructing a network public opinion data analysis model trading market and accessing a public opinion system to autonomously mark prices of the provided data analysis model according to rules; and accessing a public opinion system or a user unit to purchase or use any data analysis module from a public opinion cloud platform according to the self demand. And the data analysis model exchange subsystem shares the analysis models of the heterogeneous network access public opinion system to a platform for exchange, so as to form comprehensive application and scheduling of the network access public opinion system based on the network public opinion analysis model as a drive, and improve the utilization rate of the analysis models.

As shown in fig. 3, the method for exchanging data by using the data analysis model exchange subsystem includes:

establishing a standardized operation environment of an online public opinion data analysis model, and providing a standardized operation environment for various online public opinion data analysis models of all access public opinion systems accessing a public opinion cloud platform;

each access public opinion system standardized data analysis model ensures that the provided data analysis model can normally operate in the standardized operation environment of the public opinion cloud platform;

establishing an online public opinion data analysis model transaction market, and accessing a public opinion system to autonomously mark prices of the provided data analysis model according to rules; and

and accessing a public opinion system or a user unit to purchase or use any data analysis module from a public opinion cloud platform according to the self demand.

The operating environment supports mainstream technical frameworks and programming languages and is provided with an access interface capable of accessing a public opinion cloud platform database. And each access public opinion system reconstructs and reforms each network public opinion data analysis model which needs to be submitted to the network public opinion cloud platform according to the technical requirements of the standardized operation environment, so that the access public opinion system meets the requirements of operation on the standardized operation environment of the cloud platform.

In this embodiment, public opinion cloud platform is used to bring the network public opinion data analysis models of each accessed public opinion system (each opinion company) into a standardized operation environment, each accessed public opinion system (each opinion company) reconstructs and transforms each network public opinion data analysis model which needs to be submitted to the network public opinion cloud platform, so that the accessed public opinion system meets the requirement of operation on the standardized operation environment of the cloud platform, and migrates to the standardized operation environment according to the relevant standards, thereby realizing sharing, exchange use and transaction of the data analysis models of different accessed public opinion systems (each opinion company) on the 'public opinion cloud' platform; each access public opinion system (each opinion company) can self-selectively buy and configure any data analysis model in the public opinion cloud platform according to the customer requirements or the operation development requirements thereof, thereby providing diversified network public opinion data analysis model options which are not limited to the analysis model developed by the user and are more flexible and more adaptive to the user requirements for the customer.

In the specific using process, when any data analysis module is purchased or used by an access public sentiment system or a user unit, a request service is provided for a public sentiment cloud platform, the public sentiment cloud platform records the request, the public sentiment cloud platform loads a data analysis model to a standardized operation environment after recording, analyzes specified data accessed to the public sentiment system or the user unit, and then pushes a data analysis result to the access public sentiment system or the user unit.

The public opinion cloud platform constructs a basic data analysis module according to the requirements of accessing a public opinion system or a user unit.

In this embodiment, the data analysis modules in the internet public opinion cloud platform are divided into three categories, the first category is a data analysis module that the public opinion cloud platform establishes a foundation according to the requirements of the access public opinion systems or user units, the second category is various internet public opinion data analysis models that each access public opinion system (which is an opinion company) shares with the public opinion cloud platform, and the third category is an internet public opinion data analysis model that related colleges, scientific research institutions or professionals independently develop and upload to the public opinion cloud platform. The embodiment forms the comprehensive application and scheduling of the network access public opinion system based on the network public opinion analysis model as the drive, and improves the utilization rate of the analysis model.

The pricing rules include pricing according to the number of uses, the length of use, or the amount of data analyzed by the data analysis model. The method further comprises the step of storing transaction data of each data analysis model based on a block chain technology, and the public opinion cloud platform supervises transaction behaviors. The public opinion cloud platform establishes a uniform data analysis model transaction mode, transaction behaviors are uniformly supervised by the platform, and transaction data are stored based on a block chain technology, so that transparent credibility of each access public opinion system (an inventory company) is ensured.

Preferred embodiments II,

As shown in fig. 4, in the present embodiment, a standard library subsystem and a subsystem for optimizing the configuration of the access public opinion system are added on the basis of the first preferred embodiment.

The standard library subsystem comprises a keyword group standard module, a public opinion data format standard module, a public opinion data exchange standard module and an online public opinion data analysis model standard operation environment module.

The keyword group standard module restricts the keyword data format uploaded to the keyword group collection module by public sentiment manufacturers; the public opinion data format standard module is used for constraining the effective network public opinion data format collected by the public opinion cloud data storage module; the public opinion data exchange standard module restricts the data exchange process in the data analysis model exchange subsystem; the standardized operation environment module of the network public opinion data analysis model sets up standardized input and output standards.

The network public opinion data sharing and exchanging mechanism mainly comprises public opinion keyword standard, public opinion data format standard, public opinion data exchanging standard and the like. Meanwhile, after the public opinion data is standardized, a public opinion core database is further constructed, so that a channel is provided for sharing and exchanging the public opinion data between users and between the platforms, as shown in fig. 5.

Public opinion keyword criteria. Because the public opinion manufacturers collect more keyword data sources and types and a large amount of repeated data exists, the platform makes corresponding network public opinion keyword standards. For example, the keyword "security + (greedy officer | counterfeit | academic forgery)" means "security greedy officer", "security counterfeit", and "security academic forgery". By the method, a plurality of keywords can be connected into a character string, and the relationship among the keywords can be intuitively understood. Each manufacturer firstly needs to perform word-removing and repeated data-removing processing on respective keyword data, and then changes the format of the keyword data to meet the standard requirements of the platform before uploading the keyword data to the platform, so that the quality and the standardization level of the data are improved.

And (5) public opinion data format standard. Aiming at the problem that the public opinion data formats provided by public opinion manufacturers are not uniform, corresponding format standards are formulated to restrict the formats of the network public opinion data, as shown in table 1.

TABLE 1 network public opinion data format standard requirement field

Before uploading the public opinion data to the platform, each manufacturer should adjust the format of the public opinion data according to the standard, and splice the corresponding content in the field and the public opinion data, and upload the public opinion data to the platform in the form of JSON or XML. By the method, the platform can store and manage the public sentiment data conveniently, and the user can know the related information of the public sentiment data clearly and intuitively.

Public opinion data exchange standard. In the standard, after a user submits data to a platform according to a data format standard, the platform calculates the relevancy and similarity index of the keywords according to the data provided by the user, and verifies the data according to the relevancy and similarity index. After the audit is passed, the platform stores the analysis result and the original data into the database. When a user needs to acquire public opinion data by using the platform, the user only needs to send the keywords to the platform, and then the platform forwards the request to other public opinion manufacturers. After other public opinion manufacturers upload corresponding data, the platform can audit the data uploaded by the manufacturers at first, and after the audit is passed, the data provided by the public opinion manufacturers, the corresponding data in the platform public opinion core database and the data analysis result of the platform can be returned to the user together, so that the requirements of the user can be met to the maximum extent, and convenience can be provided for the user to perform data screening and analysis.

As shown in fig. 6, when the subsystem for optimizing configuration of the access public opinion system recommends an optimized configuration for the internet public opinion data acquisition policy of each access public opinion system, on one hand, by analyzing the internet public opinion data mutually forwarded by the public opinion cloud platform of each access public opinion system within a period of time, an optimized configuration suggestion is proposed for the data acquisition policy of each access public opinion system, and the cooperative matching of each access public opinion system in the website data acquisition aspect is promoted, so that the overall internet public opinion data acquisition efficiency is improved; on the other hand, the network public opinion data analysis models of all public opinion systems of the access public opinion cloud platform are collected, the network public opinion data analysis models with low utilization rate are screened out from the network public opinion data analysis models with similar functions and more submitted numbers and are fed back to the related access public opinion system, and research and development investment of the related models are suggested to be adjusted.

In the redundancy removal process of the collected network public opinion data acquisition model, the range of the acquisition website is negotiated and cooperated to reduce redundancy. And if the acquisition frequency of the high-configuration-rate acquisition website is higher than the data updating frequency, pushing a configuration reduction suggestion to a public opinion company of the related access public opinion system, and sending the acquisition website of the low-configuration rate to the public opinion company of the related access public opinion system by analogy to increase the configuration into the acquisition website.

Wherein, the average configuration rate is the number of public sentiment systems configured to collect the target websites/the sum of the target websites configured to collect.

In the process of reducing redundancy, the public opinion cloud platform can analyze the range and the acquisition frequency of each acquisition website accessed to the local emotional system from effective public opinion data for a period of time. Screening out some websites which are set as the network public opinion data acquisition range by most of the individual opinion systems, and completely abandoning the websites which are set to be partially too high in the overall acquisition frequency due to the network public opinion data exchange sharing mechanism of the individual access opinion systems, without worrying about the problem of missing acquisition of important information of the individual access opinion systems due to abandoning acquisition. Meanwhile, the public opinion cloud platform can analyze acquisition websites with relatively few configurations of all websites, divide the less acquired websites according to the configuration conditions of the existing acquisition websites of all access history systems, and recommend the relevant access history systems to configure the acquisition websites into the acquired websites. Therefore, through recommending relevant access profile systems, relevant collected websites are reduced and increased, so that the overall collection efficiency of all access profile systems is optimized on the whole without sacrificing the collection range and effectiveness.

In the process of removing redundancy of the collected network public opinion data analysis models, collecting the adoption rate of the network public opinion data analysis models shared on the public opinion cloud platform by user units or all access public opinion systems, screening the network public opinion data analysis models with the adoption rate lower than the average adoption rate 1/2, and feeding back the network public opinion data analysis models to the public opinion systems.

The network public opinion data acquisition and recommendation module is used for pushing configuration optimization suggestions of acquisition websites to each accessed public opinion system; some websites set as network public opinion data acquisition ranges by most public opinion systems and acquisition websites with relatively few configurations of all public opinion systems are screened out. Therefore, related collection websites are reduced and increased by recommending related access public sentiment systems.

And the network public opinion data analysis model recommendation module is used for pushing optimization suggestions of network public opinion data analysis model research and development investment to each access public opinion system. And each access public opinion system submits the network public opinion data analysis model which is researched and developed by each access public opinion system to the system. The module analyzes the repeated conditions of the types and the quantity of the analysis models on the whole for the adoption condition of the online public opinion data analysis models submitted by each access public opinion system to the online public opinion data analysis models submitted by each access public opinion system in a period of time by the user units submitted to the online public opinion data analysis models on the public opinion cloud platform. And selecting the network public opinion data analysis models with lower user unit adoption rate, and recommending and submitting the model public opinion data analysis models to reduce the research and development investment of the data analysis models. By coordinating all the access public opinion systems to reduce repeated investment and respectively centralizing resources to develop cooperative technical attack, the investment efficiency of all network data model research and development resources of the access public opinion systems is integrally optimized in an advantage complementation mode, and the diversity, the advancement and the user unit satisfaction of the network public opinion data analysis model on the whole are not sacrificed.

Preferred embodiments III,

As shown in fig. 7, in the present embodiment, a subsystem for changing the configuration of the user unit is added to the second preferred embodiment.

The subsystem is composed of the following parts.

The online public opinion technical service self-selection combination supermarket module displays the online public opinion data acquisition module and the online public opinion data analysis module of each access public opinion system in a unified way and provides user units for selection and combination. The module displays the network public sentiment data acquisition capacity (including acquisition range, acquisition frequency and other indexes) and the network public sentiment data analysis module (including module introduction, module price and the like) of each access public sentiment system in a unified way. The user unit can combine the data acquisition capacity and the data analysis module in a menu mode according to the self system construction budget and the actual service requirement. The combination is assembled according to the selection of the user unit through a unified standardized interface, and the user unit is supported to adjust the combination of the originally selected data acquisition capability and the data analysis module according to the number of the self service requirement at the later stage of using the self-selection combined cloud public opinion system, wherein the adjustment includes addition, deletion or replacement. Therefore, the requirements of the user units are met, the implementation is not limited by the range provided by a certain access public opinion system, but can be selected from the sum of data technology resources provided by all the access public opinion systems, and the flexibility of the user units in the aspect of dealing with the change of self service requirements is improved.

The cloud public opinion system user unit keyword group aggregation module is characterized in that an internet public opinion cloud platform provides a system interface for setting keyword groups for a user unit through a cloud public opinion system, so that the keyword groups required by the user unit are obtained.

The cloud public opinion system user unit keyword group distribution module is used for distributing keyword groups required by a 'cloud public opinion system' user unit to a relevant access public opinion system (selected by the user unit to become a provider of an online public opinion data acquisition module) by an online public opinion cloud platform.

The cloud public opinion system keyword group distribution module is used for distributing the network public opinion cloud platform unified keyword group to a related access public opinion system (selected by a user unit to become a network public opinion data acquisition module provider).

The standardized network public opinion data analysis model operation environment module is used for constructing a standardized network public opinion data analysis model operation environment and providing a standardized operation environment for various network public opinion data analysis models accessed to a public opinion system; each access public opinion system standardized data analysis model ensures that the provided data analysis model can normally operate in the standardized operation environment of the public opinion cloud platform;

the online public opinion data analysis model transaction module is used for constructing an online public opinion data analysis model transaction market, the access public opinion system carries out autonomous pricing on the provided data analysis model according to rules, and the access public opinion system or a user unit purchases or uses any data analysis module from a public opinion cloud platform according to self requirements.

And the charging module is used for charging the effective network public opinion information provided by the relevant access public opinion system for the user unit.

After the system is put into use, a self-selection combined supermarket of network public opinion technical service facing to a user unit is constructed, a unified display platform is provided for each public opinion company, key word groups set by the user unit are distributed to a relevant access public opinion system for collection and matching, the matched network public opinion data of the relevant access public opinion system is collected and then subjected to rearrangement processing, and the data is provided for the user unit through a cloud public opinion system;

by constructing the cloud public opinion system optimized based on user demands, on one hand, each access public opinion system (public opinion company) provides network public opinion data acquisition and network public opinion data analysis capability for user units through the cloud public opinion system, and the cost investment of each access public opinion system (public opinion company) in the aspect of marketing can be effectively reduced; on the other hand, each user unit can change the configuration of the network public opinion data acquisition and the network public opinion data analysis according to the service requirements of the user unit and the advantages and disadvantages of the network public opinion data acquisition and network public opinion data analysis capabilities provided by different manufacturers in the using process of the cloud public opinion system, so that the flexible adaptability of the cross public opinion system (public opinion company) is realized.

As shown in fig. 8, the subsystem is put into operation, and the specific steps are as follows:

the method comprises the steps that firstly, an online public opinion cloud platform collects online public opinion data acquisition modules and online public opinion data analysis modules of all access public opinion systems, and after all collected modules are combined and adjusted, an online public opinion technical capability self-selection supermarket which can be used for user units to self-select and combine is constructed.

In this step, the public opinion cloud platform displays each access to the internet public opinion data acquisition module (the acquisition capability includes acquisition range, acquisition frequency and other indexes) and the internet public opinion data analysis module (including module introduction, module price and the like) of the company of storing opinions, and the modules are displayed in a unified mode. The user can combine the data acquisition capability and the data analysis module in a 'menu type' mode according to specific conditions such as self system construction budget, actual business requirements and the like.

The public opinion cloud platform assembles an interface according to the selection combination of the user by a unified and standardized cloud public opinion system, and supports the user to select the combination by oneself.

In the step, a network public opinion cloud platform establishes a standardized network public opinion data analysis model operation environment, and provides a standardized operation environment for various network public opinion data analysis models of public opinion systems accessed to the public opinion cloud platform; and each public opinion system standardized data analysis model ensures that the provided data analysis model can normally operate in the standardized operation environment of the public opinion cloud platform. The operating environment supports mainstream technical frameworks and programming languages and is provided with an access interface capable of accessing a public opinion cloud platform database.

Secondly, after a user unit autonomously selects and combines different modules in the 'network public opinion technical ability self-selection supermarket', a 'cloud public opinion system' is generated from the user unit by the network public opinion cloud platform; after the user unit selects the combination type, the combination of the originally selected data acquisition capacity and the data analysis module is adjusted, including addition, deletion or replacement, according to the change of the self service requirement, so that the continuous adaptability to the user unit requirement is met. When the user unit selects the data acquisition module and the data analysis module, the user unit is not limited to the range that the entire situation company can provide, but can select from the total of all the data technical resources that the entire situation company can provide.

And thirdly, the online public opinion cloud platform collects keyword groups set by user units through a cloud public opinion system and distributes the keyword groups to related access public opinion systems (the keyword groups are selected by the user units to become online public opinion data acquisition module providers).

And fourthly, accessing a public opinion system to collect and match keyword groups set by related user units, defining successfully matched network public opinion data as effective network public opinion data, and pushing the effective network public opinion data to a network public opinion cloud platform.

And fifthly, in the step, the network public opinion cloud platform receives effective network public opinion data pushed by the cloud public opinion system user from the relevant access public opinion system, and repeats the effective network public opinion data by taking the link of the public opinion page as a unit. For the same link, the link of each profile company is pushed by a plurality of profile companies, the profile company which pushes the link firstly is marked with effective collection of the link, the effective collection is displayed on a cloud public profile system interface of a related user, and other profile companies which push the link are marked with ineffective collection of the link.

And sixthly, in the service period of the cloud public opinion system user, carrying out statistical charging on effective network public opinion information provided for user units by the relevant access public opinion system.

And seventhly, the user units automatically combine and select related network public opinion data analysis modules in the network public opinion data analysis model trading market of the network public opinion cloud platform.

And eighthly, the network public opinion cloud platform operates the relevant network public opinion data analysis module selected by the user unit through the operation environment of the standardized network public opinion data analysis model, and the cloud public opinion system provides a uniform system interface for the user unit to use.

And ninthly, carrying out statistical charging on the network public opinion data analysis module which is provided by the relevant access public opinion system for the user unit to use.

And tenth, changing the configuration of the network public opinion data acquisition and the network public opinion data analysis by the user unit according to the service requirements of the user unit and the advantages and disadvantages of the network public opinion data acquisition and network public opinion data analysis capabilities provided by different manufacturers.

And a public opinion cloud platform, a cloud platform fund pool accessed between public opinion systems based on a block chain and a settlement system can also be added. The system comprises the following modules:

a user platform interface. A user unit-oriented graphic system interface provided by a public opinion cloud platform, a user using the cloud public opinion system does not need to care about the operation mechanism of each item of the platform, only needs to care about the public opinion information or public opinion analysis model required by the user, and selects a corresponding operation manufacturer providing services through the interface, so that the user is flexible and configurable. The public opinion cloud platform provides public opinion information and an analysis model for users, focuses on the public opinion information and analysis results, and meets the requirement of the users for obtaining public opinions quickly and with high quality. Also, the transactions of the platform will be supported by the billing system.

A charging system based on block chains. The transaction data in the public opinion cloud platform is stored through a block chain, and the transaction data is subjected to statistics of different manufacturers through a platform background, and finally provides effective public opinion data or data of an analysis model for a certain user, and is stored based on the data. As shown in fig. 9, the block chain charging process is as follows: the user side collects charging information according to the transaction details of the user; the client encrypts the data and transmits the data to the cloud server; the cloud end records the charging information into the block chain; and the intelligent contract of the block chain automatically collects the fee.

The preferred embodiments four,

As shown in fig. 9, in the present embodiment, an access operation performance evaluation subsystem is added on the basis of the third preferred embodiment, and includes a data acquisition and exchange performance module for evaluating data acquisition and exchange performance of each access public opinion system, where the data acquisition and exchange performance is determined according to the shared data and the received data of the access public opinion system; the data analysis model efficiency module is used for evaluating the efficiency of the data analysis model of each access public opinion system, and the efficiency of the data analysis model is determined according to the time length and the times of using the data analysis model of the access public opinion system; the network public opinion cloud platform is accessed to the operation efficiency module, and the access operation efficiency is determined according to the data acquisition exchange efficiency evaluation result and/or the data analysis model efficiency evaluation result of each accessed public opinion system.

The access operation efficiency evaluation subsystem further comprises an elimination module, and the elimination module sets the last elimination rate S1 of data acquisition and exchange efficiency and the last elimination rate S2 of data analysis model efficiency for an access public sentiment system in the public sentiment cloud platform; and calculating a to-be-eliminated access public opinion system set T1 according to S1, calculating a to-be-eliminated access public opinion system set T2 according to S2, and determining an intersection T3 of T1 and T2 according to the time period to determine an eliminated access public opinion system.

The access operation efficiency evaluation subsystem also comprises an efficiency monitoring module and an alarm module; the efficiency monitoring module monitors the shared or received data of each accessed public opinion system through the public opinion cloud platform and the data analysis model; the warning module carries out early warning on the public opinion cloud platform and the related access public opinion system according to the abnormal monitoring result of the efficiency monitoring module, finds the access public opinion system which cannot normally share or receive data in time, finds the public opinion data analysis model which cannot normally run on the cloud platform in time, and informs related access public opinion enterprises to modify in time.

As shown in fig. 10, when the subsystem performs evaluation on the access operation performance of the network public opinion cloud platform, the method is as follows:

evaluating data acquisition and exchange efficiency of a heterogeneous access public opinion system of different public opinion enterprises in a public opinion cloud platform, wherein the data acquisition and exchange efficiency is determined according to shared data and received data of the access public opinion system;

evaluating the effectiveness of a data analysis model of each access public opinion system in a public opinion cloud platform, wherein the effectiveness of the data analysis model is determined according to the time length and the times of using the data analysis model of the access public opinion system;

and determining the cloud platform access operation efficiency of each access public opinion system according to the data acquisition exchange efficiency evaluation result and/or the data analysis model efficiency evaluation result of each access public opinion system.

In this embodiment, the public opinion cloud platform with a plurality of access public opinion systems can be accessed to make efficiency evaluation on each access public opinion system through the above evaluation method, and the access public opinion systems with lower operation efficiency in a clearing period are introduced and accessed to some access public opinion systems with higher operation efficiency on the market, so as to improve the operation efficiency of the whole public opinion cloud platform and form a good operation of the public opinion cloud platform. Meanwhile, the evaluation can be carried out only by adopting the data acquisition and exchange efficiency evaluation results of all the access public opinion systems, the evaluation can also be carried out only by adopting the data analysis model efficiency evaluation results of all the access public opinion systems, and the comprehensive evaluation can also be carried out by adopting the data acquisition and exchange efficiency evaluation results and the data analysis model efficiency evaluation results of all the access public opinion systems.

The data acquisition exchange efficiency is the shared data of the access public opinion system: the received data of the access public opinion system is the data of other access public opinion systems adopted or used by the access public opinion system from the public opinion cloud platform, and the shared data is the data which is shared by the access public opinion system to the public opinion cloud platform and is adopted or used by other access public opinion systems.

In this embodiment, the performance of the data analysis model of the nth access opinion system is An, the shared data of the access opinion system is Bn, and the received data of the access opinion system is Cn; the access public sentiment system receives data from the public sentiment cloud platform in a large amount, and when the data shared by the access public sentiment system is received or used more, Cn is larger, for the public sentiment cloud platform, the more shared data of the access public sentiment system is used, the more the access public sentiment system is better in efficiency, and when a certain access public sentiment system receives data from the platform in a large amount and the data shared by the access public sentiment system is not adopted by other access public sentiment systems, the public sentiment performance is relatively poorer, so for An/Bn, when the value of An of the access public sentiment system is larger, the network public sentiment data performance is better.

The performance of the data analysis model is the use duration and the use times of the data analysis model which is shared by the public opinion access system to the public opinion cloud platform and adopted or used by other public opinion access systems.

In this embodiment, each access public opinion system in the access public opinion cloud platform shares the data analysis model while sharing data, so the performance of the data analysis model is also an important index for evaluating the performance of the access public opinion system, and the longer the time or the more times the data analysis model shared by the access public opinion system is used, the better the performance of the data analysis model of the access public opinion system.

Establishing a rejection mechanism according to the evaluation result of the cloud platform access operation efficiency, setting the last rejection rate S1 of the data acquisition and exchange efficiency and the last rejection rate S2 of the data analysis model efficiency for an access public sentiment system in the public sentiment cloud platform. As a preferred embodiment, the present embodiment calculates a to-be-rejected accessing public opinion system set T1 according to S1, calculates a to-be-rejected accessing public opinion system set T2 according to S2, and finds an intersection T3 between T1 and T2 to determine a rejected accessing public opinion system at a certain time period. The end elimination rates are respectively set aiming at the two efficiencies, so that the public opinion access systems are stimulated to improve the data quality, the data resource goodness of the public opinion cloud platform is ensured, the benign competition among different public opinion access systems is promoted on the basis of comprehensively, effectively and accurately evaluating the cloud platform access operation efficiencies of the public opinion systems, and the data resources on the public opinion cloud platform are efficiently utilized.

In summary, the present embodiment evaluates the cloud platform access operation performance of the public opinion access system from two aspects, namely, the data collection exchange performance and the data analysis model performance. The embodiment evaluates the data acquisition and exchange efficiency of the access public opinion system based on the proportion of the shared data and the received data of the access public opinion system; evaluating the efficiency of the data analysis model accessed to the public opinion system based on the used time and the used times of the data analysis model; the adoption amount of the shared data by the user can reflect the effectiveness of the shared data, and the higher the adoption amount is, the higher the relevance of the data is; the service duration and times of the analysis model by the user can reflect the quality degree of the analysis model, the service duration of the analysis model with better efficiency is longer or the times of the analysis model user is more, and otherwise, the service duration is insufficient or the times of the analysis model user is insufficient. By adopting the method, the cloud platform access operation efficiency of each public opinion system can be comprehensively, effectively and accurately evaluated, the advantages and disadvantages of each public opinion system are visually reflected by the cloud platform access operation efficiency evaluation mode, the benign competition among different access public opinion systems is promoted, and the method is also beneficial to better screening of users in different access public opinion systems. On the other hand, the method and the system make efficient use of data resources on the public opinion cloud platform.

Claims

1. The utility model provides an online public opinion cloud platform which characterized in that: the platform comprises a key phrase and joint database management subsystem and a data analysis model exchange subsystem;

2. The internet public opinion cloud platform of claim 1, characterized in that: the keyword group and combined database management subsystem further comprises a specific emergency network public opinion event keyword group module, a specific emergency network public opinion event keyword group of a specific emergency network public opinion event is constructed according to actual requirements, the specific emergency network public opinion event keyword group is pushed to part or all of access public opinion systems, the access public opinion systems receive the specific emergency network public opinion event keyword group and then carry out matching, and matching data are packaged and shared to a public opinion cloud platform; the priority of the key phrase of the specific emergency network public opinion event is higher than that of the conventional cloud platform unified key phrase, the access public opinion system carries out priority matching after receiving the key phrase of the specific emergency network public opinion event and packs and shares related data to the public opinion cloud platform at the first time.

3. The internet public opinion cloud platform of claim 2, characterized in that: the key phrase and joint database management subsystem also comprises a public opinion cloud data management module which comprises

4. The internet public opinion cloud platform of claim 1, characterized in that: the platform also comprises a standard library subsystem, wherein the standard library subsystem comprises a keyword group standard module, a public opinion data format standard module, a public opinion data exchange standard module and an online public opinion data analysis model standardized operation environment module.

5. The internet public opinion cloud platform of claim 4, characterized in that: the keyword group standard module restricts the keyword data format uploaded to the keyword group collection module by public sentiment manufacturers; the public opinion data format standard module is used for constraining the effective network public opinion data format collected by the public opinion cloud data storage module; the public opinion data exchange standard module restricts the data exchange process in the data analysis model exchange subsystem; the standardized operation environment module of the network public opinion data analysis model sets up standardized input and output standards.

6. The internet public opinion cloud platform of claim 4, characterized in that: the platform also comprises an optimization subsystem for resource allocation of the accessed public sentiment system, and the subsystem comprises

7. The internet public opinion cloud platform of claim 6, characterized in that: the platform also comprises a subsystem for changing the configuration of the user unit, wherein the subsystem comprises

the network public opinion cloud platform provides a system interface for setting the keyword group for the user unit through a cloud public opinion system so as to obtain the keyword group required by the user unit;

8. The internet public opinion cloud platform of claim 1 or 7, characterized in that: the platform also includes an access operation performance evaluation subsystem, which includes

9. The internet public opinion cloud platform of claim 8, characterized in that: the access operation efficiency evaluation subsystem further comprises an elimination module, and the elimination module sets the last elimination rate S1 of data acquisition and exchange efficiency and the last elimination rate S2 of data analysis model efficiency for an access public sentiment system in the public sentiment cloud platform; and calculating a to-be-eliminated access public opinion system set T1 according to S1, calculating a to-be-eliminated access public opinion system set T2 according to S2, and determining an intersection T3 of T1 and T2 according to the time period to determine an eliminated access public opinion system.

10. The internet public opinion cloud platform of claim 8, characterized in that: the access operation efficiency evaluation subsystem further comprises an efficiency monitoring module and an alarm module; the efficiency monitoring module monitors the shared or received data of each accessed public opinion system through the public opinion cloud platform and the data analysis model; the warning module carries out early warning on the public opinion cloud platform and the related access public opinion system according to the abnormal monitoring result of the efficiency monitoring module, finds the access public opinion system which cannot normally share or receive data in time, finds the public opinion data analysis model which cannot normally run on the cloud platform in time, and informs related access public opinion enterprises to modify in time.