Public opinion early warning method and device, equipment and storage medium
A technology of public opinion and early warning information, applied in the field of communication, can solve the problems of poor accuracy of public opinion early warning, inability to obtain public opinion early warning results, misjudgment of public opinion, etc.
Pending Publication Date: 2021-09-14
CHINA UNITED NETWORK COMM GRP CO LTD
0 Cites 0 Cited by
AI-Extracted Technical Summary
Problems solved by technology
 This application provides a public opinion early warning method, device, equipment and storage medium, so as to solve the technical problems that the existing technology may...
Here, the embodiment of the present application does not use word frequency as feature to carry out intention identification, has reduced the cost of keyword storehouse maintenance, by setting up the intention labeling training data that pre-marked in the intention labeling training database, can establish before intention identification The intent recognition model is preset, so that intent recognition is performed based on this model, which reduces the cost of public opinion analysis and early warning, and improves efficiency.
 In an optional emb...
The invention provides a public opinion early warning method and device, equipment and a storage medium, and the method comprises the steps: obtaining telephone traffic data, carrying out the data preprocessing of the telephone traffic data, and obtaining effective telephone traffic data; according to a preset intention recognition model, performing intention recognition on the effective traffic data to obtain intention traffic data; according to a preset prediction telephone traffic regression model, carrying out the normalization processing on the intention telephone traffic data to obtain first normalized intention data, wherein the preset prediction telephone traffic regression model is obtained through training according to historical telephone traffic data and historical public opinion data; if the first normalized intention data is greater than the preset intention threshold, determining that the public opinion is abnormal, and generating public opinion early warning information, thereby improving the accuracy of public opinion early warning.
Semantic analysisCharacter and pattern recognition +2
EngineeringReal-time computing +3
- Experimental program(1)
 The exemplary embodiment will be described in detail herein, and examples thereof are shown in the drawings. The following description is related to the drawings, unless otherwise indicated, the same numbers in the drawings represent the same or similar elements. The embodiments described in the exemplary embodiments described below do not represent all embodiments consistent with the present disclosure. Instead, they are only examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.
 The specification and claims and the drawings described above herein, the term "first", "second", "third" and "fourth" and the like (if present) is used for distinguishing between similar objects, without having to use in describing a particular sequential or priorities. It is to be understood that the data such as use can be interchangeable in appropriate, so that the embodiments of the present application described herein can be implemented in the order other than those illustrated or described herein. Moreover, the terms "including" and "having" and any variations are intended to cover non-rowing, such as process, methods, systems, products, or equipment that contain a series of steps or units, including a series of steps or units, not necessarily limited to clearly These steps or units may include other steps or units that are not clearly listed or for these processes, methods, products, or equipment.
 Public sentiment warning communications industry, mainly to monitor emergencies, such as network anomaly, abnormal system, community complaints and other issues, the use of common methods, emergency warning recall rate can reach 99%, but less than 10% accuracy rate, the main reason is the content of the communications industry, then surgery and time factors, such as the general account of the end, calls class of problems increase dramatically, months, traffic will be a sharp increase in the use of, these are a normal phenomenon, but the existing monitoring system the above classified as emergencies, resulting in low accuracy, the staff need to spend a lot of time to check the accuracy of emergencies.
 In the presence of the special session will result in a miscarriage of justice of the prior art of public opinion, public opinion can not get accurate early warning result, the accuracy of public sentiment warning of poor technical problems.
 To solve the above problems, the present embodiment provides a method for warning the public opinion, apparatus, device, and a storage medium, wherein the method is intended to identify the use of natural language technology, intended frequency statistics, data with a large build regression models, eliminating the industry vary period caused by the impact of a purpose surge phenomenon.
 Optionally, figure 1 Architecture diagram of a warning system provided by public opinion embodiment of the present application. exist figure 1 In the architecture includes a server 101 and the operator information administration server 102.
 Wherein the information administration server 102 may collect traffic data 101 by the carrier server 774 for warning the public opinion data.
 It will be appreciated that the structure of embodiment illustrative embodiment of the present application does not constitute a specific limitation on the public sentiment warning system architecture. In this application, other possible embodiments, the architecture described above may include more or fewer components than shown, or a combination of some of the components, or some component split, or different arrangements of components, specifically, depending on the application scene determined that this is not the limit. figure 1 Components shown may be implemented in hardware, software, or software and hardware.
 It should be understood, it said processor memory may be read mode and executing instructions implemented by the processor, may be realized by a circuit chip.
 Additionally, the network architecture and traffic scenario described embodiment of the present application is to more clearly describe the technical solutions of the embodiments of the present application, not intended to limit the technical solution provided in the embodiment for the present application, known to those of ordinary skill in the art, with the emergence of new business scenarios and the evolution of network architecture, the present application provides technical solutions for similar technical problems, the same applies.
 The following embodiments with reference to specific embodiments of the aspect of the present disclosure is described in detail:
 Optionally, figure 2 The method of warning a schematic flow of public opinion to an embodiment of the present application. Execution subject of the present embodiment may apply figure 1 The information administration server 101, the specific subject may be determined based on the actual execution scenarios. like figure 2 As shown, the method includes the following steps:
 S201: acquiring traffic data, traffic data for the data pretreatment, data traffic effectively.
 Alternatively, data preprocessing of data traffic, comprising:
 , Traffic data for invalid data Sort by natural language processing technology.
 Alternatively, techniques can reptiles, Chinese division technique in which words and natural language processing techniques to achieve one or more pre-processing of data traffic.
 S202: The preset intention recognition model, the effective traffic data is intended to identify, intended to obtain traffic data.
 Alternatively, according to the preset intention before recognition model, the effective data is intended to identify traffic, traffic data is intended to obtain, further comprising:
 Intention to establish a labeled training database; intentions labeled training database intent labeled training data for natural language processing model train, get pre-intent recognition model.
 Which may be labeled by first doing a lot - the "content intended" training data to train the model of natural language processing (Natural Language Processing, NLP), intended for identifying the valid data.
 Here, for example, is not used as word frequency characteristics to identify the intended application of the present embodiment, reducing the cost of the keyword database maintained by establishing intent labeled training database marked out in advance of the intended labeled training data, it is intended to be established before the preset intention recognition recognition model, which according to this model, the intention of identifying, reducing the cost of public opinion analysis of early warning and improve the efficiency.
 Alternatively, it is also possible to identify the intent of word frequency.
 S203: Return traffic forecasting model by default, to the intent traffic data normalized to obtain a first normalized intent data.
 The preset predictive regression model has been based on historical traffic data and historical traffic data public opinion training.
 Exemplarily, image 3 Application of the present embodiment provides a graph showing actual data traffic forecast traffic preset regression model, Figure 4 An intention of this application a plot of data traffic according to an embodiment, Figure 5 Is first normalized data normalized graph of the intended processing, according to Figure 5 Conduct public sentiment warning.
 S204: If the first data is greater than a normalized preset intended intent threshold, it is determined abnormal public opinion, public opinion generating warning information.
 It will be appreciated that the intention here is preset threshold may be determined according to the actual situation, application of the present Comparative Example is not particularly limited.
 Alternatively, after the opinion generating warning information, further comprising:
 Output reports of public opinion.
 Here, embodiments of the application after determining that there is an abnormality in the public opinion, public opinion reports may be output, facilitates opinion information control, adjustment and predicted traffic and traffic information according to this public opinion.
Example embodiments of the present application firstly traffic data for the data pretreatment, data traffic effectively screened out, leaving only the valid data to facilitate subsequent analysis, in order to reduce maintenance costs keyword database is not used as a characteristic word frequency, , may be performed by a preset valid model intended to identify traffic data is intended to identify, obtain traffic data is intended, then, it is intended for traffic data normalized by preset predictive traffic regression model, since the pre- set predictive regression models based on historical traffic data and historical traffic data public opinion is trained to eliminate the influence due to a sharp increase in some special period, then surgery phenomenon brought to remove the exclusion of emergencies increased traffic on influence public opinion data analysis and improve the accuracy of early warning of public opinion.
 In an alternative embodiment, the present embodiment may also apply in advance to establish a preset traffic prediction regression model and the new business model regression, through said data traffic intended to model twice normalized, can effectively eliminate the influence of special time and new services to public sentiment warning, the corresponding, Image 6 Application schematic flowchart of another method provided warning opinion embodiment, such as Image 6 , The method comprising:
 S601: acquiring traffic data, traffic data for the data pretreatment, data traffic effectively.
 S602: The preset intention recognition model, the effective traffic data is intended to identify, intended to obtain traffic data.
 Wherein the step of S601-S602 of the implementation of the steps S201-S202 of the implementation class Similarly, this will not be repeated herein.
 S603: acquiring historical traffic data and historical data public opinion.
 S604: create a default prediction regression models based on historical traffic data and historical traffic data public opinion.
 The preset predictive regression models need to get rid of traffic incidents increased traffic, exemplary, such as June 5, a local network failure occurs, the data can not be taken June 5 of regression model, June 5 need to replace the normal data or May 5 of July 5.
 Exemplary, Figure 7 One kind of the present application on May traffic graph according to an embodiment, Figure 8 One kind of the present application, June traffic graph according to an embodiment, in which traffic occurs abnormal June 5, then the normal data according to the data replacement or 5 May July 5, and can be corrected after curve, Figure 9 June traffic graph illustrating a modified embodiment of the embodiment provided in the present application.
 In order to normalize the data traffic intended process, application of the present embodiment, a pre-established preset traffic prediction regression model, model combines historical traffic data and historical data of public opinion, the burst can be effectively removed events increased traffic, resulting in precise traffic regression model and the actual traffic match, and further improve the accuracy of early warning of public opinion.
 S605: Return traffic forecasting model by default, to the intent traffic data normalized to obtain a first normalized intent data.
 S606: acquiring new business traffic data.
 S607: the establishment of new business models based on regression of new business traffic data.
 Exemplary, Figure 10 Application of the present embodiment provides a graph is not new traffic, Figure 11 The latter new traffic graph according to an embodiment of the present application.
 S608: According to the new business model for the regression first normalized second intent data normalized to obtain a second normalized intent data.
 S609: If the second normalized data is greater than the preset intended intent threshold, it is determined abnormal public opinion, public opinion generating warning information.
 This embodiment may further application is to establish new service input system, input system embedded in public sentiment warning of new business process, the recording of new business point of time, i.e., the new service traffic data, the data can be used to create a new business growth regression model, in order to achieve the intent of the first normalized secondary data normalization, eliminate the influence of public opinion analysis of data brought new business to further improve the accuracy of early warning of public opinion.
 Figure 12 Schematic structural diagram of one kind of public sentiment warning application according to an embodiment of the apparatus, such as Figure 12 Apparatus shown in the embodiment of the present application comprises: a first obtaining module 1201, a first processing module 1202, second module 1203 and the alarm processing module 1204. The warning means here may be the above opinion information administration server 102 itself, or implemented chip functions of the information administration server 102 or an integrated circuit. It should be noted here that the first acquiring module 1201, a first processing module 1202, a second processing module 1203 and division module 1204 warning only a logical division of functions, both of which may be physically integrated, or may be independently of.
 Wherein the first acquiring module, for acquiring traffic data, traffic data for the data pretreatment, data traffic effectively;
 A first processing module, according to a preset intention recognition model, the effective data is intended to identify traffic, data traffic intended to obtain;
 A second processing module, according to a preset regression model predictive traffic, data traffic intended for normalized, to obtain a first normalized data is intended, wherein the predetermined prediction regression model based on historical traffic traffic data and historical data public opinion is trained ;;
 Warning means for, if the first data is greater than a normalized preset intended intent threshold, it is determined abnormal public opinion, public opinion generating warning information.
 Alternatively, the first processing module according to the preset intention recognition model, the effective traffic data is intended to identify, before the traffic intended to obtain the data, said apparatus further comprising:
 The first establishment module for establishing intent labeled training database;
 Training module for the training of intent annotation database intentions labeled training data for natural language processing model train, get pre-intent recognition model.
 Alternatively, in a second processing module according to a preset regression model predictive traffic, data traffic intended for the normalization process, before the first normalization intended to obtain data, said apparatus further comprising:
 A second acquiring module, for acquiring historical traffic data and historical data of public opinion;
 The second establishment module, configured based on historical traffic data and historical data public opinion, to create a default regression model to predict traffic.
 Alternatively, the alarm module when the normalized data is greater than the preset intended intent threshold, it is determined abnormal public opinion, public opinion before the warning information generation, said apparatus further comprising:
 A third acquiring module, for acquiring the new service traffic data;
 The third establishment module for establishing new business based on regression models of new business traffic data;
 A third processing module, according to the new traffic regression model intents first normalized data normalized second, is intended to obtain a second normalized data;
 Accordingly, the alarm module is configured to:
 If the second normalized data is greater than the preset intended intent threshold, it is determined abnormal public opinion, public opinion generating warning information.
 Alternatively, the first acquisition module is configured to:
 , Traffic data for invalid data Sort by natural language processing technology.
 Alternatively, after the warning information warning module generates opinion, further comprising:
 Generation module for generating reports public opinion.
 Figure 13 Present application a schematic structure of a public sentiment warning apparatus according to an embodiment, the warning device may be a public opinion information administration server 102. Components shown herein, their connections and relationships, and their functions only as an example and does not limit the implementation of the present application and / or the requirements described herein.
 like Figure 13 As shown in the public sentiment warning apparatus comprising: a processor 1301 and memory 1302, the various components using different buses connected to each other, and may be mounted on a common motherboard or in other manners as desired. The processor 1301 may process instructions for execution within the public sentiment warning device, comprising a display graphical information stored in memory or on the memory to the external input / output device (such as a display device coupled to the interface) on. In other embodiments, if desired, a plurality of processors and / or multiple buses used along with multiple memories and a plurality of memory. Figure 13 To a processor 1301 for example.
 Memory 1302 as a non-transitory computer-readable storage medium, for storing non-transitory software program, program and non-transitory computer-executable modules, such as the public sentiment warning device of the present application in an embodiment of a method corresponding to program instructions / modules (e.g. , with Figure 12 Shown, a first acquiring module 1201, a first processing module 1202, second module 1203 and the alarm processing module 1204). The processor 1301 through a non-transitory storage running software programs, instructions, and a memory module 1302, so that various functions of the applications and data processing platform performing authentication, i.e., to achieve the above methods public sentiment warning apparatus in the embodiment.
 Public sentiment warning device may further comprise: an input device 1303 and an output device 1304. A processor 1301, memory 1302, input devices 1303 and output devices 1304 may be connected through a bus or other means, Figure 13 In connection with the bus connection.
 The input device 1303 may receive input numeric or character information, and the user is provided with public sentiment warning generating apparatus and a control function related to the key input signal, such as a touch screen, a keyboard, a mouse, or more mouse buttons, trackballs, joysticks, etc. input means. Output device 1304 may be a display device public sentiment warning device, etc. Output devices. The display device may include, without limitation, a liquid crystal display (LCD), a light emitting diode (LED) displays and plasma displays. In some embodiments, the display device may be a touch screen.
 Public sentiment warning apparatus of the embodiment of the present application, the present application may be used to perform the foregoing method aspect of the embodiments, principles and techniques which achieve a similar effect, are not repeated herein.
The present application embodiment also provides a computer readable storage medium that stores a computer execution instruction in the computer readable storage medium, and the computer execution instruction is performed when executed by the processor to implement the prevention warning method described in any of the above.
 The present application embodiment also provides a computer program product, including a computer program, and the computer program is executed by the processor, and is used to implement the prevention warning method of any of the above.
 In several embodiments provided herein, it should be understood that the disclosed systems, apparatus, and methods can be implemented in other ways. For example, the embodiment described above is merely schematic, for example, the division of the unit, only one logic function division, and can have additional division mode, such as a plurality of units or components, may be combined or integrated. To another system, or some features can be ignored, or not executed. Another point, the coupling or direct coupling or communication connections displayed or discussed may be an electrical, mechanical or other form.
 Further, each of the functional units in the various embodiments of the present application may be integrated into one processing unit, or each unit can be generated separately, or two or more units can be integrated into one unit. The above-described integrated units can be implemented in the form of hardware, or may be implemented in the form of a software functional unit.
 Other embodiments of the present disclosure will be readily apparent to those skilled in the art after considered the specification and practicing the disclosed invention. The present application is intended to cover any variations, uses, or adaptive changes in the disclosure, these variations, uses, or adaptive changes follow the general principles of the present disclosure and include known common sense or customary techniques in the art from the present disclosure. . The specification and examples are considered to be exemplary, and the true scope and spirit of the present disclosure are pointed out by the following claims.
 It should be understood that the present disclosure is not limited to the exact structure described above and shown in the figures, and various modifications and changes can be made without departing from their extent. The scope of this disclosure is limited only by the appended claims.
Description & Claims & Application Information
We can also present the details of the Description, Claims and Application information to help users get a comprehensive understanding of the technical details of the patent, such as background art, summary of invention, brief description of drawings, description of embodiments, and other original content. On the other hand, users can also determine the specific scope of protection of the technology through the list of claims; as well as understand the changes in the life cycle of the technology with the presentation of the patent timeline. Login to view more.