An opinion monitoring system and method based on big data

Abstract

The present invention discloses a public opinion monitoring system and method based on big data, which belongs to the technical field of public opinion monitoring. The present invention includes: S10: based on big data, obtaining comment information in corresponding events on the Internet, classifying useful information in the comment information according to the public opinion analysis perspective, predicting the public opinion trend of the corresponding event based on various types of comment information, and judging whether to conduct public opinion monitoring on the corresponding event based on the predicted public opinion trend; S20: predicting the propagation of the corresponding event based on the forwarding and search conditions of the corresponding event at each moment, and the public opinion trend of the corresponding event predicted in S10; S30: determining the public opinion management time of the corresponding event. The present invention can determine the public opinion management time of the corresponding event, ensure that the corresponding event guides the public opinion in a positive direction before the public opinion becomes uncontrollable, realize comprehensive analysis and monitoring of public opinion events, and improve the monitoring effect of the system.

Description

Technical Field

[0001] The present invention relates to the technical field of public opinion monitoring, and in particular to a public opinion monitoring system and method based on big data. Background Art

[0002] Public opinion monitoring refers to the integration of Internet information collection technology and information intelligent processing technology to achieve user network public opinion monitoring through automatic capture, automatic classification and clustering, topic detection, and special focus on massive amounts of Internet information.

[0003] The existing public opinion monitoring system only collects and analyzes part of the information in the event when monitoring public opinion, and is unable to achieve comprehensive analysis and monitoring of public opinion events. In addition, the public opinion monitoring technology functions in the existing system are relatively simple and cannot predict the public opinion monitoring time, that is, it is impossible to achieve effective control of public opinion. In addition, when analyzing public opinion, the analysis process is relatively cumbersome and the analysis efficiency is too low. Summary of the Invention

[0004] The purpose of the present invention is to provide a public opinion monitoring system and method based on big data to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solution: a method for monitoring public opinion based on big data, characterized in that the method includes:

[0006] S10: Based on big data, we obtain the comments on the corresponding events on the Internet, classify the useful information in the comments according to the perspective of public opinion analysis, and predict the direction of public opinion on the corresponding events based on the various types of comments.

[0007] Based on the predicted trend of public opinion, determine whether to conduct public opinion monitoring on the corresponding event;

[0008] S20: Predicting the spread of the corresponding event based on the forwarding and search status of the corresponding event at each moment and the public opinion trend of the corresponding event predicted in S10;

[0009] S30: Determine the public opinion management time for the corresponding event.

[0010] Furthermore, the S10 includes:

[0011] S101: Acquire comment information in corresponding events on the Internet at time intervals t, and divide the acquired comment information into primary comment information and secondary comment information according to whether there is a reply relationship between the comment information. Primary comment information refers to comment information that has no reply information and is not a reply information itself, and secondary comment information is reply information to the primary comment information. Determine whether the comment opinions of the primary comment information and the matching secondary comment information are consistent. If not, modify the matching secondary comment information to be the primary comment information, or use it as the secondary comment information of the primary comment information that is consistent with its comment opinion. If consistent, there is no need to modify the secondary comment information. Acquire the degree of correlation between the primary comment information and the corresponding event. If the correlation degree is lower than a set threshold, it indicates that the corresponding primary comment information and the matching secondary comment information are useless information. Delete the useless information and the comment information with neutral comment opinions in the comment information. The remaining information after deletion is useful information.

[0012] S102: Determine the number of controversial points of the corresponding event based on the useful information in S101, randomly combine the determined controversial points, and determine the number of public opinion analysis angles for the corresponding event based on the random combinations. The specific determination formula is:

[0013]

[0014] Where v = 1, 2, …, n, represents the number of controversial points in the combination, n represents the total number of controversial points in the corresponding event, n! represents the factorial of n, (nv)! represents the factorial of nv, and W represents the number of public opinion analysis angles for the corresponding event. The public opinion analysis angle refers to the controversial angle corresponding to the controversial point of the corresponding event. For example, if the controversial point of the corresponding event is that students should focus on extracurricular activities, the corresponding public opinion analysis angle is whether students should focus on extracurricular activities.

[0015] Classify the useful information in S101 according to the perspective of public opinion analysis, put the main comments belonging to the same category into the same set, and predict the direction of public opinion on the corresponding event based on the main comments in the set and the secondary comments that match the main comments in the set;

[0016] S103: If the public opinion trend of the corresponding event is positive, then there is no need to monitor the public opinion of the corresponding event. If the public opinion trend of the corresponding event is negative, then there is a need to monitor the public opinion of the corresponding event.

[0017] Furthermore, the specific method of predicting the direction of public opinion on the corresponding event in S102 is:

[0018] A. Extract the comments of the main comment information in the collection, using the formula The support rate of each extracted comment is calculated, where j = 1, 2, ..., m, represents the number of the main comment information in the set, m represents the total number of main comment information in the set, i = 1, 2, ..., r, represents the number of the extracted comment view, r represents the total number of extracted comment view, a ji =0 or a ji =1, when a ji =1 means that the main comment information numbered j supports the comment opinion numbered i. ji = 0 means that the main comment information numbered j does not support the comment opinion numbered i, b j F represents the total number of secondary review information that matches the primary review information numbered j, i Indicates the support rate of the comment numbered i in the set;

[0019] B. Based on the support rate of each extracted comment calculated in A, predict the direction of public opinion on the corresponding event. The specific prediction formula is:

[0020]

[0021] Among them, p=1,2,…,W, represents the number corresponding to the set, α p Indicates the proportional coefficient corresponding to the public opinion analysis angle corresponding to the set numbered p, F ip represents the support rate of the opinion numbered i in the set p, β i =0 or β i =1, when β i =1, it means that the comment numbered i is a positive comment. i =0, it means that the comment numbered i is a negative comment, and E represents the reference value of the public opinion trend of the corresponding event;

[0022] When 0.6≤E≤1, it means that the public opinion trend of the corresponding event is positive; when 0≤E<0.6, it means that the public opinion trend of the corresponding event is negative.

[0023] Furthermore, the S20 includes:

[0024] S201: Calculate the propagation rate of the corresponding event at each moment based on the reference value of the public opinion trend of the corresponding event predicted at each moment in S102. Propagation rate = [(E T+t*(z-1) -E T+t*z ) / (1-E T+t*(z-1) )], where T represents the initial time of obtaining the comment information in the corresponding event, z represents the number of times the comment information in the corresponding event is obtained, z≥2, E T+t*zIt represents the reference value of the public opinion trend of the corresponding event at the time point T+t*z;

[0025] S202: Obtain the forwarding and search status of the corresponding event at each time, and predict the spread of the corresponding event at each time based on the obtained information. The specific prediction formula is:

[0026] S=γ*(Q T+t*z -Q T+t*(z-1) )+(1-γ)*(P T+t*z -P T+t*(z-1) );

[0027] Among them, γ represents the proportional coefficient, Q T+t*z represents the forwarding volume of the corresponding event at time point T+t*z, P T+t*z It represents the search volume of the corresponding event at the time point T+t*z, and S represents the degree of spread of the corresponding event at the time point T+t*z.

[0028] Furthermore, S30 determines the public opinion management time of the corresponding event based on the propagation rate of the corresponding event at each moment predicted in S201 and the propagation degree of the corresponding event at each moment predicted in S202. The specific determination method is: calculate the product of the propagation rate and the propagation degree of the corresponding event at each moment, determine the maximum value of the calculated product, and use the time corresponding to the determined maximum value as the public opinion management time of the corresponding event.

[0029] A public opinion monitoring system based on big data, comprising a public opinion monitoring and analysis module, a public opinion dissemination prediction module, and a public opinion management time determination module;

[0030] The public opinion monitoring and analysis module is used to obtain comment information on the corresponding event on the Internet, classify the useful information in the comment information according to the public opinion analysis perspective, predict the public opinion trend of the corresponding event based on the various types of comment information, and determine whether to monitor the public opinion of the corresponding event based on the predicted public opinion trend. Based on the judgment result, the predicted public opinion trend of the corresponding event is transmitted to the public opinion dissemination prediction module;

[0031] The public opinion propagation prediction module is used to receive the public opinion trend of the corresponding event transmitted by the public opinion monitoring and analysis module, and based on the received information, combined with the forwarding and search conditions of the corresponding event at each moment, predict the propagation of the corresponding event, and transmit the prediction results to the public opinion management time determination module;

[0032] The public opinion management time determination module is used to receive the prediction results transmitted by the public opinion dissemination situation prediction module, and determine the public opinion management time of the corresponding event based on the received information.

[0033] Furthermore, the public opinion monitoring and analysis module includes an information division unit, a public opinion analysis angle determination unit, a public opinion trend prediction unit, and a public opinion monitoring unit;

[0034] The information division unit obtains comment information in the corresponding event on the Internet at a time interval t, divides the obtained comment information into main comment information and secondary comment information according to whether there is a reply relationship between the comment information, judges whether the comment opinions of the main comment information and the matched secondary comment information are consistent, performs relevant processing on the secondary comment information according to the judgment result, obtains the degree of correlation between the main comment information and the corresponding event, and if the degree of correlation is lower than a set threshold, it indicates that the corresponding main comment information and the matched secondary comment information are useless information, deletes the useless information and the comment information with neutral comment opinions in the comment information, and the remaining information after deletion is useful information, which is transmitted to the public opinion analysis angle determination unit and the public opinion trend prediction unit;

[0035] The public opinion analysis angle determination unit receives the useful information transmitted by the information division unit, determines the number of controversial points of the corresponding event based on the received useful information, randomly combines the determined controversial points, determines the number of public opinion analysis angles of the corresponding event based on the random combination, and transmits the determined number of public opinion analysis angles to the public opinion trend prediction unit;

[0036] The public opinion trend prediction unit receives the number of public opinion analysis angles transmitted by the public opinion analysis angle determination unit and the useful information transmitted by the information division unit, classifies the received useful information according to the public opinion analysis angles, puts the main comment information belonging to the same classification into the same set, and predicts the public opinion trend of the corresponding event based on the main comment information in the set and the secondary comment information matching the main comment information in the set, and transmits the prediction result to the public opinion monitoring unit, and transmits the public opinion trend reference value to the public opinion dissemination situation prediction module;

[0037] The public opinion monitoring unit receives the prediction results transmitted by the public opinion trend prediction unit, selects whether to perform public opinion monitoring on the corresponding event based on the received information, and transmits the selection result to the public opinion dissemination situation prediction module.

[0038] Furthermore, the public opinion propagation situation prediction module includes a propagation rate prediction unit and a propagation degree prediction unit;

[0039] The propagation rate prediction unit receives the selection result transmitted by the public opinion monitoring unit. If public opinion monitoring is selected for the corresponding event, the public opinion trend reference value transmitted by the public opinion trend prediction unit is received. Based on the received information, the propagation rate of the corresponding event at each moment is calculated, and the calculation result is transmitted to the public opinion management time determination module. If public opinion monitoring is not selected for the corresponding event, there is no need to receive the public opinion trend reference value transmitted by the public opinion trend prediction unit.

[0040] The propagation degree prediction unit obtains the forwarding and search conditions of the corresponding event at each moment, predicts the propagation degree of the corresponding event at each moment based on the obtained information, and transmits the predicted propagation degree to the public opinion management time determination module.

[0041] Furthermore, the public opinion management time determination module receives the propagation rate transmitted by the propagation rate prediction unit and the propagation degree transmitted by the propagation degree prediction unit, and determines the public opinion management time of the corresponding event based on the received information.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] 1. When conducting public opinion analysis on each event, the present invention filters out useful information from each event, thereby improving the system's efficiency in analyzing public opinion. The filtered useful information is classified and analyzed according to the public opinion analysis perspective, thereby avoiding incorrect judgment of the expression meaning of useful information due to different analysis perspectives when analyzing useful information, thereby weakening the connection between information and reducing the effect of public opinion analysis on the event.

[0044] 2. The present invention classifies and analyzes the useful information in each event from the perspective of public opinion analysis, calculates the support rate of the comments and opinions extracted from the useful information belonging to the same category, and predicts the direction of public opinion on the corresponding event based on the calculation results. This process realizes comprehensive analysis and monitoring of public opinion events, thereby improving the monitoring effect of the system.

[0045] 3. The present invention predicts the propagation rate and degree of the corresponding event at each moment through the reference value of the predicted public opinion trend of the corresponding event, and determines the public opinion management time of the corresponding event based on the predicted value to ensure that the corresponding event guides public opinion in a positive direction before it becomes uncontrollable. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0047] Figure 1This is a workflow diagram of a public opinion monitoring system and method based on big data of the present invention;

[0048] Figure 2 It is a structural diagram of the working principle of a public opinion monitoring system and method based on big data of the present invention. DETAILED DESCRIPTION

[0049] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0050] See also Figure 1 and Figure 2 The present invention provides a technical solution: a method for monitoring public opinion based on big data, the method comprising:

[0051] S10: Based on big data, we obtain the comments on the corresponding events on the Internet, classify the useful information in the comments according to the perspective of public opinion analysis, and predict the direction of public opinion on the corresponding events based on the various types of comments.

[0052] Based on the predicted trend of public opinion, determine whether to conduct public opinion monitoring on the corresponding event;

[0053] The S10 includes:

[0054] S101: Acquire comment information in corresponding events on the Internet at time intervals t, and divide the acquired comment information into primary comment information and secondary comment information according to whether there is a reply relationship between the comment information. Primary comment information refers to comment information that has no reply information and is not a reply information itself, and secondary comment information is reply information to the primary comment information. Determine whether the comment opinions of the primary comment information and the matching secondary comment information are consistent. If not, modify the matching secondary comment information to be the primary comment information, or use it as the secondary comment information of the primary comment information that is consistent with its comment opinion. If consistent, there is no need to modify the secondary comment information. Acquire the degree of correlation between the primary comment information and the corresponding event. If the correlation degree is lower than a set threshold, it indicates that the corresponding primary comment information and the matching secondary comment information are useless information. Delete the useless information and the comment information with neutral comment opinions in the comment information. The remaining information after deletion is useful information.

[0055] S102: Determine the number of controversial points of the corresponding event based on the useful information in S101, randomly combine the determined controversial points, and determine the number of public opinion analysis angles for the corresponding event based on the random combinations. The specific determination formula is:

[0056]

[0057] Where v = 1, 2, …, n, represents the number of controversial points in the combination, n represents the total number of controversial points in the corresponding event, n! represents the factorial of n, (nv)! represents the factorial of nv, and W represents the number of public opinion analysis angles for the corresponding event.

[0058] The useful information in S101 is classified according to the perspective of public opinion analysis. The main comments belonging to the same category are put into the same set. Based on the main comments in the set and the secondary comments that match the main comments in the set, the public opinion trend of the corresponding event is predicted. The specific prediction method is as follows:

[0059] A. Extract the comments of the main comment information in the collection, using the formula The support rate of each extracted comment is calculated, where j = 1, 2, ..., m, represents the number of the main comment information in the set, m represents the total number of main comment information in the set, i = 1, 2, ..., r, represents the number of the extracted comment view, r represents the total number of extracted comment view, a ji =0 or a ji =1, when a ji =1 means that the main comment information numbered j supports the comment opinion numbered i. ji = 0 means that the main comment information numbered j does not support the comment opinion numbered i, b j F represents the total number of secondary review information that matches the primary review information numbered j, i Indicates the support rate of the comment numbered i in the set;

[0060] B. Based on the support rate of each extracted comment calculated in A, predict the direction of public opinion on the corresponding event. The specific prediction formula is:

[0061]

[0062] Among them, p=1,2,…,W, represents the number corresponding to the set, α p Indicates the proportional coefficient corresponding to the public opinion analysis angle corresponding to the set numbered p, F ip represents the support rate of the opinion numbered i in the set p, β i =0 or βi =1, when β i =1, it means that the comment numbered i is a positive comment. i =0, it means that the comment numbered i is a negative comment, and E represents the reference value of the public opinion trend of the corresponding event;

[0063] When 0.6≤E≤1, it means that the public opinion trend of the corresponding event is positive; when 0≤E<0.6, it means that the public opinion trend of the corresponding event is negative;

[0064] S103: If the public opinion trend for the corresponding event is positive, then there is no need to monitor the public opinion for the corresponding event. If the public opinion trend for the corresponding event is negative, then there is a need to monitor the public opinion for the corresponding event.

[0065] S20: Predicting the spread of the corresponding event based on the forwarding and search status of the corresponding event at each moment and the public opinion trend of the corresponding event predicted in S10;

[0066] The S20 includes:

[0067] S201: Calculate the propagation rate of the corresponding event at each moment based on the reference value of the public opinion trend of the corresponding event predicted at each moment in S102. Propagation rate = [(E T+t*(z-1) -E T+t*z ) / (1-E T+t*(z-1) )], where T represents the initial time of obtaining the comment information in the corresponding event, z represents the number of times the comment information in the corresponding event is obtained, z≥2, E T+t*z It represents the reference value of the public opinion trend of the corresponding event at the time point T+t*z;

[0068] S202: Obtain the forwarding and search status of the corresponding event at each time, and predict the spread of the corresponding event at each time based on the obtained information. The specific prediction formula is:

[0069] S=γ*(Q T+t*z -Q T+t*(z-1) )+(1-γ)*(P T+t*z -P T+t*(z-1) );

[0070] Among them, γ represents the proportional coefficient, Q T+t*z represents the forwarding volume of the corresponding event at time point T+t*z, P T+t*z represents the search volume of the corresponding event at the time point T+t*z, and S represents the spread degree of the corresponding event at the time point T+t*z;

[0071] S30: Determine the time for public opinion management of the corresponding event;

[0072] S30 determines the public opinion management time of the corresponding event based on the propagation rate of the corresponding event at each moment predicted in S201 and the propagation degree of the corresponding event at each moment predicted in S202. The specific determination method is: calculate the product of the propagation rate and the propagation degree of the corresponding event at each moment, determine the maximum value of the calculated product, and use the time corresponding to the determined maximum value as the public opinion management time of the corresponding event.

[0073] A public opinion monitoring system based on big data, the system includes a public opinion monitoring and analysis module, a public opinion dissemination prediction module and a public opinion management time determination module;

[0074] The public opinion monitoring and analysis module is used to obtain comment information on the corresponding event on the Internet, classify the useful information in the comment information according to the public opinion analysis perspective, predict the public opinion trend of the corresponding event based on the various types of comment information, and determine whether to conduct public opinion monitoring on the corresponding event based on the predicted public opinion trend. Based on the judgment result, the predicted public opinion trend of the corresponding event is transmitted to the public opinion dissemination prediction module;

[0075] The public opinion monitoring and analysis module includes an information division unit, a public opinion analysis angle determination unit, a public opinion trend prediction unit, and a public opinion monitoring unit;

[0076] The information division unit obtains comment information in the corresponding event on the Internet at time intervals t, divides the obtained comment information into primary comment information and secondary comment information according to whether there is a reply relationship between the comment information, judges whether the comment opinions of the primary comment information and the matching secondary comment information are consistent, performs relevant processing on the secondary comment information according to the judgment result, and obtains the degree of correlation between the primary comment information and the corresponding event. If the correlation degree is lower than a set threshold, it means that the corresponding primary comment information and the matching secondary comment information are useless information. The useless information and the comment information with neutral comment opinions in the comment information are deleted. The remaining information after deletion is useful information, and the useful information is transmitted to the public opinion analysis angle determination unit and the public opinion trend prediction unit;

[0077] The public opinion analysis angle determination unit receives the useful information transmitted by the information division unit, determines the number of controversial points of the corresponding event based on the received useful information, randomly combines the determined controversial points, and based on the random combination, determines the number of public opinion analysis angles for the corresponding event, and transmits the determined number of public opinion analysis angles to the public opinion trend prediction unit;

[0078] The public opinion trend prediction unit receives the number of public opinion analysis angles transmitted by the public opinion analysis angle determination unit and the useful information transmitted by the information division unit, classifies the received useful information according to the public opinion analysis angles, puts the main comment information belonging to the same category into the same set, and predicts the public opinion trend of the corresponding event based on the main comment information in the set and the secondary comment information matching the main comment information in the set, and transmits the prediction result to the public opinion monitoring unit, and transmits the public opinion trend reference value to the public opinion dissemination prediction module;

[0079] The public opinion monitoring unit receives the prediction results transmitted by the public opinion trend prediction unit, selects whether to monitor public opinion for the corresponding event based on the received information, and transmits the selection result to the public opinion dissemination prediction module;

[0080] The public opinion propagation prediction module is used to receive the public opinion trend of the corresponding event transmitted by the public opinion monitoring and analysis module. Based on the received information and combined with the forwarding and search status of the corresponding event at each moment, it predicts the propagation of the corresponding event and transmits the prediction results to the public opinion management time determination module;

[0081] The public opinion propagation prediction module includes a propagation rate prediction unit and a propagation degree prediction unit;

[0082] The propagation rate prediction unit receives the selection result transmitted by the public opinion monitoring unit. If public opinion monitoring is selected for the corresponding event, the public opinion trend reference value transmitted by the public opinion trend prediction unit is received. Based on the received information, the propagation rate of the corresponding event at each moment is calculated, and the calculation result is transmitted to the public opinion management time determination module. If public opinion monitoring is not selected for the corresponding event, there is no need to receive the public opinion trend reference value transmitted by the public opinion trend prediction unit.

[0083] The propagation degree prediction unit obtains the forwarding and search status of the corresponding event at each time, predicts the propagation degree of the corresponding event at each time based on the obtained information, and transmits the predicted propagation degree to the public opinion management time determination module;

[0084] The public opinion management time determination module receives the propagation rate transmitted by the propagation rate prediction unit and the propagation degree transmitted by the propagation degree prediction unit, and determines the public opinion management time of the corresponding event based on the received information.

[0085] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0086] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for monitoring public opinion based on big data, characterized by: The method comprises: S10: Based on big data, obtain comments on the corresponding event on the Internet, classify useful information in the comments according to the perspective of public opinion analysis, predict the direction of public opinion on the corresponding event based on the various types of comments, and determine whether to conduct public opinion monitoring on the corresponding event based on the predicted direction of public opinion; The S10 includes: S101: obtaining comment information on corresponding events on the Internet at time intervals t, dividing the obtained comment information into primary comment information and secondary comment information according to whether there is a reply relationship between the comment information, judging whether the comment opinions of the primary comment information and the matched secondary comment information are consistent, if not, modifying the matched secondary comment information to the primary comment information, or as the secondary comment information of the primary comment information consistent with its comment opinion, if consistent, then there is no need to modify the secondary comment information, obtaining the degree of correlation between the primary comment information and the corresponding event, if the degree of correlation is lower than a set threshold, indicating that the corresponding primary comment information and the matched secondary comment information are useless information, deleting the useless information and the comment information with neutral comment opinions in the comment information, and the remaining information after deletion is useful information; S102: Determine the number of controversial points of the corresponding event based on the useful information in S101, randomly combine the determined controversial points, and determine the number of public opinion analysis angles for the corresponding event based on the random combinations. The specific determination formula is: Where v = 1, 2, …, n, represents the number of controversial points in the combination, n represents the total number of controversial points in the corresponding event, n! represents the factorial of n, (nv)! represents the factorial of nv, and W represents the number of public opinion analysis angles for the corresponding event. The useful information in S101 is classified according to the perspective of public opinion analysis, and the main comments belonging to the same category are put into the same set. Based on the main comments in the set and the secondary comments matching the main comments in the set, the public opinion trend of the corresponding event is predicted. The specific method is as follows: A. Extract the comments of the main comment information in the collection, using the formula The support rate of each extracted comment is calculated, where j = 1, 2, ..., m, represents the number of the main comment information in the set, m represents the total number of main comment information in the set, i = 1, 2, ..., r, represents the number of the extracted comment, r represents the total number of extracted comment, a ji =0 or a ji =1, when a ji =1 means that the main comment information numbered j supports the comment opinion numbered i. ji = 0 means that the main comment information numbered j does not support the comment opinion numbered i, b j F represents the total number of secondary review information that matches the primary review information numbered j, i Indicates the support rate of the comment numbered i in the set; B. Based on the support rate of each extracted comment calculated in A, predict the direction of public opinion on the corresponding event. The specific prediction formula is: Among them, p=1,2,…,W, represents the number corresponding to the set, α p Indicates the proportional coefficient corresponding to the public opinion analysis angle corresponding to the set numbered p, F ip represents the support rate of the opinion numbered i in the set p, β i =0 or β i =1, when β i =1, it means that the comment numbered i is a positive comment. i =0, it means that the comment numbered i is a negative comment, and E represents the reference value of the public opinion trend of the corresponding event; When 0.6≤E≤1, it means that the public opinion trend of the corresponding event is positive; when 0≤E<0.6, it means that the public opinion trend of the corresponding event is negative; S103: If the public opinion trend for the corresponding event is positive, then there is no need to monitor the public opinion for the corresponding event. If the public opinion trend for the corresponding event is negative, then there is a need to monitor the public opinion for the corresponding event. S20: Predicting the spread of the corresponding event based on the forwarding and search status of the corresponding event at each moment, and the public opinion trend of the corresponding event predicted in S10; S30: Determine the public opinion management time for the corresponding event.

2. The method for monitoring public opinion based on big data according to claim 1, characterized in that: The S20 includes: S201: Calculate the propagation rate of the corresponding event at each moment based on the reference value of the public opinion trend of the corresponding event predicted at each moment in S102. Propagation rate = [(E T+t*(z-1) -E T+t*z ) / (1-E T+t*(z-1) )], where T represents the initial time of obtaining the comment information in the corresponding event, z represents the number of times the comment information in the corresponding event is obtained, z≥2, E T+t*z It represents the reference value of the public opinion trend of the corresponding event at the time point T+t*z; S202: Obtain the forwarding and search status of the corresponding event at each time, and predict the spread of the corresponding event at each time based on the obtained information. The specific prediction formula is: S=γ*(Q T+t*z -Q T+t*(z-1) )+(1-γ)*(P T+t*z -P T+t*(z-1) ); Among them, γ represents the proportional coefficient, Q T+t*z represents the forwarding volume of the corresponding event at time point T+t*z, P T+t*z It represents the search volume of the corresponding event at the time point T+t*z, and S represents the degree of spread of the corresponding event at the time point T+t*z.

3. The method for monitoring public opinion based on big data according to claim 2, characterized in that: The S30 determines the public opinion management time of the corresponding event based on the propagation rate of the corresponding event at each moment predicted in S201 and the propagation degree of the corresponding event at each moment predicted in S202. The specific determination method is: calculate the product of the propagation rate and the propagation degree of the corresponding event at each moment, determine the maximum value of the calculated product, and use the time corresponding to the determined maximum value as the public opinion management time of the corresponding event.

4. A big data-based public opinion monitoring system applied to the big data-based public opinion monitoring method according to any one of claims 1 to 3, characterized in that: The system includes a public opinion monitoring and analysis module, a public opinion dissemination situation prediction module and a public opinion management time determination module; The public opinion monitoring and analysis module is used to obtain comment information on the corresponding event on the Internet, classify the useful information in the comment information according to the public opinion analysis perspective, predict the public opinion trend of the corresponding event based on the various types of comment information, and determine whether to monitor the public opinion of the corresponding event based on the predicted public opinion trend. Based on the judgment result, the predicted public opinion trend of the corresponding event is transmitted to the public opinion dissemination prediction module; The public opinion propagation prediction module is used to receive the public opinion trend of the corresponding event transmitted by the public opinion monitoring and analysis module, and based on the received information, combined with the forwarding and search conditions of the corresponding event at each moment, predict the propagation of the corresponding event, and transmit the prediction results to the public opinion management time determination module; The public opinion management time determination module is used to receive the prediction results transmitted by the public opinion dissemination situation prediction module, and determine the public opinion management time of the corresponding event based on the received information.

5. The public opinion monitoring system based on big data according to claim 4 is characterized by: The public opinion monitoring and analysis module includes an information division unit, a public opinion analysis angle determination unit, a public opinion trend prediction unit and a public opinion monitoring unit; The information division unit obtains comment information in the corresponding event on the Internet at a time interval t, divides the obtained comment information into main comment information and secondary comment information according to whether there is a reply relationship between the comment information, judges whether the comment opinions of the main comment information and the matched secondary comment information are consistent, performs relevant processing on the secondary comment information according to the judgment result, obtains the degree of correlation between the main comment information and the corresponding event, and if the degree of correlation is lower than a set threshold, it indicates that the corresponding main comment information and the matched secondary comment information are useless information, deletes the useless information and the comment information with neutral comment opinions in the comment information, and the remaining information after deletion is useful information, which is transmitted to the public opinion analysis angle determination unit and the public opinion trend prediction unit; The public opinion analysis angle determination unit receives the useful information transmitted by the information division unit, determines the number of controversial points of the corresponding event based on the received useful information, randomly combines the determined controversial points, determines the number of public opinion analysis angles of the corresponding event based on the random combination, and transmits the determined number of public opinion analysis angles to the public opinion trend prediction unit; The public opinion trend prediction unit receives the number of public opinion analysis angles transmitted by the public opinion analysis angle determination unit and the useful information transmitted by the information division unit, classifies the received useful information according to the public opinion analysis angles, puts the main comment information belonging to the same classification into the same set, and predicts the public opinion trend of the corresponding event based on the main comment information in the set and the secondary comment information matching the main comment information in the set, and transmits the prediction result to the public opinion monitoring unit, and transmits the public opinion trend reference value to the public opinion dissemination situation prediction module; The public opinion monitoring unit receives the prediction results transmitted by the public opinion trend prediction unit, selects whether to perform public opinion monitoring on the corresponding event based on the received information, and transmits the selection result to the public opinion dissemination situation prediction module.

6. The public opinion monitoring system based on big data according to claim 5 is characterized by: The public opinion propagation situation prediction module includes a propagation rate prediction unit and a propagation degree prediction unit; The propagation rate prediction unit receives the selection result transmitted by the public opinion monitoring unit. If public opinion monitoring is selected for the corresponding event, the public opinion trend reference value transmitted by the public opinion trend prediction unit is received. Based on the received information, the propagation rate of the corresponding event at each moment is calculated, and the calculation result is transmitted to the public opinion management time determination module. If public opinion monitoring is not selected for the corresponding event, there is no need to receive the public opinion trend reference value transmitted by the public opinion trend prediction unit. The propagation degree prediction unit obtains the forwarding and search conditions of the corresponding event at each moment, predicts the propagation degree of the corresponding event at each moment based on the obtained information, and transmits the predicted propagation degree to the public opinion management time determination module.

7. The public opinion monitoring system based on big data according to claim 6 is characterized by: The public opinion management time determination module receives the propagation rate transmitted by the propagation rate prediction unit and the propagation degree transmitted by the propagation degree prediction unit, and determines the public opinion management time of the corresponding event based on the received information.

Images