A method for processing network public opinion data

Through the public opinion analysis system, the public opinion keywords are monitored, the changes in the warehouse selection of the target object are calculated, and the proportion of related objects is obtained, which solves the problem of inaccurate monitoring of public opinion data in the existing technology, and real-time accurate analysis and monitoring of public opinion data is achieved.

CN118535784BActive Publication Date: 2024-11-19FAST PAGE INFORMATION TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410644365.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-05-23
Publication Date
2024-11-19
Estimated Expiration
2044-05-23

AI Technical Summary

Technical Problem

The existing technology is difficult to accurately monitor the network public opinion situation and obtain feedback from related data when the keyword has been determined, and lacks the ability to accurately analyze the changes in public opinion data.

Method used

Through the public opinion analysis system, the public opinion keywords are monitored in real time, the target object's warehouse selection changes are calculated, and the ratio of rising-related objects and half-liter related objects are automatically obtained, forming feedback information, including linking upward and downward object chains, and providing the direction of public opinion data.

Benefits of technology

It realizes accurate analysis and monitoring of public opinion data, and can automatically identify related objects when public opinion changes, improving the real-time monitoring and analysis efficiency of public opinion data, which is simple and practical.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118535784B_ABST
    Figure CN118535784B_ABST
Patent Text Reader

Abstract

The invention discloses a method for processing network public opinion data, and relates to the technical field of network public opinion data processing. The method monitors the real-time change of the number of warehouse selections determined by the frequency data of the target object. When a sudden change occurs in the consumption rule of the number of warehouse selections of any target object, the ratio of the same-frequency change of the liter-related objects and the half-liter-related objects of the corresponding target object can be automatically obtained, which are marked as the liter ratio and the half-liter ratio respectively. The selection ratio is calculated according to the liter ratio and the half-liter ratio, and information is fed back according to the selection ratio. The feedback information includes two groups, one group is the suspension of the target object and the reduction of the related chain; the other group is composed of the supplementary related chain and the replenishment target object. The application can reasonably analyze the number of occurrences in the target object, and can distinguish the related objects at the same time. When the corresponding target object increases or decreases, accurate analysis and grasp can be achieved, so as to better grasp the public opinion data situation. The invention is simple, effective, and easy to use.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of network public opinion data processing, and specifically relates to a method for processing network public opinion data. Background Art

[0002] The patent with the publication number CN109471965A discloses a network public opinion data collection, processing method and monitoring platform based on big data, the method includes continuously capturing effective first network public opinion data from the received massive structured or semi-structured data; preprocessing the first network public opinion data respectively to obtain second network public opinion data; aggregating the second network public opinion data and inputting it into the deep learning model for deep processing to obtain first network public opinion information; using at least one of topic recognition and tracking, text tendency analysis, hot spot discovery, and automatic summary technology to analyze the first network public opinion information to obtain second network public opinion information; and displaying the second network public opinion information through big data visualization technology. The present invention can realize the automatic collection and processing of massive network public opinion data, greatly improve the processing efficiency of network public opinion data, and timely discover the development trend of public opinion hot spots and various events.

[0003] Regarding online public opinion, how to accurately monitor the corresponding public opinion situation in the system when some keywords are determined, and at the same time obtain related data feedback, this is a problem; based on this, the present application provides a solution. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a method for processing network public opinion data, which specifically includes the following steps:

[0005] With the help of the public opinion analysis system, the public opinion keywords that have been mastered are monitored in real time to obtain two sets of feedback information, one set is the associated ascending objects and ascending associated chains; the other set is composed of descending associated chains and descending target objects;

[0006] Provide administrators with public opinion data trends based on feedback information;

[0007] The public opinion analysis system specifically includes:

[0008] The data processing end is used to monitor the real-time changes of the bin selection number of the target object determined by the frequency data. When the value of the bin selection number of the frequency data of any target object suddenly changes, the mutation here refers to the abnormal increase or decrease in the bin selection number of the target object. The proportion of the same-frequency changes of the corresponding target object's ascending related objects and half-elevation related objects will be automatically obtained, marked as the ascending proportion and the half-elevation proportion respectively. The selection ratio is calculated according to the ascending proportion and the half-elevation proportion, and feedback information is formed according to the selection ratio; the feedback information includes two groups, one group is the associated ascending object and the ascending associated chain; the other group consists of the descending associated chain and the descending target object.

[0009] Furthermore, the frequency data and bin selection number of the target object are obtained by the data collection end.

[0010] Furthermore, the data collection end mines the frequency data of the target object in the following ways:

[0011] First, the target object is monitored, and any target object is obtained and marked as the selected target object. The frequency data of the selected target object is obtained, including the number of occurrences and the number of clicks. The number of bin selections is calculated based on the number of occurrences and the number of clicks. The bin selection number is calculated by the following formula, which is:

[0012] Number of bin selections = 0.39*number of occurrences + 0.61*number of clicks;

[0013] Get the number of remaining warehouse selections for the corresponding selected target object. This data is obtained once a day.

[0014] Furthermore, a sudden change in the consumption pattern of the frequency data of any target object is determined in the following way:

[0015] First, get the corresponding bin selection number, first use the date as the horizontal axis and the remaining bin selection number as the vertical axis to create a two-dimensional model, then connect the points to get the slope of each point;

[0016] It starts to acquire from the initial time, and obtains several slope data continuously, which are marked as the bin-selected slope data group i, i = 1, ..., n, indicating that there are n slope data. The slope change value B is calculated once each time a slope data is added. The specific calculation formula is:

[0017]

[0018] In the formula, |*| means taking the absolute value of the value in brackets, and P is the mean value of the corresponding Zi;

[0019] The calculation continues until the ramp value B is reached. When B exceeds X1, a mutation signal is generated; X1 is a preset value.

[0020] Furthermore, the specific method of obtaining the feedback information composed of the associated ascending object and the ascending association chain is as follows:

[0021] When a mutation signal is generated, the slope data corresponding to the mutation moment will be obtained. When it is a positive number, a forward analysis is performed. The specific method of the forward analysis is:

[0022] Automatically obtain the selected target object at this time, and its corresponding liter-related objects and half-liter-related objects;

[0023] The ratio of the upgrade-related objects whose number of warehouse selections increases to all upgrade-related objects is obtained, and it is marked as the upgrade ratio. Then, the ratio of the half-elevation-related objects whose number of warehouse selections increases to all related objects is obtained, and it is marked as the half-elevation ratio.

[0024] Calculate the selection ratio based on the liter ratio and the half-liter ratio: selection ratio = 0.68*liter ratio + 0.32*half-liter ratio;

[0025] When the selection ratio exceeds X2, an accumulated growth signal is generated, and the selected target object is marked as an associated ascending object; at the same time, all ascending related objects and half-elevation related objects form an ascending associated chain, and the associated ascending objects and the ascending associated chain are merged to form feedback information; X2 is a preset value.

[0026] Furthermore, the specific method of obtaining the feedback information consisting of the descending association chain and the descending target object is as follows:

[0027] Automatically obtain the selected target object at this time, and its corresponding liter-related objects and half-liter-related objects;

[0028] The ratio of the upgrade-related objects whose number of warehouse selections is reduced to all upgrade-related objects is obtained, and it is marked as the upgrade ratio. Then, the ratio of the half-elevation-related objects whose number of warehouse selections is reduced to all related objects is obtained, and it is marked as the half-elevation ratio.

[0029] Calculate the selection ratio based on the liter ratio and the half-liter ratio: selection ratio = 0.68*liter ratio + 0.32*half-liter ratio;

[0030] When the selection ratio exceeds X2, a bin selection under-selection signal is generated, and the selected target object is marked as a descending target object. At the same time, all ascending related objects and half-ascending related objects form a descending association chain, and the descending association chain and the descending target object are fused to form feedback information.

[0031] Further, the liter-related object and the half-liter-related object are specified by the administrator.

[0032] Furthermore, the processing end is used to transmit the feedback information to the feedback end.

[0033] Furthermore, the liter-related objects and the half-liter-related objects are obtained by the lifting and lowering analysis, and the lifting and lowering analysis is specifically performed as follows:

[0034] First, obtain the warehouse selection number of all target objects in the past year. The past three years refer to the number of warehouse selections from the current time forward.

[0035] Select any target object and obtain the bin selection number corresponding to the target object. First, create a two-dimensional model with the date as the horizontal axis and the remaining bin selection number as the vertical axis. Then connect the points to obtain the slope of each point.

[0036] It starts to acquire from the initial time, and obtains several slope data continuously, which are marked as the bin-selected slope data group i, i = 1, ..., n, indicating that there are n slope data. The slope change value B is calculated once each time a slope data is added. The specific calculation formula is:

[0037]

[0038] In the formula, |*| means taking the absolute value of the value in brackets, and P is the mean value of the corresponding Zi;

[0039] Continue to calculate until the ramp value B, when B exceeds X1, a mutation signal is generated;

[0040] When a mutation signal is generated, the slope data corresponding to the mutation moment will be obtained. When it is a positive number, a positive marking signal is generated, otherwise a reverse marking signal is generated;

[0041] After all target objects are processed, one target object is selected and marked as the selected target object. The forward marking signal and reverse marking signal generated by the target object are collected to obtain the ratio of the signals consistent with the selected target object.

[0042] Here, the consistent signal means that when the selected target object generates a forward marking signal, other target objects also generate forward marking signals, and when the selected target object generates a reverse marking signal, other target objects also generate reverse marking signals;

[0043] The proportion of the selected target object is obtained and marked as the same-frequency proportion. When the same-frequency proportion exceeds X3, the one exceeding X3 is marked as an ascending-related object, and the one with the same-frequency proportion between X4 and X3 is marked as a half-ascending-related object, where the range includes the endpoints; X3 and X4 are both preset values.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] This application monitors the real-time changes of the number of warehouse selections determined by the frequency data of the target object. When the consumption pattern of the number of warehouse selections of any target object suddenly changes, the ratio of the same-frequency changes of the liter-related objects and half-liter-related objects of the corresponding target object will be automatically obtained, marked as the liter ratio and the half-liter ratio respectively, and the selection ratio will be calculated according to the liter ratio and the half-liter ratio. According to the selection ratio, feedback information is fed back, and the feedback information includes two groups, one group is the suspension of the target object and the reduction of the associated chain; the other group is composed of the supplementary associated chain and the replenishment target object;

[0046] The present application can reasonably analyze the number of occurrences in the target object and distinguish the related objects at the same time, and can accurately analyze and grasp the increase or decrease in the corresponding target object, so as to better grasp the public opinion data situation; the present invention is simple, effective, and easy to use. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 is a flow chart of the method of the present invention;

[0048] Figure 2 This is a block diagram of the public opinion analysis system of the present invention. DETAILED DESCRIPTION

[0049] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] See also Figure 1 The present application provides a method for processing network public opinion data, which specifically includes the following steps:

[0051] With the help of the public opinion analysis system, the public opinion keywords that have been mastered are monitored in real time to obtain two sets of feedback information, one set is the associated ascending objects and ascending associated chains; the other set is composed of descending associated chains and descending target objects;

[0052] Provide administrators with public opinion data trends based on feedback information;

[0053] The public opinion analysis system specifically includes: data collection end, data processing end and feedback end;

[0054] Among them, the data collection end is used to mine the frequency data of the target object;

[0055] First, the subject matter is monitored. The subject matter is the corresponding Internet, and the target object is the sensitive word. In this embodiment, the sensitive word is set and added by the administrator;

[0056] Get any target object, mark it as the selected target object, get the frequency data of the selected target object, the frequency data includes the number of occurrences and clicks, calculate the bin selection number according to the number of occurrences and clicks, and the bin selection number is calculated by the following formula, specifically:

[0057] Number of bin selections = 0.39*number of occurrences + 0.61*number of clicks;

[0058] Get the number of remaining warehouse selections for the corresponding selected target object. This data is obtained once a day.

[0059] The data collection end is used to transmit the target object and its corresponding bin selection number to the data processing end, and the data processing end is used to perform up-change analysis on the target object and its corresponding bin selection number. The specific method of up-change analysis is as follows:

[0060] First, get the corresponding bin selection number, first use the date as the horizontal axis and the remaining bin selection number as the vertical axis to create a two-dimensional model, then connect the points to get the slope of each point;

[0061] It starts to acquire from the initial time, and obtains several slope data continuously, which are marked as the bin-selected slope data group i, i = 1, ..., n, indicating that there are n slope data. The slope change value B is calculated once each time a slope data is added. The specific calculation formula is:

[0062]

[0063] In the formula, |*| means taking the absolute value of the value in brackets, and P is the mean value of the corresponding Zi;

[0064] Continue to calculate until the ramp value B, when B exceeds X1, a mutation signal is generated; X1 is the preset value;

[0065] When a mutation signal is generated, the slope data corresponding to the mutation moment will be obtained. When it is a positive number, it means that the slope is greater than zero, that is, the slope data is increasing, and the value of the bin selection number has increased. At this time, forward analysis is performed. The specific method of forward analysis is:

[0066] Automatically obtain the selected target object at this time, and its corresponding liter-related objects and half-liter-related objects;

[0067] Here, the upgrade-related objects and the half-upgrade-related objects are other target objects associated with the selected target object. The upgrade-related objects are represented by the increase in the number of bin selections of the corresponding selected target object, and vice versa.

[0068] Half-liter related objects, that is, when the number of selected target objects increases, half-liter related objects will increase in most cases; liter related objects and half-liter related objects are specified by the administrator;

[0069] The ratio of the upgrade-related objects whose number of warehouse selections increases to all upgrade-related objects is obtained, and it is marked as the upgrade ratio. Then, the ratio of the half-elevation-related objects whose number of warehouse selections increases to all related objects is obtained, and it is marked as the half-elevation ratio.

[0070] Calculate the selection ratio based on the liter ratio and the half-liter ratio: selection ratio = 0.68*liter ratio + 0.32*half-liter ratio;

[0071] When the selection ratio exceeds X2, an accumulation growth signal is generated, and the selected target object is marked as an associated ascending object; at the same time, all ascending related objects and half-elevation related objects form an ascending associated chain, and the associated ascending objects and the ascending associated chain are merged to form feedback information; X2 is a preset value;

[0072] When the slope data is less than zero, it means that the slope data is decreasing, and the value of the remaining bin selection number is decreasing; reverse analysis is performed, and the specific method of reverse analysis is:

[0073] Automatically obtain the selected target object at this time, and its corresponding liter-related objects and half-liter-related objects;

[0074] The ratio of the upgrade-related objects whose number of warehouse selections is reduced to all upgrade-related objects is obtained, and it is marked as the upgrade ratio. Then, the ratio of the half-elevation-related objects whose number of warehouse selections is reduced to all related objects is obtained, and it is marked as the half-elevation ratio.

[0075] Calculate the selection ratio based on the liter ratio and the half-liter ratio: selection ratio = 0.68*liter ratio + 0.32*half-liter ratio;

[0076] When the selection ratio exceeds X2, a bin selection under-selection signal is generated, and the selected target object is marked as a descending target object. At the same time, all ascending related objects and half-ascending related objects form a descending association chain, and the descending association chain and the descending target object are fused to form feedback information;

[0077] The processing end is used to transmit the feedback information to the feedback end.

[0078] As the second embodiment of the present application, this embodiment is different from the first embodiment in that, in this embodiment, the rise-related objects and the half-rise-related objects are obtained by rise-fall analysis, and the specific method of the rise-fall analysis is:

[0079] First, obtain the warehouse selection number of all target objects in the past year. The past three years refer to the number of warehouse selections from the current time forward.

[0080] Select any target object and obtain the bin selection number corresponding to the target object. First, create a two-dimensional model with the date as the horizontal axis and the remaining bin selection number as the vertical axis. Then connect the points to obtain the slope of each point.

[0081] It starts to acquire from the initial time, and obtains several slope data continuously, which are marked as the bin-selected slope data group i, i = 1, ..., n, indicating that there are n slope data. The slope change value B is calculated once each time a slope data is added. The specific calculation formula is:

[0082]

[0083] In the formula, |*| means taking the absolute value of the value in brackets, and P is the mean value of the corresponding Zi;

[0084] Continue to calculate until the ramp value B, when B exceeds X1, a mutation signal is generated;

[0085] When a mutation signal is generated, the slope data corresponding to the mutation moment will be obtained. When it is a positive number, a positive marking signal is generated, otherwise a reverse marking signal is generated;

[0086] After all target objects are processed, one target object is selected and marked as the selected target object. The forward marking signal and reverse marking signal generated by the target object are collected to obtain the ratio of the signals consistent with the selected target object.

[0087] Here, the consistent signal means that when the selected target object generates a forward marking signal, other target objects also generate forward marking signals, and when the selected target object generates a reverse marking signal, other target objects also generate reverse marking signals;

[0088] Get the proportion of the selected target object and mark it as the same-frequency proportion. When the same-frequency proportion exceeds X3, mark it as an ascending-related object, and mark it as a half-ascending-related object when the same-frequency proportion is between X4 and X3. The range here includes the endpoints. X3 and X4 are both preset values. Generally, X3 is 0.85 and X4 is 0.65.

[0089] like Figure 2 As shown, as the third embodiment of the present application, this embodiment is implemented on the basis of the second embodiment, and the data processing end includes a processor and a sub-processing unit;

[0090] The processor is used to perform the up-shift analysis mentioned in the first embodiment;

[0091] The sub-processing unit is used to perform the lifting and lowering analysis mentioned in the second embodiment.

[0092] Some of the data in the above formula are calculated by removing the dimensions and taking their numerical values. The formula is a formula that is closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0093] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A method for processing network public opinion data, characterized in that: The method specifically comprises the following steps: With the help of the public opinion analysis system, the public opinion keywords that have been mastered are monitored in real time to obtain two sets of feedback information, one set is the associated ascending objects and ascending associated chains; the other set is composed of descending associated chains and descending target objects; Provide administrators with public opinion data trends based on feedback information; The public opinion analysis system specifically includes: The data processing end is used to monitor the real-time changes of the bin selection number of the target object determined by the frequency data. When the value of the bin selection number of the frequency data of any target object mutates, the mutation here refers to the abnormal increase or decrease of the bin selection number of the target object. The ratio of the same-frequency changes of the corresponding target object's ascending related objects and half-elevation related objects will be automatically obtained, marked as the ascending ratio and the half-elevation ratio respectively. The selection ratio is calculated based on the ascending ratio and the half-elevation ratio, and feedback information is formed based on the selection ratio; the feedback information includes two groups, one group is the associated ascending object and the ascending association chain; the other group is composed of the descending association chain and the descending target object; the target object is a sensitive word, and the sensitive word is set and added by the administrator; The frequency data and bin selection number of the target object are obtained by the data collection end; The data collection end mines the frequency data of the target object in the following ways: First, the target object is monitored, and any target object is obtained and marked as the selected target object. The frequency data of the selected target object is obtained, including the number of occurrences and the number of clicks. The number of bin selections is calculated based on the number of occurrences and the number of clicks. The bin selection number is calculated by the following formula, which is: Number of bin selections = 0.39*number of occurrences + 0.61*number of clicks; Get the number of remaining warehouse selections for the corresponding selected target object. This data is obtained once a day. The specific method of obtaining the feedback information consisting of the associated ascending object and the ascending association chain is: When a mutation signal is generated, the slope data corresponding to the mutation moment will be obtained. When it is a positive number, a forward analysis is performed. The specific method of the forward analysis is: Automatically obtain the selected target object at this time, and its corresponding liter-related objects and half-liter-related objects; The ratio of the upgrade-related objects whose number of warehouse selections increases to all upgrade-related objects is obtained, and it is marked as the upgrade ratio. Then, the ratio of the half-elevation-related objects whose number of warehouse selections increases to all related objects is obtained, and it is marked as the half-elevation ratio. Calculate the selection ratio based on the liter ratio and the half-liter ratio: selection ratio = 0.68*liter ratio + 0.32*half-liter ratio; When the selection ratio exceeds X2, an accumulation growth signal is generated, and the selected target object is marked as an associated ascending object; at the same time, all ascending related objects and half-elevation related objects form an ascending associated chain, and the associated ascending objects and the ascending associated chain are merged to form feedback information; X2 is a preset value; The specific method of obtaining the feedback information consisting of the descending association chain and the descending target object is as follows: Automatically obtain the selected target object at this time, and its corresponding liter-related objects and half-liter-related objects; The ratio of the upgrade-related objects whose number of warehouse selections is reduced to all upgrade-related objects is obtained, and it is marked as the upgrade ratio. Then, the ratio of the half-elevation-related objects whose number of warehouse selections is reduced to all related objects is obtained, and it is marked as the half-elevation ratio. Calculate the selection ratio based on the liter ratio and the half-liter ratio: selection ratio = 0.68*liter ratio + 0.32*half-liter ratio; When the selection ratio exceeds X2, a bin selection under-selection signal is generated, and the selected target object is marked as a descending target object. At the same time, all ascending related objects and half-ascending related objects form a descending association chain, and the descending association chain and the descending target object are fused to form feedback information; The liter-related objects and half-liter-related objects are specified by the administrator.

2. A method for processing network public opinion data according to claim 1, characterized in that: The sudden change in the consumption pattern of the frequency data of any target object is determined in the following way: First, get the corresponding bin selection number, first use the date as the horizontal axis and the remaining bin selection number as the vertical axis to create a two-dimensional model, then connect the points to get the slope of each point; It starts to acquire from the initial time, and obtains several slope data continuously, which are marked as the bin-selected slope data group i, i = 1, ..., n, indicating that there are n slope data. The slope change value B is calculated once each time a slope data is added. The specific calculation formula is: In the formula, |*| means taking the absolute value of the value in brackets, and P is the mean value of the corresponding Zi; The calculation continues until the ramp value B is reached. When B exceeds X1, a mutation signal is generated; X1 is a preset value.

3. According to the method for processing network public opinion data of claim 1, it is characterized in that: The processing end is used to transmit the feedback information to the feedback end.

4. The method for processing network public opinion data according to claim 1, characterized in that: The liter-related objects and half-liter-related objects are also obtained by the lifting and lowering analysis. The specific method of lifting and lowering analysis is as follows: First, obtain the warehouse selection number of all target objects in the past year. The past three years refer to the number of warehouse selections from the current time forward. Select any target object and obtain the bin selection number corresponding to the target object. First, create a two-dimensional model with the date as the horizontal axis and the remaining bin selection number as the vertical axis. Then connect the points to obtain the slope of each point. It starts to acquire from the initial time, and obtains several slope data continuously, which are marked as the bin-selected slope data group i, i = 1, ..., n, indicating that there are n slope data. The slope change value B is calculated once each time a slope data is added. The specific calculation formula is: In the formula, |*| means taking the absolute value of the value in brackets, and P is the mean value of the corresponding Zi; Continue to calculate until the ramp value B, when B exceeds X1, a mutation signal is generated; When a mutation signal is generated, the slope data corresponding to the mutation moment will be obtained. When it is a positive number, a positive marking signal is generated, otherwise a reverse marking signal is generated; After all target objects are processed, one target object is selected and marked as the selected target object. The forward marking signal and reverse marking signal generated by the target object are collected to obtain the ratio of the signals consistent with the selected target object. Here, the consistent signal means that when the selected target object generates a forward marking signal, other target objects also generate forward marking signals, and when the selected target object generates a reverse marking signal, other target objects also generate reverse marking signals; The proportion of the selected target object is obtained and marked as the same-frequency proportion. When the same-frequency proportion exceeds X3, the one exceeding X3 is marked as an ascending-related object, and the one with the same-frequency proportion between X4 and X3 is marked as a half-ascending-related object, where the range includes the endpoints; X3 and X4 are both preset values.