A themed data processing method and system

Abstract

The present invention is applicable to the field of data processing technology, and in particular to a thematic data processing method and system, the method comprising: S100: establishing a word set of thematic keywords, classifying the word set into a preset data platform for matching, defining the matching result as an initial event, extracting the initial viewpoint, creating a monitoring tree, and embedding the hierarchical architecture. By constructing a monitoring tree, the present invention can monitor the spread of the initial event in real time, help users to grasp the public opinion dynamics related to themselves in a timely manner, identify potential crisis signals, and quickly make public relations responses to avoid the spread and fermentation of negative information. It can also understand the current public concerns and emotional tendencies, adjust marketing strategies, and through a linked list, can help users fully understand the complete path from the emergence to the spread of a topic, formulate more accurate marketing and public relations strategies, and provide decision support for users.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a themed data processing method and system. Background Art

[0002] Thematic data refers to data collected, organized and processed around a specific theme or topic. Thematic data in social networks usually refers to topics or events that attract a lot of user attention, discussion and interaction within a specific time period, which are often referred to as hot searches or hot spots. Analyzing hot topics can help understand their popularity changes, user emotional tendencies and dissemination paths, and thus be applied to brand marketing, public relations management, market research and content creation.

[0003] Therefore, “how to monitor the propagation path of subjectified data and conduct quantitative analysis” is the technical problem that the present invention needs to solve. Summary of the invention

[0004] The purpose of the present invention is to provide a thematic data processing method and system to solve the problem of "how to monitor the propagation path of thematic data and conduct quantitative analysis" raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A themed data processing method, the method comprising:

[0007] S100: Establish a word set of thematic keywords, classify the word set into a preset data platform for matching, define the matching result as an initial event, extract an initial viewpoint, create a monitoring tree, and embed a hierarchical structure, hang the pre-built root node, branch node, and leaf node layer by layer into the hierarchical structure, and transfer the initial event and initial viewpoint into the root node;

[0008] S200: Collect browsing data and user ID of the initial event, extract comments of the user ID on the initial event, obtain a first forwarding viewpoint, collect browsing data and user ID of the first forwarding viewpoint, obtain a second forwarding viewpoint, and so on, obtain a third forwarding viewpoint, ..., nth forwarding viewpoint, synchronize the first forwarding viewpoint to a branch node, and synchronize the second forwarding viewpoint to a leaf node;

[0009] S300: construct a sentiment word query dictionary, query the sentiment values ​​of all forwarded opinions, arrange the second forwarded opinions in the branch node in the order of the sentiment values ​​from small to large, determine the quartiles, define the interval composed of the first quartile and the third quartile as the fluctuation range, calculate the average value of the sentiment value within the fluctuation range, construct a number block corresponding to the hierarchical structure, transfer the fluctuation range and the average value into the number block, integrate the number blocks, and determine a linked list;

[0010] S400: Integrate a monitoring mechanism into the linked list, compare changes in sentiment values, and determine a treatment plan.

[0011] Furthermore, the S100 includes:

[0012] Using the word set, a semantic network is constructed, and matching results are optimized;

[0013] The matching results are formatted, initial viewpoints are extracted using viewpoint mining technology, the matching results are numbered, and the corresponding relationship between the matching results and the initial viewpoints is configured.

[0014] Furthermore, the S100 further includes:

[0015] Locate the bottom of the monitoring tree, transfer the root node to the bottom, mount the branch node to the root node, and mount the leaf node to the branch node, wherein the number of the root node is one, and the number of the branch nodes and the number of the leaf nodes are both several;

[0016] The root node, branch node and leaf node are transferred into a hierarchical structure, wherein the number of layers in the hierarchical structure is three, a multi-dimensional analysis strategy is introduced into the monitoring tree, and a communication link within the hierarchical structure is established.

[0017] Furthermore, the S100 further includes:

[0018] Establishing a growth mechanism for the hierarchical architecture, creating child nodes, and hanging them down to the leaf nodes;

[0019] Counting the number of first forwarded viewpoints in the branch node, obtaining the number of subscriptions of the user ID corresponding to the first forwarded viewpoint, and determining the propagation weight of the user ID according to different magnitudes of the number of subscriptions;

[0020] The user ID whose dissemination weight is greater than a preset threshold is defined as an opinion leader;

[0021] The sentiment value of the comment corresponding to the opinion leader is calculated, and if the sentiment value is less than a preset threshold, an emergency strategy is triggered.

[0022] Furthermore, the emergency strategy includes:

[0023] Establish a real-time sentiment monitoring system and embed a crisis management plan that includes, at a minimum, calling a response team, establishing a communications strategy, and coordinating media coverage;

[0024] Construct a multi-level threshold to obtain a first threshold and a second threshold; wherein the first threshold is less than the second threshold

[0025] When the emotion value is less than a second threshold, activating the real-time emotion monitoring system;

[0026] When the emotion value is less than a first threshold, the crisis management plan is triggered to start.

[0027] Further, the S200 includes:

[0028] Determining a propagation round of the initial event, and setting the number of child nodes using the propagation round;

[0029] Inserting a finalization mechanism into the hierarchy.

[0030] Further, the S300 includes:

[0031] Compare the user IDs whose sentiment values ​​are lower than the threshold and mark them as abnormal users;

[0032] The comments of the abnormal user are deleted from the preset platform, and the fluctuation range and average value are updated.

[0033] Furthermore, the system comprises:

[0034] The transfer module is used to establish a word set of thematic keywords, classify the word set into a preset data platform for matching, define the matching result as an initial event, extract the initial viewpoint, create a monitoring tree, and embed the hierarchical structure, hang the pre-built root node, branch node and leaf node layer by layer into the hierarchical structure, and transfer the initial event and initial viewpoint to the root node;

[0035] An extraction module is used to collect the browsing data and user ID of the initial event, extract the comments of the user ID on the initial event, obtain a first forwarding viewpoint, collect the browsing data and user ID of the first forwarding viewpoint, obtain a second forwarding viewpoint, and so on, obtain a third forwarding viewpoint, ..., an nth forwarding viewpoint, synchronize the first forwarding viewpoint to a branch node, and synchronize the second forwarding viewpoint to a leaf node;

[0036] A determination module is used to construct a sentiment word query dictionary, query the sentiment values ​​of all forwarded opinions, arrange the second forwarded opinions in the branch node in the order of the sentiment values ​​from small to large, determine the quartiles, define the interval composed of the first quartile and the third quartile as the fluctuation range, calculate the average value of the sentiment value within the fluctuation range, construct a number block corresponding to the hierarchical structure, transfer the fluctuation range and the average value into the number block, integrate the number block, and determine the linked list;

[0037] The comparison module is used to integrate a monitoring mechanism into the linked list, compare changes in sentiment values, and determine a disposal plan.

[0038] Furthermore, the transfer module includes:

[0039] A building unit, used to build a semantic network using the word set and optimize the matching results;

[0040] A configuration unit, used to format the matching results, extract the initial viewpoints using viewpoint mining technology, number the matching results, and configure the corresponding relationship between the matching results and the initial viewpoints;

[0041] A mounting unit, used to locate the bottom of the monitoring tree, transfer the root node to the bottom, mount the branch node to the root node, and mount the leaf node to the branch node, wherein the number of the root node is one, and the number of the branch nodes and the number of the leaf nodes are both several;

[0042] A construction unit, used to transfer the root node, branch node and leaf node into a hierarchical structure, wherein the number of layers in the hierarchical structure is three, introduce a multidimensional analysis strategy into the monitoring tree, and build a communication link within the hierarchical structure;

[0043] A hanging unit, used to establish a growth mechanism of the hierarchical architecture, create child nodes, and hang them down to the leaf nodes;

[0044] A statistical unit, used to count the number of first forwarded viewpoints in the branch node, obtain the number of subscriptions of the user ID corresponding to the first forwarded viewpoint, and determine the propagation weight of the user ID according to different magnitudes of the subscription number;

[0045] A definition unit, used to define a user ID whose dissemination weight is greater than a preset threshold as an opinion leader;

[0046] The trigger unit is used to calculate the sentiment value of the comment corresponding to the opinion leader, and if the sentiment value is less than a preset threshold, trigger the emergency strategy to start.

[0047] Furthermore, the extraction module includes:

[0048] A setting unit, used to determine a propagation round of the initial event, and set the number of child nodes using the propagation round;

[0049] A termination unit is used to insert a termination mechanism into the hierarchical structure.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] By defining the initial event and extracting the initial opinion, it is possible to timely discover the user's first reaction or most original opinion to the initial event. These opinions are of great significance for understanding the propagation path of the initial event, analyzing public opinion trends, and formulating subsequent public relations strategies. By constructing a monitoring tree, it is possible to monitor the spread of the initial event in real time, helping users to grasp the public opinion trends related to themselves in a timely manner, identify potential crisis signals, and make public relations responses quickly to avoid the spread and fermentation of negative information. It can also understand the current public concerns and emotional tendencies and adjust marketing strategies. Through the linked list, it can help users fully understand the complete path of the topic from its emergence to its spread, formulate more accurate marketing and public relations strategies, and provide decision-making support for users. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A flowchart of a thematic data processing method provided by an embodiment of the present invention;

[0053] Figure 2 A first sub-flow chart of the subject-oriented data processing method provided by an embodiment of the present invention;

[0054] Figure 3 A second sub-flow chart of the subject-oriented data processing method provided by an embodiment of the present invention;

[0055] Figure 4 A third sub-flow chart of the subject-oriented data processing method provided by an embodiment of the present invention;

[0056] Figure 5 A block diagram of the composition of a thematic data processing system provided by an embodiment of the present invention;

[0057] Figure 6 A block diagram of the composition of the transfer module in the subject data processing system provided by an embodiment of the present invention;

[0058] Figure 7 A block diagram of the composition of the extraction module in the subject-oriented data processing system provided by an embodiment of the present invention;

[0059] Figure 8 A block diagram of the composition of a determination module in a thematic data processing system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0061] In Example 1, Figure 1 The implementation process of the subject-oriented data processing method provided by the embodiment of the present invention is shown and described in detail below:

[0062] S100: Establish a word set of thematic keywords, classify the word set into a preset data platform for matching, define the matching result as an initial event, extract the initial viewpoint, create a monitoring tree, and embed a hierarchical architecture, and hang the pre-built root nodes, branch nodes, and leaf nodes layer by layer into the hierarchical architecture, and transfer the initial event and initial viewpoint to the root node.

[0063] Establish a word set of thematic keywords, where the word set is the collection of all keywords related to a certain thematic data, and transfer the set to a preset data platform, where the preset data platform can be a social platform such as Weibo and Moments, that is, a news platform such as News Network and NetEase News; compare the word set with the hot searches and news in the data platform, define the initial event, and determine the initial opinion based on the specific content of the comments and reposts of the initial event, where the initial opinion is the opinion of the user who first published the initial event; determine the hierarchical structure, first, transfer the root node to the hierarchical structure, hang the branch node under the root node, and then hang the leaf node under the branch node. It should be noted that the number of root nodes here is one, and the number of branch nodes and leaf nodes is multiple; integrate all root nodes, branch nodes and leaf nodes, create a monitoring tree, and finally, transfer the initial event and initial opinion to the root node.

[0064] For example, in the data platform, user A proposed event A for the first time. At the same time, he also expressed his opinion on event A. After the event was published, users B, C and D browsed event A, among which users B and D forwarded event A and continued to comment. Users E and F browsed the forwarding of users B or D, among which user E commented. In this event, event A is the initial event, and the initial opinion is A's opinion. Event A and user A's opinion are transferred to the root node, and two branch nodes are created. The comments of users B and D are transferred to the branch nodes, and the comment of E is transferred to the leaf node.

[0065] S200: Collect browsing data and user ID of the initial event, extract comments of the user ID on the initial event, obtain a first forwarding viewpoint, collect browsing data and user ID of the first forwarding viewpoint, obtain a second forwarding viewpoint, and so on, obtain a third forwarding viewpoint, ..., nth forwarding viewpoint, synchronize the first forwarding viewpoint to a branch node, and synchronize the second forwarding viewpoint to a leaf node.

[0066] Collect the browsing data and user ID of the initial event, extract the ID of the user who commented on the initial event, and determine the comment of the extracted user ID as the first forwarded viewpoint, collect the browsing data of the first forwarded viewpoint again, repeat the above steps to obtain the second forwarded viewpoint.

[0067] In the above example, the first forwarded opinion is the comment of users B and D, and the second forwarded opinion is the comment of user E.

[0068] S300: Construct a sentiment word query dictionary, query the sentiment values ​​of all forwarded opinions, arrange the second forwarded opinions in the branch node according to the sentiment value from small to large, determine the quartiles, and define the interval composed of the first quartile and the third quartile as the fluctuation range, and calculate the average value of the sentiment value within the fluctuation range, construct a number block corresponding to the hierarchical structure, and transfer the fluctuation range and the average value into the number block, integrate the number blocks, and determine the linked list.

[0069] Construct a sentiment word query dictionary to determine the sentiment values ​​of all forwarded opinions, where the forwarded opinions here include: the first forwarded opinion, the second forwarded opinion, etc. The number of the first forwarded opinion, the second forwarded opinion, etc. may be one or more. If there are more than one, they need to be sorted in descending order according to the sentiment value; for example, sort all the first forwarded opinions in the branch nodes.

[0070] In the sorting results, the first quartile and the third quartile are found, and the interval formed by the first quartile and the third quartile is determined as the fluctuation range, and the average value of the sentiment value within the fluctuation range is calculated.

[0071] In the above steps, it is also necessary to delete outliers in the sorting results, where outliers are greater than Q3+1.5IQR and Q1-1.5IQR, and IQR is the interquartile range.

[0072] Construct blocks corresponding to the hierarchical structure, that is, construct blocks corresponding to the root node, branch node and leaf node respectively; transfer the sentiment value of the initial opinion in the root node to the first block, transfer the fluctuation range and average value in the branch node to the second block, and transfer the fluctuation range and average value in the leaf node to the third block. The order of the blocks is based on the time when the comments are generated. Integrate all the generated blocks to generate a linked list.

[0073] S400: Integrate a monitoring mechanism into the linked list, compare changes in sentiment values, and determine a treatment plan.

[0074] A monitoring mechanism is integrated into the linked list, where the monitoring mechanism is a method of monitoring the fluctuation range and the average value. When the fluctuation range or the average value exceeds a certain interval, a disposal plan is triggered and initiated, where the disposal plan is prepared in advance by the staff.

[0075] In real life, by constructing a linked list, we can determine the user's sentiment value after each forwarding and comment on the initial event. This sentiment value also represents the user's sentiment trend towards the initial event.

[0076] In Example 2, Figure 2 The implementation process of the subject-oriented data processing method provided by the embodiment of the present invention is shown, and S100 is described in detail below, as follows:

[0077] S101: Utilize the word set to construct a semantic network and optimize the matching results.

[0078] According to the word set of keywords, a semantic network is constructed, wherein the semantic network is the relationship between keywords. The matching results of keywords are optimized by using the semantic network so as to more completely define the initial event.

[0079] S102: Format the matching results, extract the initial viewpoints using viewpoint mining technology, number the matching results, and configure the corresponding relationship between the matching results and the initial viewpoints.

[0080] The matching results are divided into formats. The matching results here are the initial events. The opinion mining technology in the prior art is used to extract the initial opinions in the initial events. The initial events and initial opinions are numbered. The advantage of this is that the corresponding relationship between the initial events and the initial opinions can be determined.

[0081] In Example 3, Figure 2 The implementation process of the subject-oriented data processing method provided by the embodiment of the present invention is shown, and S100 is further described in detail below, as follows:

[0082] S103: Locate the bottom of the monitoring tree, move the root node to the bottom, mount the branch node to the root node, and mount the leaf node to the branch node, wherein the number of the root node is one, and the number of the branch nodes and the number of the leaf nodes are both several.

[0083] Determine the bottom of the monitoring tree and move the root node to the bottom. In addition to storing the initial event and initial viewpoint, the root node can also coordinate and manage all branch nodes and leaf nodes; the branch nodes and leaf nodes are included in the monitoring tree.

[0084] According to practical experience, in a single thematic data processing process, the number of root nodes is one, while the number of branch nodes and leaf nodes is determined by actual conditions.

[0085] S104: transferring the root node, branch nodes and leaf nodes into a hierarchical structure, wherein the number of layers in the hierarchical structure is three, introducing a multi-dimensional analysis strategy into the monitoring tree, and building a communication link within the hierarchical structure.

[0086] Create a hierarchical structure. The hierarchical structure is the round of initial event propagation. A single forwarding and comment is regarded as a round, that is, a layer. In order to facilitate the representation of the hierarchical structure, the number of layers is assumed to be three, and a multidimensional analysis strategy is introduced into the monitoring tree. The multidimensional analysis strategy means that users from different industries or different age groups may have very different views on a certain event. This requires the use of a multidimensional analysis strategy to adjust the initial view, the first forwarding view, etc. The adjustment method can be to set different weights for different forwarding views. The setting of the communication link can manually correct the data in the root node, branch node or leaf node.

[0087] In Example 4, Figure 2 The implementation process of the subject-oriented data processing method provided by the embodiment of the present invention is shown, and S100 is further described in detail below, as follows:

[0088] S105: Establishing a growth mechanism of the hierarchical architecture, creating child nodes, and hanging them down to the leaf nodes.

[0089] Insert a growth mechanism into the hierarchical structure, in which the hierarchy of the child nodes is not limited. The growth mechanism is the propagation process of the initial event. When the second forwarded opinion in the leaf node continues to be forwarded and evaluated by users, the first child node should be generated and mounted to the leaf node. When the forwarded opinion in the first leaf node is forwarded and commented on, the second child node can be mounted in the first child node, and so on.

[0090] S106: Count the number of first forwarded viewpoints in the branch node, obtain the number of subscriptions of the user ID corresponding to the first forwarded viewpoint, and determine the propagation weight of the user ID according to different magnitudes of the subscription number.

[0091] Count the number of the first forwarded opinions in the branch node, which is how many users have viewed the initial opinion and forwarded the comment; at the same time, the number of subscriptions of the user ID should also be recorded, which is the number of fans in life. According to the number of fans, the communication weight of each user ID is determined.

[0092] S107: defining the user ID whose dissemination weight is greater than a preset threshold as an opinion leader.

[0093] User IDs with a dissemination weight greater than a preset threshold are defined as opinion leaders. Opinion leaders are also commonly known as Internet celebrities, and their comments are more likely to be widely disseminated.

[0094] S108: Calculate the sentiment value of the comment corresponding to the opinion leader, and if the sentiment value is less than a preset threshold, trigger the initiation of an emergency strategy.

[0095] When the opinion domain makes a comment on an event, the sentiment value of the comment is calculated. If the sentiment value is less than the preset threshold, the emergency strategy is initiated.

[0096] In Example 5, the implementation process of the subject-oriented data processing method provided by the embodiment of the present invention is described in detail below for the emergency strategy, as follows:

[0097] Establish a real-time sentiment monitoring system and embed a crisis management plan that includes, at a minimum, calling a response team, establishing a communications strategy, and coordinating media coverage;

[0098] Construct a multi-level threshold to obtain a first threshold and a second threshold; wherein the first threshold is less than the second threshold;

[0099] When the emotion value is less than a second threshold, activating the real-time emotion monitoring system;

[0100] When the emotion value is less than a first threshold, the crisis management plan is triggered to start.

[0101] Build a real-time emotion monitoring system and import it into the monitoring tree. At the same time, embed a crisis management plan into the real-time emotion monitoring system. When the emotion value of the opinion leader is less than the second threshold, activate the real-time emotion monitoring system to monitor the comments and reposts of the opinion leader, and promptly discover the dynamics of the opinion leader in the preset platform. When the emotion value is less than the first threshold, initiate the crisis management plan, which includes: calling the response team (public relations team), establishing communication with the opinion leader, or using the media to refute rumors, official releases, etc.

[0102] In Example 6, Figure 3 The implementation process of the subject-oriented data processing method provided by the embodiment of the present invention is shown, and S200 is further described in detail as follows:

[0103] S201: Determine the propagation round of the initial event, and use the propagation round to set the number of child nodes.

[0104] Determine the propagation rounds of the initial event, where the number of propagation rounds is equal to the number of hierarchical structures minus 1. By setting the propagation rounds, the number of child nodes can be better determined, where the number of child nodes is equal to the propagation rounds minus 2.

[0105] S202: Inserting a termination mechanism into the hierarchical architecture.

[0106] Insert a termination mechanism into the hierarchical structure, wherein the termination mechanism is that if there is no further forwarding and commenting on the initial event within a preset time period, the hierarchical structure will be terminated, thus ending the growth of the hierarchical structure.

[0107] In Example 7, Figure 4 The implementation process of the subject-oriented data processing method provided by the embodiment of the present invention is shown, and S300 is further described in detail below, as follows:

[0108] S301: compare the user IDs whose sentiment values ​​are lower than the threshold and mark them as abnormal users.

[0109] Find the user IDs whose sentiment values ​​are lower than the threshold from the monitoring tree and mark the corresponding user IDs as abnormal users.

[0110] S302: Deleting the abnormal user's comments from the preset platform, and updating the fluctuation range and average value.

[0111] The forwarded opinions corresponding to the abnormal user's comments are deleted from the monitoring tree, and the fluctuation range and average value are updated. It should be noted that by determining the fluctuation range and average value, the emotional trend of the user in the entire propagation process of the initial event can be determined.

[0112] Figure 5 The following is a structural block diagram of a themed data processing system according to an embodiment of the present invention. The themed data processing system 1 includes:

[0113] The transfer module 11 is used to establish a word set of thematic keywords, classify the word set into a preset data platform for matching, define the matching result as an initial event, extract the initial viewpoint, create a monitoring tree, and embed the hierarchical structure, hang the pre-built root node, branch node and leaf node layer by layer into the hierarchical structure, and transfer the initial event and initial viewpoint to the root node;

[0114] The extraction module 12 is used to collect the browsing data and user ID of the initial event, extract the comments of the user ID on the initial event, obtain the first forwarding viewpoint, collect the browsing data and user ID of the first forwarding viewpoint, obtain the second forwarding viewpoint, and so on, obtain the third forwarding viewpoint, ..., the nth forwarding viewpoint, synchronize the first forwarding viewpoint to the branch node, and synchronize the second forwarding viewpoint to the leaf node;

[0115] The determination module 13 is used to construct a sentiment word query dictionary, query the sentiment values ​​of all forwarded opinions, arrange the second forwarded opinions in the branch node in the order of the sentiment values ​​from small to large, determine the quartiles, define the interval composed of the first quartile and the third quartile as the fluctuation range, calculate the average value of the sentiment value within the fluctuation range, construct a number block corresponding to the hierarchical structure, transfer the fluctuation range and the average value into the number block, integrate the number block, and determine the linked list;

[0116] The comparison module 14 is used to integrate a monitoring mechanism into the linked list, compare the changes in the sentiment value, and determine a treatment plan.

[0117] Figure 6 The structure diagram of the subject-oriented data processing system provided by the embodiment of the present invention is shown, and the transfer module 11 includes:

[0118] A forming unit 111, used to form a semantic network using the word set and optimize the matching result;

[0119] A configuration unit 112, configured to format the matching results, extract the initial viewpoints using viewpoint mining technology, number the matching results, and configure the corresponding relationship between the matching results and the initial viewpoints;

[0120] The mounting unit 113 is used to locate the bottom of the monitoring tree, transfer the root node to the bottom, mount the branch node to the root node, and mount the leaf node to the branch node, wherein the number of the root node is one, and the number of the branch nodes and the leaf nodes are both several;

[0121] A building unit 114 is used to transfer the root node, branch node and leaf node into a hierarchical structure, wherein the number of layers in the hierarchical structure is three, introduce a multi-dimensional analysis strategy into the monitoring tree, and build a communication link within the hierarchical structure;

[0122] A hanging unit 115 is used to establish a growth mechanism of the hierarchical structure, create child nodes, and hang them down to the leaf nodes;

[0123] A statistical unit 116, configured to count the number of first forwarded viewpoints in the branch node, obtain the number of subscriptions of the user ID corresponding to the first forwarded viewpoint, and determine the propagation weight of the user ID according to different magnitudes of the subscription number;

[0124] A defining unit 117, configured to define a user ID whose dissemination weight is greater than a preset threshold as an opinion leader;

[0125] The trigger unit 118 is used to calculate the sentiment value of the comment corresponding to the opinion leader, and if the sentiment value is less than a preset threshold, trigger the emergency strategy to start.

[0126] Figure 7 The following is a structural block diagram of a themed data processing system provided by an embodiment of the present invention, wherein the themed data processing system comprises:

[0127] A setting unit 121, configured to determine a propagation round of the initial event, and set the number of child nodes using the propagation round;

[0128] A termination unit 122, configured to insert a termination mechanism into the hierarchical architecture;

[0129] Figure 8 The structure diagram of the subject-oriented data processing system provided by the embodiment of the present invention is shown, and the determination module 13 includes:

[0130] A marking unit 131, used to compare the user IDs whose sentiment values ​​are lower than a threshold value and mark them as abnormal users;

[0131] The updating unit 132 is used to delete the comments of the abnormal user from the preset platform and update the fluctuation range and the average value.

[0132] The transfer module 11 is mainly used to complete step S100, the extraction module 12 is mainly used to complete step S200, the determination module 13 is mainly used to complete step S300, and the comparison module 14 is mainly used to complete step S400;

[0133] The assembly unit 111 is mainly used to complete step S101, the configuration unit 112 is mainly used to complete step S102, the mounting unit 113 is mainly used to complete step S103, the construction unit 114 is mainly used to complete step S104, the hanging unit 115 is mainly used to complete step S105, the statistics unit 116 is mainly used to complete step S106, the definition unit 117 is mainly used to complete step S107, and the trigger unit 118 is mainly used to complete step S108;

[0134] The setting unit 121 is mainly used to complete step S201, and the termination unit 122 is mainly used to complete step S202;

[0135] The marking unit 131 is mainly used to complete step S301, and the updating unit 132 is mainly used to complete step S302.

[0136] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0137] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

[0138] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A themed data processing method, characterized in that: The method comprises: S100: Establish a word set of thematic keywords, classify the word set into a preset data platform for matching, define the matching result as an initial event, extract the initial viewpoint, create a monitoring tree, and embed it into a hierarchical structure, hang the pre-built root node, branch node and leaf node layer by layer into the hierarchical structure, and transfer the initial event and initial viewpoint into the root node; S200: Collect browsing data and user ID of the initial event, extract comments of the user ID on the initial event, obtain a first forwarding viewpoint, collect browsing data and user ID of the first forwarding viewpoint, obtain a second forwarding viewpoint, and so on, obtain a third forwarding viewpoint, ..., nth forwarding viewpoint, synchronize the first forwarding viewpoint to a branch node, and synchronize the second forwarding viewpoint to a leaf node; S300: construct a sentiment word query dictionary, query the sentiment values ​​of all forwarded opinions, arrange the second forwarded opinions in the branch node in the order of the sentiment values ​​from small to large, determine the quartiles, define the interval composed of the first quartile and the third quartile as the fluctuation range, calculate the average value of the sentiment value within the fluctuation range, construct a number block corresponding to the hierarchical structure, transfer the fluctuation range and the average value into the number block, integrate the number blocks, and determine a linked list; S400: Integrate a monitoring mechanism into the linked list, compare changes in sentiment values, and determine a treatment plan.

2. The themed data processing method according to claim 1, characterized in that: The S100 includes: Using the word set, a semantic network is constructed, and matching results are optimized; The matching results are formatted, initial viewpoints are extracted using viewpoint mining technology, the matching results are numbered, and the corresponding relationship between the matching results and the initial viewpoints is configured.

3. The themed data processing method according to claim 1, characterized in that: The S100 further includes: Locate the bottom of the monitoring tree, transfer the root node to the bottom, mount the branch node to the root node, and mount the leaf node to the branch node, wherein the number of the root node is one, and the number of the branch nodes and the number of the leaf nodes are both several; The root node, branch node and leaf node are transferred into a hierarchical structure, wherein the number of layers in the hierarchical structure is three, a multi-dimensional analysis strategy is introduced into the monitoring tree, and a communication link within the hierarchical structure is established.

4. The themed data processing method according to claim 3, characterized in that: The S100 further includes: Establishing a growth mechanism for the hierarchical architecture, creating child nodes, and hanging them down to the leaf nodes; Counting the number of first forwarded viewpoints in the branch node, obtaining the number of subscriptions of the user ID corresponding to the first forwarded viewpoint, and determining the propagation weight of the user ID according to different magnitudes of the number of subscriptions; The user ID whose dissemination weight is greater than a preset threshold is defined as an opinion leader; The sentiment value of the comment corresponding to the opinion leader is calculated, and if the sentiment value is less than a preset threshold, an emergency strategy is triggered.

5. The themed data processing method according to claim 4, characterized in that: The emergency response strategies include: Establish a real-time sentiment monitoring system and embed a crisis management plan that includes, at a minimum, calling a response team, establishing a communications strategy, and coordinating media coverage; Constructing a multi-level threshold to obtain a first threshold and a second threshold, wherein the first threshold is less than the second threshold; When the emotion value is less than a second threshold, activating the real-time emotion monitoring system; When the emotion value is less than a first threshold, the crisis management plan is triggered to start.

6. The themed data processing method according to claim 1, characterized in that: The S200 includes: Determining a propagation round of the initial event, and setting the number of child nodes using the propagation round; Inserting a finalization mechanism into the hierarchy.

7. The themed data processing method according to claim 4, characterized in that: The S300 includes: Compare the user IDs whose sentiment values ​​are lower than the threshold and mark them as abnormal users; The comments of the abnormal user are deleted from the monitoring tree, and the fluctuation range and average value are updated.

8. A themed data processing system, characterized in that: The system comprises: The transfer module is used to establish a word set of thematic keywords, classify the word set into a preset data platform for matching, define the matching result as an initial event, extract the initial viewpoint, create a monitoring tree, and embed the hierarchical structure, hang the pre-built root node, branch node and leaf node layer by layer into the hierarchical structure, and transfer the initial event and initial viewpoint to the root node; An extraction module is used to collect the browsing data and user ID of the initial event, extract the comments of the user ID on the initial event, obtain a first forwarding viewpoint, collect the browsing data and user ID of the first forwarding viewpoint, obtain a second forwarding viewpoint, and so on, obtain a third forwarding viewpoint, ..., an nth forwarding viewpoint, synchronize the first forwarding viewpoint to a branch node, and synchronize the second forwarding viewpoint to a leaf node; A determination module is used to construct a sentiment word query dictionary, query the sentiment values ​​of all forwarded opinions, arrange the second forwarded opinions in the branch node in the order of the sentiment values ​​from small to large, determine the quartiles, define the interval composed of the first quartile and the third quartile as the fluctuation range, calculate the average value of the sentiment value within the fluctuation range, construct a number block corresponding to the hierarchical structure, transfer the fluctuation range and the average value into the number block, integrate the number block, and determine the linked list; The comparison module is used to integrate a monitoring mechanism into the linked list, compare changes in sentiment values, and determine a disposal plan.

9. The themed data processing system according to claim 8, wherein the import module comprises: A building unit, used to build a semantic network using the word set and optimize the matching results; A configuration unit, used to format the matching results, extract the initial viewpoints using viewpoint mining technology, number the matching results, and configure the corresponding relationship between the matching results and the initial viewpoints; A mounting unit, used to locate the bottom of the monitoring tree, transfer the root node to the bottom, mount the branch node to the root node, and mount the leaf node to the branch node, wherein the number of the root node is one, and the number of the branch nodes and the number of the leaf nodes are both several; A construction unit, used to transfer the root node, branch node and leaf node into a hierarchical structure, wherein the number of layers in the hierarchical structure is three, introduce a multidimensional analysis strategy into the monitoring tree, and build a communication link within the hierarchical structure; A hanging unit, used to establish a growth mechanism of the hierarchical architecture, create child nodes, and hang them down to the leaf nodes; A statistical unit, used to count the number of first forwarded viewpoints in the branch node, obtain the number of subscriptions of the user ID corresponding to the first forwarded viewpoint, and determine the propagation weight of the user ID according to different magnitudes of the subscription number; A defining unit, used to define a user ID whose dissemination weight is greater than a preset threshold as an opinion leader; The trigger unit is used to calculate the sentiment value of the comment corresponding to the opinion leader, and if the sentiment value is less than a preset threshold, trigger the emergency strategy to start.

10. The themed data processing system according to claim 8, wherein the extraction module comprises: A setting unit, used to determine a propagation round of the initial event, and set the number of child nodes using the propagation round; A termination unit is used to insert a termination mechanism into the hierarchical structure.

Images