A multi-level response command method and system based on event characteristic judgment

Abstract

This invention discloses a multi-level response command method and system based on event characteristic judgment. The method includes filtering online information to obtain public opinion data; constructing a public opinion risk classification model; processing the public opinion data based on the public opinion risk classification model to obtain a public opinion risk value, and classifying public opinion according to the public opinion risk value; determining a response mechanism based on the public opinion classification; and constructing a command decision based on the response mechanism. This invention obtains public opinion data by filtering online information, classifies the public opinion data using the constructed public opinion risk classification model, establishes a graded response mechanism for public opinion data, and uses the response mechanism to assist in decision-making. This effectively improves the ability to detect and handle public opinion, reduces the time required for public opinion detection and handling, increases the control ability of relevant entities over public opinion, correctly guides the direction of online public opinion, and reduces the social impact of negative public opinion.

Description

Technical Field

[0001] This invention specifically relates to the field of public opinion management technology, and more specifically to a multi-level response command method and system based on event characteristic judgment. Background Technology

[0002] Due to the anonymity of online identities, the blurring of personal social boundaries, and the complexity of information, stimulating, unique, or conflict-inducing content often attracts the attention of many netizens and becomes a major form of online event. While there are relevant handling methods available, they cannot accurately predict the trajectory of events, meaning they cannot precisely determine the current stage of development or the appropriate approach. Improper handling can be counterproductive. Summary of the Invention

[0003] The purpose of this invention is to provide a multi-level response command method and system based on event characteristic judgment.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A multi-level response command method based on event characteristics includes the following steps:

[0006] Online information is filtered to obtain public opinion data, wherein the public opinion data is obtained by inputting the online information into a topic classification model. The public opinion data includes: public opinion release data, public opinion flow curve, and public opinion response data. The public opinion release data is used to characterize the characteristics of public opinion release, the public opinion release curve is used to characterize the characteristics of public opinion spread at different times, and the public opinion response data is used to characterize the response characteristics during the spread of public opinion.

[0007] A public opinion risk classification model is constructed, and a public opinion risk value is calculated based on the public opinion data based on the public opinion risk classification model, so as to classify public opinion according to the public opinion risk value;

[0008] The response mechanism is determined based on the public opinion classification, and command and decision-making are constructed based on the response mechanism.

[0009] As a further aspect of the present invention: to filter network information to obtain public opinion data, the method further includes: denoising and segmenting the collected network information into Chinese words, and using Boolean model or vector space model to filter and aggregate web page information by topic in order to form a unified text feature format.

[0010] As a further aspect of the present invention: the public opinion risk classification model is as follows:

[0011] ∑α i =1,∑K i≤100, i≤N, N=3, where K represents a risk assessment item, which includes the risk assessment value of the release of public opinion, the risk assessment value of the flow, and the risk assessment value of the reaction. α represents a risk assessment factor, which is determined based on the corresponding risk assessment values ​​of the release, the flow, and the reaction. The risk assessment value of the release is determined based on the public opinion release data, the risk assessment value of the flow is determined based on the public opinion flow curve, and the risk assessment value of the reaction is determined based on the public opinion reaction data.

[0012] As a further aspect of the present invention: the risk value F is 0-100. When the risk value is in the range of 0-30, the risk level of the public opinion data is the first risk level; when the risk value is in the range of 31-60, the risk level of the public opinion data is the second risk level; and when the risk value is in the range of 61-100, the risk level of the public opinion data is the third risk level.

[0013] As a further aspect of the present invention: the risk assessment factor corresponding to the release risk assessment value is determined based on the source distortion rate, content error rate, and content controversy rate; the risk assessment factor corresponding to the flow risk assessment value is determined based on the possibility of flow breadth and the possibility of flow depth; the risk assessment factor corresponding to the reaction risk assessment value is determined based on the audience's subjective attention rate and the audience's negative reaction rate.

[0014] This invention also provides a multi-level response command system based on event characteristic judgment, comprising:

[0015] The public opinion filtering module is used to filter online information to obtain public opinion data. The public opinion data is obtained by inputting the online information into a topic classification model. The public opinion data includes: public opinion release data, public opinion flow curve, and public opinion response data. The public opinion release data is used to characterize the characteristics of public opinion release, the public opinion release curve is used to characterize the characteristics of public opinion spread at different times, and the public opinion response data is used to characterize the response characteristics during the spread of public opinion.

[0016] The public opinion classification module is used to construct a public opinion risk classification model, and calculate the public opinion risk value based on the public opinion risk classification model, so as to classify the public opinion according to the public opinion risk value.

[0017] The public opinion response module is used to determine the response mechanism based on the public opinion classification and to construct command and decision-making based on the response mechanism.

[0018] As a further aspect of the present invention, it also includes a data preprocessing module, which is used to denoise and segment Chinese words in the collected network information, and to perform topic filtering and aggregation on web page information using Boolean model or vector space model to form a unified text feature format.

[0019] As a further aspect of the present invention: the public opinion risk classification model is as follows:

[0020] ∑α i =1,∑K i =100, i≤N, N=3, where K represents a risk assessment item, which includes the risk assessment value of the release of public opinion, the risk assessment value of liquidity, and the risk assessment value of the reaction. α represents a risk assessment factor, which is determined based on the corresponding risk assessment value of the release of public opinion, the risk assessment value of liquidity, and the risk assessment value of the reaction.

[0021] As a further aspect of the present invention: the risk assessment factor corresponding to the release risk assessment value is determined based on the source distortion rate, content error rate, and content controversy rate; the risk assessment factor corresponding to the flow risk assessment value is determined based on the possibility of flow breadth and the possibility of flow depth; the risk assessment factor corresponding to the reaction risk assessment value is determined based on the audience's subjective attention rate and the audience's negative reaction rate.

[0022] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention obtains public opinion data by screening network information, classifies the public opinion data using a constructed public opinion risk classification model, establishes a graded response mechanism for public opinion data, and uses the response mechanism to assist in decision-making, thereby effectively improving the ability to discover and process public opinion, reducing the time for discovering and processing public opinion, increasing the control ability of relevant entities over public opinion, correctly guiding the direction of online public opinion, and reducing the social impact of negative public opinion. Attached Figure Description

[0023] Figure 1 This is a flowchart of a multi-level response command method based on event characteristics.

[0024] Figure 2 This is a flowchart of step S10 in the multi-level response command method based on event characteristics.

[0025] Figure 3 This is a block diagram of a multi-level response command system based on event characteristics. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0027] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0028] It should be understood that although the terms "first," "second," etc., may be used in the embodiments of the present invention to describe various information, this information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another.

[0029] Based on this, please refer to Figures 1-2 In this embodiment of the invention, a multi-level response command method based on event characteristic judgment includes the following steps:

[0030] S10. Filter online information to obtain public opinion data, wherein the public opinion data is obtained by inputting the online information into a topic classification model, and the public opinion data includes: public opinion release data, public opinion flow curve, and public opinion response data. The public opinion release data is used to characterize the characteristics of public opinion release, the public opinion release curve is used to characterize the characteristics of public opinion spread at different times, and the public opinion response data is used to characterize the response characteristics during the spread of public opinion.

[0031] In step S10 of this embodiment of the invention, the method for filtering network information to obtain public opinion data includes the following steps:

[0032] S11. Using web crawler technology, network information with a target topic is actively crawled to the local machine and stored. In this embodiment, the network information collection based on the topic adopts web page data crawling that supports custom URLs in order to obtain the network information with the target topic required by the user.

[0033] S12. The network information with the target topic is preprocessed to obtain preprocessed data. In this embodiment, the information preprocessing will denoise and segment Chinese words in the collected information, and then use Boolean model or vector space model to filter and aggregate the web page information by topic, and unify the format of text features.

[0034] It should be noted that Chinese word segmentation refers to dividing a sequence of Chinese characters into individual words for semantic recognition. Currently, commonly used Chinese word segmentation methods include string matching, statistical methods, and understanding methods. Extracting the main text information from collected web pages requires developing corresponding collection rules and constructing classifiers for different websites to achieve automatic extraction of web page content. To facilitate subsequent analysis, the segmented content needs to use specific text representation techniques. Chinese web pages also need to be converted to a unified encoding format to facilitate the representation and extraction of text features. Extraction techniques based on page content and topic relevance are used, employing methods based on web page frames, word frequency, semantics, clustering, and vector space.

[0035] S13. Perform topic identification on the preprocessed data to determine whether the network information is public opinion data. The topic identification method is generally implemented by text classification method. The process is divided into two steps: first, the topic model is trained to obtain the topic classification model, and then the subsequent documents are classified through the classification model to generate classification results. In this embodiment, the topic classification model can adopt a neural network model.

[0036] S20. Construct a public opinion risk classification model, and calculate the public opinion risk value based on the public opinion data according to the public opinion risk classification model, so as to classify the public opinion according to the public opinion risk value; it should be noted that classifying public opinion events means classifying various public opinion events in the network information according to the severity of the event, that is, the level of risk, so as to facilitate the subsequent construction of a response mechanism based on the risk level.

[0037] The specific public opinion risk classification model is as follows:

[0038] ∑α i =1,∑K i ≤100, i≤N, N=3, where K represents a risk assessment item, which includes the risk assessment value of the release of public opinion, the risk assessment value of the flow, and the risk assessment value of the reaction. α represents a risk assessment factor, which is determined based on the corresponding risk assessment values ​​of the release, the flow, and the reaction. The risk assessment value of the release is determined based on the public opinion release data, the risk assessment value of the flow is determined based on the public opinion flow curve, and the risk assessment value of the reaction is determined based on the public opinion reaction data.

[0039] The aforementioned risk value F ranges from 0 to 100. When the risk value is in the range of 0 to 30, the risk level of the public opinion data is the first risk level. In this embodiment of the invention, the first risk level is regarded as public opinion. When the risk value is in the range of 31 to 60, the risk level of the public opinion data is the second risk level. In this embodiment of the invention, the second risk level is regarded as negative public opinion. When the risk value is in the range of 61 to 100, the risk level of the public opinion data is the third risk level. In this embodiment of the invention, the third risk level is regarded as major public opinion.

[0040] The risk assessment factor corresponding to the release risk assessment value is determined based on the source distortion rate, content error rate, and content controversy rate. In this assessment, the source distortion rate is judged based on historical data, that is, the proportion of distorted information in messages disseminated through this source in the past. For example, if 10 out of 100 messages disseminated by this source were distorted, the source distortion rate is 10%. The content error rate is determined based on the current public opinion content to determine the degree to which the source deviates from the truth. For example, if the ratio of true to false information in the disseminated messages is 1:1, the content error rate is 50%. Content controversy indicates the degree of debate caused by ambiguity in discourse. It should be noted that the content controversy index is mainly determined by the content error rate. The lower the content error rate, the greater the authenticity, and the lower the possibility of controversy, and vice versa. The risk assessment factor corresponding to the release risk assessment value = (1 - source distortion rate) * (1 - content error rate) * (1 - content controversy rate).

[0041] The risk assessment factor corresponding to the flow risk assessment value is determined based on the probability of flow breadth and the probability of flow depth. Specifically, the probability of flow breadth is determined by the flow time above the preset heat line in the public opinion flow curve. That is, if public opinion is at a high level for a relatively long time, it indicates a large spread. The probability of flow depth is determined by the number of inflection points above the preset heat line. The more inflection points, the more likely the public opinion will have secondary spread, and the higher the probability of continued spread. When calculating, the risk assessment factor corresponding to the flow risk assessment value = (flow time above the preset heat line / total spread time) * number of inflection points above the preset heat line.

[0042] The risk assessment factor corresponding to the reaction risk assessment value is determined based on the audience's subjective attention rate and the audience's negative reaction rate. The higher the audience's subjective attention rate, the greater the risk of public opinion dissemination, while the audience's negative reaction rate indicates a higher possibility of negative emotions arising from public opinion. The audience's subjective attention rate is calculated based on user search volume and public opinion exposure volume, while the audience's negative reaction rate is determined based on the negative reaction volume and the overall reaction volume. The risk assessment factor corresponding to the reaction risk assessment value = audience's subjective attention rate + audience's negative reaction rate.

[0043] It should be noted that the above evaluation factors need to be normalized before being substituted into the model for calculation.

[0044] S30. Determine the response mechanism based on the level of public opinion, and construct command and decision-making based on the response mechanism.

[0045] In step S30 of the present invention, the response mechanism includes a first response mechanism corresponding to the first risk level, which is a response mechanism; a second response mechanism corresponding to the second risk level, which is a guidance mechanism; and a third response mechanism corresponding to the third risk level, which is a processing mechanism.

[0046] Furthermore, the response method of the response mechanism includes the following steps:

[0047] Step 1: Obtain the URL address of public opinion data;

[0048] Step 2: Respond to the public opinion data at the webpage location with the URL address. The response processing includes deleting, receiving, commenting, following, and forwarding the public opinion data.

[0049] The response method of the guidance mechanism is as follows: clarify the public opinion data. Clarification channels include newspapers, radio, television and the Internet. Clarification is carried out through mainstream media, statements from relevant institutions and reprints by major portal websites to correctly guide the direction of online public opinion.

[0050] The response method of the processing mechanism includes the following steps:

[0051] Step 1: Obtain public opinion data related to the specified domain; the specified domain is provided by a first entity.

[0052] Step 2: Generate corresponding handling plan information based on the public opinion data. The handling plan information includes quickly activating the online public opinion emergency plan, taking effective technical measures to suppress the situation, using correct public opinion propaganda to avoid adverse effects, and handling relevant responsible persons.

[0053] Step 3: Send the processing plan information to the first entity so that the first entity can take corresponding actions based on the processing plan information.

[0054] Please see Figure 3 The present invention also discloses a multi-level response command system based on event characteristic judgment, comprising:

[0055] The public opinion filtering module 100 is used to filter online information to obtain public opinion data. The public opinion data is obtained by inputting the online information into a topic classification model. The public opinion data includes: public opinion release data, public opinion flow curve, and public opinion response data. The public opinion release data is used to characterize the characteristics of public opinion release, the public opinion release curve is used to characterize the characteristics of public opinion spread at different times, and the public opinion response data is used to characterize the response characteristics during the spread of public opinion.

[0056] The public opinion classification module 200 is used to construct a public opinion risk classification model, and calculate the public opinion risk value based on the public opinion risk classification model, so as to classify the public opinion according to the public opinion risk value.

[0057] The public opinion response module 300 is used to determine the response mechanism based on the public opinion classification and to construct command and decision-making based on the response mechanism.

[0058] As a further aspect of the present invention, it also includes a data preprocessing module, which is used to denoise and segment Chinese words in the collected network information, and to perform topic filtering and aggregation on web page information using Boolean model or vector space model to form a unified text feature format.

[0059] As a further aspect of the present invention: the public opinion risk classification model is as follows:

[0060] ∑α i =1,∑K i =100, i≤N, N=3, where K represents a risk assessment item, which includes the risk assessment value of the release of public opinion, the risk assessment value of liquidity, and the risk assessment value of the reaction. α represents a risk assessment factor, which is determined based on the corresponding risk assessment value of the release of public opinion, the risk assessment value of liquidity, and the risk assessment value of the reaction.

[0061] As a further aspect of the present invention: the risk assessment factor corresponding to the release risk assessment value is determined based on the source distortion rate, content error rate, and content controversy rate; the risk assessment factor corresponding to the flow risk assessment value is determined based on the possibility of flow breadth and the possibility of flow depth; the risk assessment factor corresponding to the reaction risk assessment value is determined based on the audience's subjective attention rate and the audience's negative reaction rate.

[0062] Furthermore, some embodiments may include a storage medium having a program for executing the methods described herein on a computer, having stored thereon at least one instruction, at least one program segment, code set, or instruction set, which, when loaded and executed by a processor, implements the steps in the above-described method embodiments. Examples of computer-readable recording media include hardware devices specifically configured for storing and executing program commands: magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floppy disks; and ROM, RAM, flash memory, etc. Examples of program commands may include machine language code written by a compiler and high-level language code executed by a computer using an interpreter, etc.

[0063] Those skilled in the art will understand that implementing all or part of the processes in the methods of the above embodiments can be accomplished by instructing related hardware with at least one instruction, at least one program, code set, or instruction set. The at least one instruction, at least one program, code set, or instruction set can be stored in a non-volatile computer-readable storage medium. When executed, the at least one instruction, at least one program, code set, or instruction set can include the processes of the embodiments of the above methods. Furthermore, any references to memory, storage, database, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory.

[0064] In summary, this invention obtains public opinion data by filtering online information, classifies the public opinion data using a constructed public opinion risk classification model, establishes a graded response mechanism for public opinion data, and uses the response mechanism to assist in decision-making. This effectively improves the ability to discover and process public opinion, reduces the time required for discovery and processing, increases the control of relevant entities over public opinion, correctly guides the direction of online public opinion, and reduces the social impact of negative public opinion.

[0065] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0066] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A multi-level response command method based on event characteristic judgment, characterized in that, Includes the following steps: Online information is filtered to obtain public opinion data, wherein the public opinion data is obtained by inputting the online information into a topic classification model. The public opinion data includes: public opinion release data, public opinion flow curve, and public opinion response data. The public opinion release data is used to characterize the characteristics of public opinion release, the public opinion release curve is used to characterize the characteristics of public opinion spread at different times, and the public opinion response data is used to characterize the response characteristics during the spread of public opinion. A public opinion risk classification model is constructed, and the public opinion data is processed based on the public opinion risk classification model to obtain a public opinion risk value, so as to classify the public opinion based on the public opinion risk value; A response mechanism is determined based on the public opinion classification, and a command and decision-making system is constructed based on the response mechanism. The public opinion risk classification model is as follows: ∑α i =1,∑K i ≤100, i≤N, N=3, where K represents a risk assessment item, which includes the risk assessment value of the release of public opinion, the risk assessment value of the flow, and the risk assessment value of the reaction. α represents a risk assessment factor, which is determined based on the corresponding risk assessment values ​​of the release, the flow, and the reaction. The risk assessment value of the release is determined based on the public opinion release data, the risk assessment value of the flow is determined based on the public opinion flow curve, and the risk assessment value of the reaction is determined based on the public opinion reaction data. The risk assessment factors corresponding to the release risk assessment value are determined based on the source distortion rate, content error rate, and content controversy rate; the risk assessment factors corresponding to the flow risk assessment value are determined based on the probability of flow breadth and the probability of flow depth; the risk assessment factors corresponding to the reaction risk assessment value are determined based on the audience's subjective attention rate and the audience's negative reaction rate. The source distortion rate is judged based on historical data, the content error rate is determined based on the current public opinion content to determine the degree to which the source deviates from the truth, the content controversy rate is determined by the content error rate, and the risk assessment factor corresponding to the release risk assessment value is (1-source distortion rate)*(1-content error rate)*(1-content controversy rate). The breadth of the flow is determined based on the time the flow takes above the preset heat line in the public opinion flow curve, while the depth of the flow is determined based on the number of inflection points that occur above the preset heat line. The risk assessment factor corresponding to the flow risk assessment value is (the time the flow takes above the preset heat line / the total time of dissemination) * the number of inflection points that occur above the preset heat line. The audience's subjective attention rate is calculated based on user search volume and public opinion exposure. The audience's negative reaction rate is determined based on the negative reaction volume and the overall reaction volume. The risk assessment factor corresponding to the reaction risk assessment value = audience's subjective attention rate + audience's negative reaction rate.

2. The multi-level response command method based on event characteristic judgment according to claim 1, characterized in that, The process of filtering online information to obtain public opinion data previously included: denoising the collected online information, segmenting Chinese words, and using Boolean models or vector space models to filter and aggregate web page information by topic in order to form a unified text feature format.

3. The multi-level response command method based on event characteristic judgment according to claim 1, wherein the risk value F is 0-100, and when the risk value is in the range of 0-30, the risk level of the public opinion data is the first risk level; When the risk value is in the range of 31-60, the risk level of public opinion data is the second risk level. When the risk value is in the range of 61-100, the risk level of public opinion data is the third risk level.

4. The multi-level response command method based on event characteristic judgment according to claim 3, characterized in that, The response mechanism includes a first response mechanism corresponding to the first risk level, which is a response mechanism; and a second response mechanism corresponding to the second risk level, which is a guidance mechanism. The third response mechanism corresponds to the third risk level; the third response mechanism is a handling mechanism.

5. A multi-level response command system based on event characteristic judgment, characterized in that, include: The public opinion filtering module is used to filter online information to obtain public opinion data. The public opinion data is obtained by inputting the online information into a topic classification model. The public opinion data includes: public opinion release data, public opinion flow curve, and public opinion response data. The public opinion release data is used to characterize the characteristics of public opinion release, the public opinion release curve is used to characterize the characteristics of public opinion spread at different times, and the public opinion response data is used to characterize the response characteristics during the spread of public opinion. The public opinion classification module is used to construct a public opinion risk classification model, and calculate the public opinion risk value based on the public opinion risk classification model, so as to classify the public opinion according to the public opinion risk value. The public opinion response module is used to determine the response mechanism based on the public opinion classification and to construct command and decision-making based on the response mechanism. The public opinion risk classification model is as follows: ∑α i =1,∑K i ≤100, i≤N, N=3, where K represents a risk assessment item, which includes the risk assessment value of the release of public opinion, the risk assessment value of the flow, and the risk assessment value of the reaction. α represents a risk assessment factor, which is determined based on the corresponding risk assessment values ​​of the release, the flow, and the reaction. The risk assessment value of the release is determined based on the public opinion release data, the risk assessment value of the flow is determined based on the public opinion flow curve, and the risk assessment value of the reaction is determined based on the public opinion reaction data. The risk assessment factors corresponding to the release risk assessment value are determined based on the source distortion rate, content error rate, and content controversy rate; the risk assessment factors corresponding to the flow risk assessment value are determined based on the probability of flow breadth and the probability of flow depth; the risk assessment factors corresponding to the reaction risk assessment value are determined based on the audience's subjective attention rate and the audience's negative reaction rate. The source distortion rate is judged based on historical data, the content error rate is determined based on the current public opinion content to determine the degree to which the source deviates from the truth, the content controversy rate is determined by the content error rate, and the risk assessment factor corresponding to the release risk assessment value is (1-source distortion rate)*(1-content error rate)*(1-content controversy rate). The breadth of the flow is determined based on the time the flow takes above the preset heat line in the public opinion flow curve, while the depth of the flow is determined based on the number of inflection points that occur above the preset heat line. The risk assessment factor corresponding to the flow risk assessment value is (the time the flow takes above the preset heat line / the total time of dissemination) * the number of inflection points that occur above the preset heat line. The audience's subjective attention rate is calculated based on user search volume and public opinion exposure. The audience's negative reaction rate is determined based on the negative reaction volume and the overall reaction volume. The risk assessment factor corresponding to the reaction risk assessment value = audience's subjective attention rate + audience's negative reaction rate.

6. The multi-level response command system based on event characteristic judgment according to claim 5, characterized in that, It also includes a data preprocessing module, which is used to denoise and segment Chinese words in the collected network information, and to use Boolean model or vector space model to filter and aggregate web page information by topic in order to form a unified text feature format.

Images