WO2019196209A1

WO2019196209A1 - Event information analysis method, readable storage medium, terminal device and apparatus

Info

Publication number: WO2019196209A1
Application number: PCT/CN2018/093346
Authority: WO
Inventors: 陈一恋; 汪伟; 王晓伟; 罗傲雪; 肖京
Original assignee: 平安科技（深圳）有限公司
Priority date: 2018-04-08
Filing date: 2018-06-28
Publication date: 2019-10-17
Also published as: CN108763272A; CN108763272B

Abstract

The present application belongs to the technical field of computers, and relates in particular to an event information analysis method, a computer-readable storage medium, a terminal device and an apparatus. The method comprises: first, acquiring initial search results corresponding to a preset initial keyword by means of a preset network search engine, and screening extension keywords out from the initial search results; next, acquiring extension search results corresponding to the extension keywords by means of the network search engine, and then extracting target event sentences in the initial search results and the extension search results; and finally, matching the target event sentences by means of a preset regular expression, and if the matching is successful, determining matching fields in the target event sentences to be event main-bodies which correspond to the target event sentences. Since extension keywords are introduced, broader search results may be obtained, automatic matching for event main-bodies is achieved by using the regular expression, and analysis efficiency is thus greatly improved.

Description

Event information analysis method, readable storage medium, terminal device and device

The present application claims priority to Chinese Patent Application No. 201101305412.4, entitled "An Event Information Analysis Method, Computer Readable Storage Medium, and Terminal Equipment", filed on April 8, 2018, the entire disclosure of which is incorporated herein by reference. The content is incorporated herein by reference.

Technical field

The present application belongs to the field of computer technology, and in particular, to an event information analysis method, a computer readable storage medium, a terminal device and a device.

Background technique

With the continuous development of the social economy, people are increasingly aware of the importance of information, and whether it can quickly and accurately obtain information has become one of the important factors affecting the success of enterprises. Among them, the analysis and sorting of various news events in the Internet is an effective way to obtain information. However, the search results obtained by simple keyword search are often limited and underrepresented, and these searches are obtained. After the results, it is often necessary to further analyze the main body of the event, for example, specific companies, organizations, institutions, etc., but at present the analysis is mainly done by professionals through manual analysis, and the analysis efficiency is low.

technical problem

In view of this, the embodiment of the present application provides an event information analysis method, a computer readable storage medium, a terminal device, and a device, so as to solve the problem that the search result obtained by the existing event information analysis method is limited and the analysis efficiency is low. .

Technical solution

A first aspect of the embodiment of the present application provides an event information analysis method, which may include:

Acquiring initial search results corresponding to preset initial keywords by a preset web search engine;

An extended keyword is selected in the initial search result, where the extended keyword is a word whose similarity with the initial keyword is greater than a preset similarity threshold;

Acquiring an extended search result corresponding to the extended keyword by the network search engine;

Extracting the initial search result and the target event statement in the extended search result, where the target event statement is a statement including an event keyword and a preset matching field, where the event keyword is the initial keyword or Extended keyword

Matching the target event statement by a preset regular expression;

If the matching is successful, the matching field in the target event statement is determined as the event body corresponding to the target event statement.

A second aspect of embodiments of the present application provides a computer readable storage medium storing computer readable instructions that, when executed by a processor, implement the event information analysis method described above step.

A third aspect of an embodiment of the present application provides an event information analysis terminal device including a memory, a processor, and computer readable instructions stored in the memory and executable on the processor, the processor executing The computer readable instructions implement the steps of the event information analysis method described above.

A fourth aspect of the embodiments of the present application provides an event information analyzing apparatus, which may include a module for implementing the steps of the event information analyzing method.

Beneficial effect

Compared with the prior art, the embodiment of the present application has the beneficial effects that: based on the initial keyword, the embodiment of the present application further introduces an extended keyword to expand it, and can obtain a broader search result, and The use of regular expressions achieves automatic matching of event subjects, greatly improving the efficiency of analysis.

DRAWINGS

FIG. 1 is a flowchart of an embodiment of an event information analysis method according to an embodiment of the present application;

2 is a schematic flow chart of an initial search result storage process;

3 is a schematic flow chart of filtering out extended keywords in initial search results;

4 is a schematic flow chart of a process of selecting an extended search result;

FIG. 5 is a structural diagram of an embodiment of an event information analysis apparatus according to an embodiment of the present application;

FIG. 6 is a schematic block diagram of an event information analysis terminal device according to an embodiment of the present application.

Embodiments of the invention

Referring to FIG. 1, an embodiment of an event information analysis method in an embodiment of the present application may include:

Step S101: Acquire an initial search result corresponding to the preset initial keyword by using a preset web search engine.

In this embodiment, one or more network search engines may be used to perform automatic search in the Internet according to actual conditions. When searching, you can specify the search scope and search only under certain websites. For example, if you need to search for financial information, you can specify one or more financial websites as the search scope, only in the search scope. Search within, or you can search without searching for the entire Internet.

The initial keyword can be determined according to the actual analysis field. For example, in the field of quantitative investment, the investor is more concerned about the debt situation of the investor, especially if the debt default of the investor is used, the “debt default” can be adopted as the The initial keyword is used to search to obtain related webpage content, that is, the initial search result.

It should be noted that the initial search results obtained by the search may be extremely large, and if all of the content is stored, it will consume huge storage resources. Therefore, in the present embodiment, the number of the initial search results is set in advance, and it is recorded as PageNum, and only the search results within the number are stored. The value of PageNum can be determined according to the storage capacity of the preset storage medium. The relationship between the two is positive, that is, the larger the storage capacity, the larger the value of PageNum. Conversely, the smaller the storage capacity, the value of PageNum is also The smaller.

The specific storage process may include the steps shown in FIG. 2:

Step S1011: Perform a hash operation on the initial search result to obtain a hash value of the initial search result.

In this embodiment, a method of hashing the complete content of the webpage may be adopted, but such an operation process consumes a large amount of time. Therefore, for the sake of simplicity, only the digest content of the webpage may be hashed. The way to speed up the operation, specifically:

First, the summary content of the initial search result is obtained according to the following formula:

SubContent=Head(PageContent)∪Tail(PageContent)

Wherein, PageContent is the webpage text in the initial search result, Head (PageContent) is the first M characters of the webpage text in the initial search result, and Tail (PageContent) is the post of the webpage text in the initial search result. N characters, M and N are integers greater than 1, and SubContent is the summary content of the initial search result.

Then, the hash value of the initial search result is calculated according to the following formula:

Key=Hash(SubContent)=Hash[Head(PageContent)∪Tail(PageContent)]

The hash is a preset hash function, and the Key is a hash value of the initial search result.

Step S1012: Search for a hash value of the initial search result in a preset hash value set.

The hash value set is used to record the hash value of the webpage that has been stored in the preset storage medium, and the calculation process of each of the hash values is similar to that in step S1011. For details, refer to the content in step S1011. , will not repeat them here.

If the search fails, that is, the hash value of the initial search result is not found in the hash value set, step S1013 is performed, and if the search is successful, the search is found in the hash value set. The hash value of the initial search result is then executed in step S1014.

Step S1013: Add a hash value of the initial search result into the hash value set, and store the initial search result in the storage medium.

If the hash value set is represented by a HashList, the process of adding the hash value of the initial search result to the hash value set may be expressed as:

HashList=HashList∪Key.

In step S1014, the initial search result is discarded.

If the hash value of the initial search result is found in the hash value set, it indicates that the same content as the initial search result has been stored in the storage medium, and it is not necessary to store it again.

Step S102: Filter out the extended keyword in the initial search result.

The extended keyword is a word whose similarity with the initial keyword is greater than a preset similarity threshold.

Specifically, step S102 may include the steps as shown in FIG. 3:

Step S1021, respectively calculating a degree of literal overlap between each word in the initial search result and the initial keyword.

For example, the literal overlap between each word in the initial search result and the initial keyword may be separately calculated according to the following formula:

Where w is any word in the initial search result, For the initial keyword,

For w and

The number of words that are included together,

The number of words contained in w and

The maximum number of words included,

For w with

The literal overlap between them.

Step S1022, respectively calculating a search overlap degree between each word in the initial search result and the initial keyword.

For example, the search overlap between each word in the initial search result and the initial keyword may be separately calculated according to the following formula:

among them,

For the initial search results in w and

The number of pages that appear together,

The number of pages that appear in the initial search results and

The maximum number of pages that appear,

For w with

The degree of overlap between searches.

Step S1023, respectively calculating the similarity between each word in the initial search result and the initial keyword.

For example, the similarity between each word in the initial search result and the initial keyword may be separately calculated according to the following formula:

Where k ₁ and k ₂ are preset weight coefficients, and k ₁ +k ₂ =1,

For w with

The similarity between the two.

Step S1024, determining a word whose similarity with the initial keyword is greater than the similarity threshold as the extended keyword.

For example, if the initial keyword is “debt default”, through the above process, it can be determined that the extended keywords are “debt dispute”, “debt litigation”, “debt storm”, “debt collapse”, “debt warning”. , "debt rights protection" and other similar words.

Step S103: Acquire an extended search result corresponding to the extended keyword by using the network search engine.

It should be noted that the search results obtained by the search may be extremely large, and if all of the content is stored, it will consume huge storage resources. Therefore, in this embodiment, only a part of the extended search result can be selected and stored by the steps shown in FIG. 4:

In step S1031, the importance scores of the respective extended keywords are respectively calculated.

For example, the importance scores of each of the extended keywords may be separately calculated according to the following formula:

Where ew is any of the extended keywords, freq(ew) is the frequency at which ew appears in the initial search result, and Freq(ew) is the frequency at which ew appears in the preset sample corpus. This frequency is passed. The large-scale statistics of the language materials that have actually appeared in the actual use of the language are fixed, and can be obtained directly by looking up the table, and ExWord is a collection composed of each of the extended keywords. Max[Freq(ExWord)] is the maximum value of the frequency of occurrence of each of the extended keywords in the sample corpus, namely:

Max[Freq(ExWord)]=max[Freq(ew ₁ ), Freq(ew ₂ ), ..., Freq(ew _s ), ..., Freq(ew _S )]

ew _s s is the number of keyword expansion, 1≤s≤S, S is the number of the expanded keyword, ln degree of importance scores logarithmic function, Score (ew) of ew is a natural.

It can be seen from the above process that the importance score of an extended keyword is positively correlated with the frequency of occurrence in the initial search result, and negatively correlated with the frequency of occurrence in the sample corpus. That is to say, if the frequency of occurrence of an extended keyword in normal language use is less, and the more frequently it appears in the initial search result, the importance score is higher.

Step S1032, respectively calculating the number of cuts of the extended search results corresponding to the respective extended keywords.

For example, the number of interception of the extended search result corresponding to each of the extended keywords may be separately calculated according to the following formula:

Where α is a preset scale factor, PageNum is the preset number of the initial search results, and ExPageNum(ew) is the number of cuts of the extended search result corresponding to ew.

Step S1033: Obtain an extended search result corresponding to each of the extended keywords according to the intercepted number.

It can be seen from the above process that the intercepted number of the extended search result corresponding to each of the extended keywords is positively correlated with the importance score, and the higher the score of the extended keyword is, the extended search result is intercepted. The more the number.

Step S104, extracting the initial search result and the target event statement in the extended search result.

The target event statement is a statement including an event keyword and a preset matching field, and the event keyword is the initial keyword or the extended keyword.

The matching field is a candidate event body. Specifically, the matching field may be a specific company, organization, institution name, or the like. Preferably, the event body database may be set in advance, and the event subjects that may be involved are saved in the event body database. For example, you can extract all bond issuance information from the bond management department's database, or obtain these bond issuance information from other third-party agencies or networks, and identify specific issuers, and save those issuers in the event body database. in.

It should be noted that the event subject database is constantly being updated. For example, when a certain issued bond has been fully executed as scheduled, the bond has no possibility of default, and if the corresponding issuer does not Other bonds that are being issued can be deleted from the event body database at this time. For example, if a newly issued bond appears and its corresponding issuer is not stored in the event body database, the issuer can Added to the event body database.

Step S105, matching the target event statement by a preset regular expression.

If the matching is successful, step S106 is performed.

Step S106, determining the matching field in the target event statement as an event body corresponding to the target event statement.

For example, you can use the following regular match expression to match:

"(.*)+keyword+(.*event_keyword)"

The event_keyword is the event keyword, and the keyword is the matching field.

If one of the target event statements is:

"Announcement of debt default of company A", if the above regular expression matches the target event statement successfully, then it can be determined that the subject of the event is "company A", that is, "company A" in which the debt default occurs.

It should be noted that the above is a positive match, that is, determining that a matching field is an event body, and optionally, a negative matching may also be performed, that is, determining that a matching field is not an event body.

For example, you can use the following regular match expression to make a negative match:

"(.*:|:)+keyword+(.*about)"

If one of the target event statements is:

"Company A: Bank B's announcement about the debt default of the company", if the above regular expression matches the target event statement successfully, then it can be determined that "Bank B" is not the event subject corresponding to the target event statement. That is, it is not "Bank B" that causes a debt default.

In the above analysis of debt default events, this method makes the extraction of debt default events more accurate and efficient, and is a kind of welfare for users, especially for investment institutions such as banks, which can achieve the effect of rapid warning.

In summary, the embodiment of the present application further introduces an extended keyword to expand on the basis of the initial keyword, and can obtain a wider search result, and realizes the event subject by using the regular expression. The automatic matching greatly improves the analysis efficiency.

Corresponding to an event information analysis method according to the above embodiment, FIG. 5 is a structural diagram of an embodiment of an event information analysis apparatus provided by an embodiment of the present application.

In this embodiment, an event information analyzing apparatus may include:

The initial search module 501 is configured to obtain an initial search result corresponding to the preset initial keyword by using a preset web search engine;

The extended keyword screening module 502 is configured to filter out an extended keyword in the initial search result, where the extended keyword is a word whose similarity with the initial keyword is greater than a preset similarity threshold;

An extended search module 503, configured to obtain, by using the network search engine, an extended search result corresponding to the extended keyword;

The target event statement extraction module 504 is configured to extract the initial search result and the target event statement in the extended search result, where the target event statement is a statement including an event keyword and a preset matching field, where the event key The word is the initial keyword or the extended keyword;

a regular matching module 505, configured to match the target event statement by using a preset regular expression;

The event body determining module 506 is configured to determine, when the matching is successful, the matching field in the target event statement as an event body corresponding to the target event statement.

Further, the event information analyzing apparatus may further include:

a hash operation module, configured to perform a hash operation on the initial search result to obtain a hash value of the initial search result;

a hash value searching module, configured to search, in a preset hash value set, a hash value of the initial search result, where the hash value set is used to record a webpage that has been stored in a preset storage medium Greek value

a search result storage module, configured to add a hash value of the initial search result to the hash value set if a hash value of the initial search result is not found in the hash value set, And storing the initial search result in the storage medium;

And a search result discarding module, configured to discard the initial search result if a hash value of the initial search result is found in the hash value set.

Further, the hash operation module may include:

a summary content obtaining unit, configured to obtain a summary content of the initial search result according to the following formula:

SubContent=Head(PageContent)∪Tail(PageContent)

Wherein, PageContent is the webpage text in the initial search result, Head (PageContent) is the first M characters of the webpage text in the initial search result, and Tail (PageContent) is the post of the webpage text in the initial search result. N characters, M and N are integers greater than 1, and SubContent is a summary content of the initial search result;

a hash value calculation unit for calculating a hash value of the initial search result according to the following formula:

Key=Hash(SubContent)=Hash[Head(PageContent)∪Tail(PageContent)]

Further, the extended keyword screening module may include:

a literal overlap calculation unit, configured to respectively calculate a literal overlap between each word in the initial search result and the initial keyword according to the following formula:

Where w is any word in the initial search result,

For the initial keyword,

For w and

The number of words that are included together,

The number of words contained in w and

The maximum number of words included,

For w with

Literal overlap between them;

Searching the overlap calculation unit for respectively calculating search overlap between each word in the initial search result and the initial keyword according to the following formula:

among them,

For the initial search results in w and

The number of pages that appear together,

The number of pages that appear in the initial search results and

The maximum number of pages that appear,

For w with

Search overlap between;

a similarity calculation unit, configured to respectively calculate a similarity between each word in the initial search result and the initial keyword according to the following formula:

Where k ₁ and k ₂ are preset weight coefficients, and k ₁ +k ₂ =1,

For w with

Similarity between

And an extended keyword determining unit, configured to determine, as the extended keyword, a word whose similarity with the initial keyword is greater than the similarity threshold.

Further, the extended search module may include:

The importance score calculation unit is configured to separately calculate the importance scores of each of the extended keywords according to the following formula:

Where ew is any of the extended keywords, freq(ew) is the frequency at which ew appears in the initial search result, and Freq(ew) is the frequency at which ew appears in the preset sample corpus, ExWord The set of the expanded keywords, max[Freq(ExWord)] is the maximum value of the frequency of occurrence of each of the extended keywords in the sample corpus, ln is a natural logarithm function, and Score(ew) is ew Importance score

The interception number calculation unit is configured to separately calculate the number of interception of the extended search result corresponding to each of the extended keywords according to the following formula:

Where ew _s is the extended keyword with the serial number s, 1≤s≤S, S is the number of the extended keywords, α is a preset proportional coefficient, and PageNum is the preset initial search result. The number, ExPageNum(ew) is the number of intercepted search results corresponding to ew;

And an extended search result obtaining unit, configured to respectively obtain extended search results corresponding to the respective extended keywords according to the intercepted number.

FIG. 6 is a schematic block diagram of an event information analysis terminal device provided by an embodiment of the present application. For convenience of description, only parts related to the embodiment of the present application are shown.

In this embodiment, the event information analysis terminal device 6 may be a computing device such as a desktop computer, a notebook, a palmtop computer, or a cloud server. The event information analysis terminal device 6 may include a processor 60, a memory 61, and computer readable instructions 62 stored in the memory 61 and operable on the processor 60, such as performing the event information analysis method described above. Computer readable instructions. When the processor 60 executes the computer readable instructions 62, the steps in the foregoing various event information analysis method embodiments are implemented, such as steps S101 to S106 shown in FIG. Alternatively, the processor 60, when executing the computer readable instructions 62, implements the functions of the various modules/units in the various apparatus embodiments described above, such as the functions of the modules 501 through 506 shown in FIG.

The functional units in the various embodiments of the present application may be stored in a computer readable storage medium if implemented in the form of a software functional unit and sold or used as a separate product. Based on such understanding, the technical solution of the present application, in essence or the contribution to the prior art, or all or part of the technical solution may be embodied in the form of a software product stored in a storage medium. A number of computer readable instructions are included to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present application.

Claims

An event information analysis method, comprising:

Acquiring initial search results corresponding to preset initial keywords by a preset web search engine;

An extended keyword is selected in the initial search result, where the extended keyword is a word whose similarity with the initial keyword is greater than a preset similarity threshold;

Acquiring an extended search result corresponding to the extended keyword by the network search engine;

Extracting the initial search result and the target event statement in the extended search result, where the target event statement is a statement including an event keyword and a preset matching field, where the event keyword is the initial keyword or Extended keyword

Matching the target event statement by a preset regular expression;

If the matching is successful, the matching field in the target event statement is determined as the event body corresponding to the target event statement.
The event information analysis method according to claim 1, further comprising: after obtaining the initial search result corresponding to the preset initial keyword by the preset network search engine, the method further comprising:

Performing a hash operation on the initial search result to obtain a hash value of the initial search result;

Searching, in a preset hash value set, a hash value of the initial search result, where the hash value set is used to record a hash value of a webpage that has been stored in a preset storage medium;

If the hash value of the initial search result is not found in the hash value set, adding a hash value of the initial search result to the hash value set, and the initial search result Stored in the storage medium;

If the hash value of the initial search result is found in the hash value set, the initial search result is discarded.
The event information analysis method according to claim 2, wherein the hashing the initial search result to obtain the hash value of the initial search result comprises:

Obtain the summary content of the initial search result according to the following formula:

SubContent=Head(PageContent)∪Tail(PageContent)

Wherein, PageContent is the webpage text in the initial search result, Head (PageContent) is the first M characters of the webpage text in the initial search result, and Tail (PageContent) is the post of the webpage text in the initial search result. N characters, M and N are integers greater than 1, and SubContent is a summary content of the initial search result;

Calculating the hash value of the initial search result according to the following formula:

Key=Hash(SubContent)=Hash[Head(PageContent)∪Tail(PageContent)]

The hash is a preset hash function, and the Key is a hash value of the initial search result.
The event information analysis method according to claim 1, wherein the filtering out the extended keywords in the initial search result comprises:

Calculating the literal overlap between each word in the initial search result and the initial keyword according to the following formula:

Where w is any word in the initial search result,
For the initial keyword,
For w and
The number of words that are included together,
The number of words contained in w and
The maximum number of words included,
For w with
Literal overlap between them;

Calculating the search overlap between each word in the initial search result and the initial keyword according to the following formula:

among them,
For the initial search results in w and
The number of pages that appear together,
The number of pages that appear in the initial search results and
The maximum number of pages that appear,
For w with
Search overlap between;

Calculating the similarity between each word in the initial search result and the initial keyword according to the following formula:

Where k 1 and k 2 are preset weight coefficients, and k 1 +k 2 =1,
For w with
Similarity between

A word having a degree of similarity to the initial keyword greater than the similarity threshold is determined as the extended keyword.
The event information analysis method according to any one of claims 1 to 4, wherein the obtaining, by the network search engine, the extended search result corresponding to the extended keyword comprises:

Calculate the importance scores of each of the extended keywords according to the following formula:

Where ew is any of the extended keywords, freq(ew) is the frequency at which ew appears in the initial search result, and Freq(ew) is the frequency at which ew appears in the preset sample corpus, ExWord The set of the expanded keywords, max[Freq(ExWord)] is the maximum value of the frequency of occurrence of each of the extended keywords in the sample corpus, ln is a natural logarithm function, and Score(ew) is ew Importance score

Calculating the number of interception of the extended search result corresponding to each of the extended keywords according to the following formula:

Where ew s is the extended keyword with the serial number s, 1≤s≤S, S is the number of the extended keywords, α is a preset proportional coefficient, and PageNum is the preset initial search result. The number, ExPageNum(ew) is the number of intercepted search results corresponding to ew;

The extended search results corresponding to the respective extended keywords are respectively acquired according to the intercepted number.
A computer readable storage medium storing computer readable instructions, wherein the computer readable instructions, when executed by a processor, implement the following steps:

Acquiring initial search results corresponding to preset initial keywords by a preset web search engine;

An extended keyword is selected in the initial search result, where the extended keyword is a word whose similarity with the initial keyword is greater than a preset similarity threshold;

Acquiring an extended search result corresponding to the extended keyword by the network search engine;

Extracting the initial search result and the target event statement in the extended search result, where the target event statement is a statement including an event keyword and a preset matching field, where the event keyword is the initial keyword or Extended keyword

Matching the target event statement by a preset regular expression;

If the matching is successful, the matching field in the target event statement is determined as the event body corresponding to the target event statement.
The computer readable storage medium according to claim 6, wherein after the initial search result corresponding to the preset initial keyword is obtained by the preset web search engine, the method further includes:

Performing a hash operation on the initial search result to obtain a hash value of the initial search result;

Searching, in a preset hash value set, a hash value of the initial search result, where the hash value set is used to record a hash value of a webpage that has been stored in a preset storage medium;

If the hash value of the initial search result is not found in the hash value set, adding a hash value of the initial search result to the hash value set, and the initial search result Stored in the storage medium;

If the hash value of the initial search result is found in the hash value set, the initial search result is discarded.
The computer readable storage medium according to claim 7, wherein the hashing the initial search result to obtain the hash value of the initial search result comprises:

Obtain the summary content of the initial search result according to the following formula:

SubContent=Head(PageContent)∪Tail(PageContent)

Wherein, PageContent is the webpage text in the initial search result, Head (PageContent) is the first M characters of the webpage text in the initial search result, and Tail (PageContent) is the post of the webpage text in the initial search result. N characters, M and N are integers greater than 1, and SubContent is a summary content of the initial search result;

Calculating the hash value of the initial search result according to the following formula:

Key=Hash(SubContent)=Hash[Head(PageContent)∪Tail(PageContent)]

The hash is a preset hash function, and the Key is a hash value of the initial search result.
The computer readable storage medium according to claim 6, wherein the filtering out the extended keywords in the initial search result comprises:

Calculating the literal overlap between each word in the initial search result and the initial keyword according to the following formula:

Where w is any word in the initial search result,
For the initial keyword,
For w and
The number of words that are included together,
The number of words contained in w and
The maximum number of words included,
For w with
Literal overlap between them;

Calculating the search overlap between each word in the initial search result and the initial keyword according to the following formula:

among them,
For the initial search results in w and
The number of pages that appear together,
The number of pages that appear in the initial search results and
The maximum number of pages that appear,
For w with
Search overlap between;

Calculating the similarity between each word in the initial search result and the initial keyword according to the following formula:

Where k 1 and k 2 are preset weight coefficients, and k 1 +k 2 =1,
For w with
Similarity between

A word having a degree of similarity to the initial keyword greater than the similarity threshold is determined as the extended keyword.
The computer readable storage medium according to any one of claims 6 to 9, wherein the obtaining, by the network search engine, the extended search result corresponding to the extended keyword comprises:

Calculate the importance scores of each of the extended keywords according to the following formula:

Where ew is any of the extended keywords, freq(ew) is the frequency at which ew appears in the initial search result, and Freq(ew) is the frequency at which ew appears in the preset sample corpus, ExWord The set of the expanded keywords, max[Freq(ExWord)] is the maximum value of the frequency of occurrence of each of the extended keywords in the sample corpus, ln is a natural logarithm function, and Score(ew) is ew Importance score

Calculating the number of interception of the extended search result corresponding to each of the extended keywords according to the following formula:

Where ew s is the extended keyword with the serial number s, 1≤s≤S, S is the number of the extended keywords, α is a preset proportional coefficient, and PageNum is the preset initial search result. The number, ExPageNum(ew) is the number of intercepted search results corresponding to ew;

The extended search results corresponding to the respective extended keywords are respectively acquired according to the intercepted number.
An event information analysis terminal device includes a memory, a processor, and computer readable instructions stored in the memory and executable on the processor, wherein the processor executes the computer readable instructions The following steps are implemented:

Acquiring initial search results corresponding to preset initial keywords by a preset web search engine;

An extended keyword is selected in the initial search result, where the extended keyword is a word whose similarity with the initial keyword is greater than a preset similarity threshold;

Acquiring an extended search result corresponding to the extended keyword by the network search engine;

Extracting the initial search result and the target event statement in the extended search result, where the target event statement is a statement including an event keyword and a preset matching field, where the event keyword is the initial keyword or Extended keyword

Matching the target event statement by a preset regular expression;

If the matching is successful, the matching field in the target event statement is determined as the event body corresponding to the target event statement.
The event information analysis terminal device according to claim 11, wherein after the initial search result corresponding to the preset initial keyword is obtained by the preset network search engine, the method further includes:

Performing a hash operation on the initial search result to obtain a hash value of the initial search result;

Searching, in a preset hash value set, a hash value of the initial search result, where the hash value set is used to record a hash value of a webpage that has been stored in a preset storage medium;

If the hash value of the initial search result is not found in the hash value set, adding a hash value of the initial search result to the hash value set, and the initial search result Stored in the storage medium;

If the hash value of the initial search result is found in the hash value set, the initial search result is discarded.
The event information analysis terminal device according to claim 12, wherein the hashing the initial search result to obtain the hash value of the initial search result comprises:

Obtain the summary content of the initial search result according to the following formula:

SubContent=Head(PageContent)∪Tail(PageContent)

Wherein, PageContent is the webpage text in the initial search result, Head (PageContent) is the first M characters of the webpage text in the initial search result, and Tail (PageContent) is the post of the webpage text in the initial search result. N characters, M and N are integers greater than 1, and SubContent is a summary content of the initial search result;

Calculating the hash value of the initial search result according to the following formula:

Key=Hash(SubContent)=Hash[Head(PageContent)∪Tail(PageContent)]

The hash is a preset hash function, and the Key is a hash value of the initial search result.
The event information analysis terminal device according to claim 11, wherein the filtering out the extended keywords in the initial search result comprises:

Calculating the literal overlap between each word in the initial search result and the initial keyword according to the following formula:

Where w is any word in the initial search result,
For the initial keyword,
For w and
The number of words that are included together,
The number of words contained in w and
The maximum number of words included,
For w with
Literal overlap between them;

Calculating the search overlap between each word in the initial search result and the initial keyword according to the following formula:

among them,
For the initial search results in w and
The number of pages that appear together,
The number of pages that appear in the initial search results and
The maximum number of pages that appear,
For w with
Search overlap between;

Calculating the similarity between each word in the initial search result and the initial keyword according to the following formula:

Where k 1 and k 2 are preset weight coefficients, and k 1 +k 2 =1,
For w with
Similarity between

A word having a degree of similarity to the initial keyword greater than the similarity threshold is determined as the extended keyword.
The event information analysis terminal device according to any one of claims 11 to 14, wherein the obtaining, by the network search engine, the extended search result corresponding to the extended keyword comprises:

Calculate the importance scores of each of the extended keywords according to the following formula:

Where ew is any of the extended keywords, freq(ew) is the frequency at which ew appears in the initial search result, and Freq(ew) is the frequency at which ew appears in the preset sample corpus, ExWord The set of the expanded keywords, max[Freq(ExWord)] is the maximum value of the frequency of occurrence of each of the extended keywords in the sample corpus, ln is a natural logarithm function, and Score(ew) is ew Importance score

Calculating the number of interception of the extended search result corresponding to each of the extended keywords according to the following formula:

Where ew s is the extended keyword with the serial number s, 1≤s≤S, S is the number of the extended keywords, α is a preset proportional coefficient, and PageNum is the preset initial search result. The number, ExPageNum(ew) is the number of intercepted search results corresponding to ew;

The extended search results corresponding to the respective extended keywords are respectively acquired according to the intercepted number.
An event information analysis device, comprising:

An initial search module, configured to obtain an initial search result corresponding to the preset initial keyword by using a preset web search engine;

An extended keyword screening module, configured to filter out an extended keyword in the initial search result, where the extended keyword is a word whose similarity with the initial keyword is greater than a preset similarity threshold;

An extended search module, configured to obtain, by using the network search engine, an extended search result corresponding to the extended keyword;

a target event statement extracting module, configured to extract the initial search result and a target event statement in the extended search result, where the target event statement is a statement including an event keyword and a preset matching field, the event keyword Is the initial keyword or the extended keyword;

a regular matching module, configured to match the target event statement by a preset regular expression;

The event body determining module is configured to determine, when the matching is successful, the matching field in the target event statement as an event body corresponding to the target event statement.
The event information analysis device according to claim 16, further comprising:

a hash operation module, configured to perform a hash operation on the initial search result to obtain a hash value of the initial search result;

a hash value searching module, configured to search, in a preset hash value set, a hash value of the initial search result, where the hash value set is used to record a webpage that has been stored in a preset storage medium Greek value

a search result storage module, configured to add a hash value of the initial search result to the hash value set if a hash value of the initial search result is not found in the hash value set, And storing the initial search result in the storage medium;

And a search result discarding module, configured to discard the initial search result if a hash value of the initial search result is found in the hash value set.
The event information analysis apparatus according to claim 17, wherein the hash operation module comprises:

a summary content obtaining unit, configured to obtain a summary content of the initial search result according to the following formula:

SubContent=Head(PageContent)∪Tail(PageContent)

Wherein, PageContent is the webpage text in the initial search result, Head (PageContent) is the first M characters of the webpage text in the initial search result, and Tail (PageContent) is the post of the webpage text in the initial search result. N characters, M and N are integers greater than 1, and SubContent is a summary content of the initial search result;

a hash value calculation unit for calculating a hash value of the initial search result according to the following formula:

Key=Hash(SubContent)=Hash[Head(PageContent)∪Tail(PageContent)]

The hash is a preset hash function, and the Key is a hash value of the initial search result.
The event information analysis apparatus according to claim 16, wherein the extended keyword screening module comprises:

a literal overlap calculation unit, configured to respectively calculate a literal overlap between each word in the initial search result and the initial keyword according to the following formula:

Where w is any word in the initial search result,
For the initial keyword,
For w and
The number of words that are included together,
The number of words contained in w and
The maximum number of words included,
For w with
Literal overlap between them;

Searching the overlap calculation unit for respectively calculating search overlap between each word in the initial search result and the initial keyword according to the following formula:

among them,
For the initial search results in w and
The number of pages that appear together,
The number of pages that appear in the initial search results and
The maximum number of pages that appear,
For w with
Search overlap between;

a similarity calculation unit, configured to respectively calculate a similarity between each word in the initial search result and the initial keyword according to the following formula:

Where k 1 and k 2 are preset weight coefficients, and k 1 +k 2 =1,
For w with
Similarity between

And an extended keyword determining unit, configured to determine, as the extended keyword, a word whose similarity with the initial keyword is greater than the similarity threshold.
The event information analysis apparatus according to any one of claims 16 to 19, wherein the extended search module may include:

The importance score calculation unit is configured to separately calculate the importance scores of each of the extended keywords according to the following formula:

Where ew is any of the extended keywords, freq(ew) is the frequency at which ew appears in the initial search result, and Freq(ew) is the frequency at which ew appears in the preset sample corpus, ExWord The set of the expanded keywords, max[Freq(ExWord)] is the maximum value of the frequency of occurrence of each of the extended keywords in the sample corpus, ln is a natural logarithm function, and Score(ew) is ew Importance score

The interception number calculation unit is configured to separately calculate the number of interception of the extended search result corresponding to each of the extended keywords according to the following formula:

Where ew s is the extended keyword with the serial number s, 1≤s≤S, S is the number of the extended keywords, α is a preset proportional coefficient, and PageNum is the preset initial search result. The number, ExPageNum(ew) is the number of intercepted search results corresponding to ew;

And an extended search result obtaining unit, configured to respectively obtain extended search results corresponding to the respective extended keywords according to the intercepted number.