CN109740075A

CN109740075A - Event relatedness computation method, apparatus, equipment and storage medium

Info

Publication number: CN109740075A
Application number: CN201811528235.2A
Authority: CN
Inventors: 周辉; 陈文浩; 陈玉光; 郑宇宏; 陈伟娜; 韩翠云; 潘禄
Original assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Current assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Priority date: 2018-12-13
Filing date: 2018-12-13
Publication date: 2019-05-10
Anticipated expiration: 2038-12-13
Also published as: CN109740075B

Abstract

The embodiment of the present invention provides a kind of event relatedness computation method, apparatus, equipment and storage medium.The embodiment of the present invention is by obtaining User action log, according to the corresponding multiple links of first event in event base, the corresponding multiple links of second event and the User action log, calculate the degree of correlation of the first event He the second event, or, according to the corresponding multiple search keys of first event, the corresponding multiple search keys of second event and the User action log, calculate the degree of correlation of the first event He the second event, pass through the search of different user and click the degree of correlation between behavior calculating event, improves the computational accuracy of the event degree of correlation.

Description

Event relatedness computation method, apparatus, equipment and storage medium

Technical field

The present embodiments relate to field of computer technology more particularly to a kind of event relatedness computation method, apparatus, set Standby and storage medium.

Background technique

Event is the objective fact to be interacted by particular person, object, thing in specific time, locality, the generation of event Have the characteristics that objectivity, authenticity.The event degree of correlation refers to the strength of correlation between event and event.

In the prior art, the calculation method of the event degree of correlation includes: that the side of the event degree of correlation is calculated by text relevant Method and the method that the event degree of correlation is calculated by event determinant attribute.But pass through the event degree of correlation of prior art calculating Precision is lower.

Summary of the invention

The embodiment of the present invention provides a kind of event relatedness computation method, apparatus, equipment and storage medium, to improve event The computational accuracy of the degree of correlation.

In a first aspect, the embodiment of the present invention provides a kind of event relatedness computation method, comprising:

User action log is obtained, the User action log includes multiple record information, in the multiple record information The corresponding user of each record information a search behavior, the record information include at least one search key with At least one link that the user clicks；

According to the corresponding multiple links of first event in event base and the User action log, the first event is determined Corresponding multiple search keys, the first event correspond to multiple first information, the corresponding multiple links of the first event It is corresponded with the multiple first information, each search key in the corresponding multiple search keys of the first event For searching for and clicking the first information of at least one of the multiple first information；

According to the corresponding multiple links of second event in the event base and the User action log, described second is determined The corresponding multiple search keys of event, the second event correspond to multiple second information, and the second event is corresponding multiple Link and the multiple second information correspond, and each search in the corresponding multiple search keys of the second event is closed Keyword is for searching for and clicking the second information of at least one of the multiple second information；

According to the corresponding multiple links of the first event, the corresponding multiple links of the second event and user's row For log, the degree of correlation of the first event and the second event is calculated；Alternatively, corresponding multiple according to the first event The corresponding multiple search keys of search key, the second event and the User action log calculate first thing The degree of correlation of part and the second event.

Second aspect, the embodiment of the present invention provide a kind of event relatedness computation device, comprising:

Module is obtained, for obtaining User action log, the User action log includes multiple record information, described more The search behavior of the corresponding user of each record information in a record information, the record information includes at least one At least one link that search key and the user click；

First determining module, for according to the corresponding multiple links of first event in event base and the user behavior day Will, determines the corresponding multiple search keys of the first event, and the first event corresponds to multiple first information, described first The corresponding multiple links of event and the multiple first information correspond, the corresponding multiple search keys of the first event In each search key for searching for and click the first information of at least one of the multiple first information；

Second determining module, for according to the corresponding multiple links of second event in the event base and the user behavior Log, determines the corresponding multiple search keys of the second event, and the second event corresponds to multiple second information, described The corresponding multiple links of two events and the multiple second information correspond, the corresponding multiple search keys of the second event Each search key in word is for searching for and clicking the second information of at least one of the multiple second information；

First computing module, for corresponding more according to the corresponding multiple links of the first event, the second event A link and the User action log, calculate the degree of correlation of the first event and the second event；Alternatively,

Second computing module, for according to the corresponding multiple search keys of the first event, the second event pair The multiple search keys and the User action log answered, calculate the degree of correlation of the first event and the second event.

The third aspect, the embodiment of the present invention provide a kind of equipment, comprising:

Memory；

Processor；And

Computer program；

Wherein, the computer program stores in the memory, and is configured as being executed by the processor with reality Now method as described in relation to the first aspect.

Fourth aspect, the embodiment of the present invention provide a kind of computer readable storage medium, are stored thereon with computer program, The computer program is executed by processor to realize method described in first aspect.

Event relatedness computation method, apparatus, equipment and storage medium provided in an embodiment of the present invention, by obtaining user User behaviors log, according to the corresponding multiple links of first event in event base, the corresponding multiple links of second event and user's row For log, the degree of correlation of the first event He the second event is calculated, alternatively, crucial according to the corresponding multiple search of first event The corresponding multiple search keys of word, second event and the User action log calculate the first event and the second event The degree of correlation passes through the search of different user and clicks the degree of correlation between behavior calculating event, improves the meter of the event degree of correlation Calculate precision.

Detailed description of the invention

Fig. 1 is event relatedness computation method flow diagram provided in an embodiment of the present invention；

Fig. 2 be another embodiment of the present invention provides event schematic diagram；

Fig. 3 be another embodiment of the present invention provides event relatedness computation method flow diagram；

Fig. 4 be another embodiment of the present invention provides event relatedness computation method flow diagram；

Fig. 5 is the structural schematic diagram of event relatedness computation device provided in an embodiment of the present invention；

Fig. 6 is the structural schematic diagram of event relatedness computation device provided in an embodiment of the present invention；

Fig. 7 is the structural schematic diagram of equipment provided in an embodiment of the present invention.

Through the above attached drawings, it has been shown that the specific embodiment of the disclosure will be hereinafter described in more detail.These attached drawings It is not intended to limit the scope of this disclosure concept by any means with verbal description, but is by referring to specific embodiments Those skilled in the art illustrate the concept of the disclosure.

Specific embodiment

Example embodiments are described in detail here, and the example is illustrated in the accompanying drawings.Following description is related to When attached drawing, unless otherwise indicated, the same numbers in different drawings indicate the same or similar elements.Following exemplary embodiment Described in embodiment do not represent all implementations consistent with this disclosure.On the contrary, they be only with it is such as appended The example of the consistent device and method of some aspects be described in detail in claims, the disclosure.

In embodiments of the present invention, event (event) is by particular person, object, thing in specific time, locality phase interaction Objective fact, the generation of event have the characteristics that objectivity, authenticity.The degree of correlation of event refers to deposits between two events In certain incidence relation, and the intensity of correlation can be measured by a numerical value.The embodiment of the present invention is according to user's row For the degree of correlation between calculating event, which includes user's search behavior and click behavior.Search behavior specifically can be with It is that user inputs search key in a search engine, search engine searches out and the search key according to the search key Relevant multiple search results.Click behavior specifically can be in multiple search results that user searches out the search engine At least one search result is clicked.

How to be solved with technical solution of the specifically embodiment to technical solution of the present invention and the application below above-mentioned Technical problem is described in detail.These specific embodiments can be combined with each other below, for the same or similar concept Or process may repeat no more in certain embodiments.Below in conjunction with attached drawing, the embodiment of the present invention is described.

Fig. 1 is event relatedness computation method flow diagram provided in an embodiment of the present invention.The embodiment of the present invention is for existing The technical problem as above of technology provides event relatedness computation method, and specific step is as follows for this method:

Step 101 obtains User action log, and the User action log includes multiple record information, the multiple note The search behavior of the corresponding user of each record information in information is recorded, the record information includes at least one search At least one link that keyword and the user click.

In the present embodiment, user inputs search key in a search engine, and search engine is according to the search key Multiple search results relevant to the search key are searched out, in multiple search results that user searches out the search engine At least one search result clicked.The search key and user that search engine is inputted according to user tie search The click behavior of fruit generates User action log, and therefore, User action log can searching from the magnanimity of search engine record The link for the search result that rope keyword and user click.In the present embodiment, User action log may include multiple record letters Breath, the search behavior of multiple corresponding user of each record information recorded in information, that is to say, that according to one The search behavior of user generates a record information, which includes at least one search key and the user At least one link clicked.In the present embodiment, search key is denoted as query, link is uniform resource locator (Uniform Resource Locator, URL).

For example, the t1 moment, user 1 has input query1 in a search engine, and search engine searches multiple search results, User clicks some search result in multiple search result, and the corresponding URL of the search result that user clicks is denoted as URL1, then user 1 inputs query1 and clicks the search behavior that URL1 can be denoted as user 1, the corresponding note of the search behavior Recording information includes query1 and URL1, which is denoted as (query1, URL1).

In other embodiments, if user input in a search engine it is defeated in the preset time after a query Another query is entered, then the user has been inputted into two query and corresponding click behavior is denoted as once searching for the user Suo Hangwei.For example, the t1 moment, user 1 has input query1 in a search engine, and search engine searches multiple search results, The URL for the search result that user clicks is URL1.In preset time after instant tl in such as one minute, which exists Query2 is had input in search engine, search engine searches again for multiple search results, the URL for the search result that user clicks For URL2, then user 1 is inputted into query1 and click URL1, input query2 and click the primary search that URL2 is denoted as user 1 Behavior, the corresponding record information of the search behavior are denoted as (query1, URL1, query2, URL2).

If the t1 moment, user 1 has input query1 in a search engine, and search engine searches multiple search results, uses Any one search result is not clicked in family.In preset time after instant tl in such as one minute, which is being searched for Query2 is had input in engine, search engine searches again for multiple search results, and the URL for the search result that user clicks is User 1 is then inputted query1, inputs query2 and clicked the search behavior that URL2 is denoted as user 1, the search row by URL2 (query1, query2, URL2) is denoted as corresponding record information.

In other embodiments, user is in the preset time in such as one minute, if repeatedly inputting different query, Multiple query that the user inputs and corresponding click behavior are then denoted as to the search behavior of the user.In addition, due to User action log includes multiple record information, and each record information can also correspond to an identification information, for example, by each note Record information is denoted as a Session, and the identification information for recording information is denoted as session_id, then Session is represented by (session_id, query1, URL1, query2, URL2 ... ...).

Optionally, using day as chronomere's counting user user behaviors log, then may include in the User action log it is multiple not The corresponding record information with the search behavior at least once of user each in user.

Further, according to each record information in the User action log, it may be determined that go out searching in each record information Rope keyword to link pair, the search key to include two different search keys, the link to include two not Same link.Further, the search key is counted to the number occurred in the User action log, i.e. the search key Pair co-occurrence number.The link is counted to the number occurred in the User action log, i.e. the co-occurrence number of the link pair.

For example, the Session of Session and user 2 in the User action log including user 1, user's 1 Session is expressed as (session_1, query1, URL1, query2, URL2), and the Session of user 2 is expressed as (session_2, query1, URL2, query2, URL3).Search key in the Session of user 1 to for (query1, Query2), link is to for (URL1, URL2).Search key in the Session of user 2 to for (query1, query2), Link is to for (URL2, URL3).Search key is 2 to the co-occurrence number of (query1, query2), link to (URL1, URL2 co-occurrence number) is 1, and linking to the co-occurrence number of (URL2, URL3) is 1.

For another example the Session of Session and user 2 in the User action log including user 1, user's 1 Session is expressed as (session_1, query1, URL1, query2, URL2), and the Session of user 2 is expressed as (session_2, query1, URL2, query2, URL3, query3, URL1).Search key in the Session of user 1 To being (query1, query2), link to for (URL1, URL2).Search key in the Session of user 2 to for (query1, query2), (query1, query3), (query2, query3), link in the Session of user 2 to for (URL1, URL2), (URL1, URL3), (URL2, URL3).Search key is 2 to the co-occurrence number of (query1, query2), Linking to the co-occurrence number of (URL1, URL2) is 2, and search key is 1 to the co-occurrence number of (query1, query3), search Keyword is 1 to the co-occurrence number of (query2, query3), and linking to the co-occurrence number of (URL1, URL3) is 1, link pair The co-occurrence number of (URL2, URL3) is 1.Search different in User action log in statistics available whole day out according to the method Co-occurrence number of the keyword to co-occurrence number and each link pair with different links pair and each search key pair. Further, by each search key to, each link to the co-occurrence number of, each search key pair, each link pair Co-occurrence number and date storage are in the database.

Step 102, according to first event in event base it is corresponding it is multiple link and the User action log, determine described in The corresponding multiple search keys of first event, the first event correspond to multiple first information, and the first event is corresponding Multiple links and the multiple first information correspond, and each of corresponding multiple search keys of the first event are searched Rope keyword is for searching for and clicking the first information of at least one of the multiple first information.

In the present embodiment, a large amount of information processes by polymerization, attribute extraction and optimization and forms event base, in event base Including multiple events, an event can be described with information cluster, information cluster, that is, multiple information, that is to say, that an event is available Multiple information describe, which specifically can be Domestic News.For example, calculating the correlation of any two event in event base Degree, is denoted as first event for one of event, another event is denoted as second event, for describing the information of first event It is denoted as the first information, the information for describing second event is denoted as the second information, that is to say, that first event can use multiple first Information describes, and second event can describe with multiple second information.

For example, first event can be described with information 1, information 2, information 3, the URL of information 1 is denoted as URL1, information 2 URL is denoted as URL2, and the URL of information 3 is denoted as URL3, then the corresponding multiple links of first event as URL1, URL2 and URL3, can Choosing, URL1, URL2 and URL3 are constituted into the first lists of links, which is expressed as { URL1, URL2, URL3 }, First lists of links is the corresponding lists of links of first event.Further, according to the corresponding lists of links of first event and User action log, determines the corresponding multiple search keys of the first event, the corresponding multiple search of the first event Each search key in keyword is for searching for and clicking the first information of at least one of the multiple first information.

Optionally, described according to the corresponding multiple links of first event in event base and the User action log, it determines The corresponding multiple search keys of the first event, comprising: corresponding multiple according to first event described in the event base Each link in link obtains from the User action log and links at least one corresponding search key with described； By at least one corresponding search key of each of multiple links link, it is determined as the first event Corresponding multiple search keys.

For example, including Session, the Session of user 2 and the Session of user 3 of user 1 in User action log. Wherein, the Session of user 1 is expressed as (session_1, query1, URL1, query2, URL2), the Session table of user 2 It is shown as (session_2, query1, URL2, query2, URL3, query3, URL1), the Session of user 3 is expressed as (session_3, query2, URL2, query3, URL3, query4, URL4).Each search is crucial in the User action log Word to shown in the co-occurrence number of each search key pair table 1 specific as follows, in the User action log it is each link to Shown in the co-occurrence number table 2 specific as follows of each link pair:

Table 1

Search key pair	Co-occurrence number
		(query1, query2)	2
(query1, query3)	1
		(query2, query3)	2
(query2, query4)	1
		(query3, query4)	1

Table 2

Link pair	Co-occurrence number
		(URL1, URL2)	2
(URL1, URL3)	1
		(URL2, URL3)	2
(URL2, URL4)	1
		(URL3, URL4)	1

According to each link that the corresponding multiple links of first event are in URL1, URL2 and URL3, from user's row At least one corresponding search key is linked with this to obtain in log.For example, in the User action log, with URL1 pairs At least one search key answered is query1 and query3, at least one search key corresponding with URL2 is query1 And query2, at least one search key corresponding with URL3 are query2 and query3.It will be in URL1, URL2 and URL3 It is each to link at least one corresponding search key, it is determined as the corresponding multiple search keys of the first event, i.e., The corresponding multiple search keys of the first event are and the corresponding query1 and query3 of URL1, query1 corresponding with URL2 With the set of query2, query2 and query3 corresponding with URL3.By the corresponding multiple search key notes of the first event For the first search key list, which is denoted as { query1, query2, query3 }.

Step 103, according to second event in the event base it is corresponding it is multiple link and the User action log, determine The corresponding multiple search keys of the second event, the second event correspond to multiple second information, the second event pair The multiple links answered and the multiple second information correspond, every in the corresponding multiple search keys of the second event A search key is for searching for and clicking the second information of at least one of the multiple second information.

For example, the second event in event base can be described with information 3 and information 4, the URL of information 3 is denoted as URL3, information 4 URL is denoted as URL4, then the corresponding multiple links of second event are URL3 and URL4, and optionally, URL3 and URL4 is constituted Second lists of links, second lists of links are expressed as { URL3, URL4 }, which is that second event is corresponding Lists of links.Further, according to the corresponding lists of links of second event and User action log, determine that the second event is corresponding Multiple search keys, each search key in the corresponding multiple search keys of the second event is for searching for simultaneously Click the second information of at least one of the multiple second information.

Optionally, it is described according to second event in the event base it is corresponding it is multiple link and the User action log, Determine the corresponding multiple search keys of the second event, comprising: corresponding according to second event described in the event base Each link in multiple links obtains from the User action log and links at least one corresponding search key with described Word；By at least one corresponding search key of each of multiple links link, it is determined as described second The corresponding multiple search keys of event.

For example, being each link in URL3 and URL4 according to the corresponding multiple links of second event, from the user behavior It is obtained in log and links at least one corresponding search key with this.For example, in the User action log, it is corresponding with URL3 At least one search key be query2 and query3, at least one search key corresponding with URL4 be query4. By at least one corresponding search key of each link in URL3 and URL4, it is corresponding more to be determined as the second event The corresponding multiple search keys of a search key, the i.e. second event be query2 and query3 corresponding with URL3, with The set of the corresponding query4 of URL4.The corresponding multiple search keys of the second event are denoted as the second search key column Table, the second search key list are denoted as { query2, query3, query4 }.

Step 104, according to the corresponding multiple links of the first event, the corresponding multiple links of the second event and institute User action log is stated, the degree of correlation of the first event and the second event is calculated.

For example, according to corresponding first lists of links { URL1, URL2, URL3 } of first event, second event corresponding Two lists of links { URL3, URL4 } and the User action log, calculate the degree of correlation of the first event He the second event.Specifically , each link in traversal the first lists of links { URL1, URL2, URL3 } links the link currently traversed with second Each link in list { URL3, URL4 } respectively constitutes a link pair, to obtain following multiple links pair: (URL1, URL3), (URL1, URL4), (URL2, URL3), (URL2, URL4), (URL3, URL3), (URL3, URL4).Optionally, each Link to include two different links, therefore remove (URL3, URL3).Further statistics (URL1, URL3), (URL1, URL4), (URL2, URL3), (URL2, URL4), (URL3, URL4) the co-occurrence number in the User action log respectively, tool Body, can the statistical result according to shown in table 2 as above, inquiry (URL1, URL3), (URL1, URL4), (URL2, URL3), (URL2, URL4), (URL3, URL4) the co-occurrence number in the User action log respectively.Further, according to (URL1, URL3), (URL1, URL4), (URL2, URL3), (URL2, URL4), (URL3, URL4) are respectively in the User action log Co-occurrence number calculates the degree of correlation of the first event He the second event.

Step 105 multiple is searched according to the corresponding multiple search keys of the first event, the second event are corresponding Rope keyword and the User action log, calculate the degree of correlation of the first event and the second event.

For example, according to the corresponding first search key list { query1, query2, query3 } of first event, second The corresponding second search key list { query2, query3, query4 } of event and the User action log, calculate this first The degree of correlation of event and the second event.Specifically, the first search key list { query1, query2, query3 } of traversal In each search key, by the search key currently traversed and the second search key list query2, Query3, query4 } in each search key respectively constitute a search key pair, optionally, each search is crucial Word is to including two different search keys, to obtain following multiple search keys pair: (query1, query2), (query1, query3), (query1, query4), (query2, query3), (query2, query4), (query3, query4).Further statistics (query1, query2), (query1, query3), (query1, query4), (query2, Query3), (query2, query4), (query3, query4) the co-occurrence number in the User action log respectively, specifically , can according to statistical result as listed in Table 1, inquire (query1, query2), (query1, query3), (query1, Query4), (query2, query3), (query2, query4), (query3, query4) are respectively in the User action log Co-occurrence number.Further, according to (query1, query2), (query1, query3), (query1, query4), The co-occurrence in the User action log is secondary respectively by (query2, query3), (query2, query4), (query3, query4) Number, calculates the degree of correlation of the first event He the second event.

In the present embodiment, calculate the degree of correlation of the first event and the second event method can selecting step 104 or Step 105, that is to say, that in the present embodiment, can be corresponding with second event more according to the corresponding multiple links of first event A chain fetches the degree of correlation for calculating the first event and the second event, or crucial according to the corresponding multiple search of first event Word and the corresponding multiple search keys of second event calculate the degree of correlation of the first event He the second event.

As illustrated in fig. 2, it is assumed that there is 4 events in event base, i.e. event 1, event 2, event 3, event 4 are only shown herein Meaning property explanation, does not limit the number of event in the event base.The method according to the present embodiment can determine that in 4 events The degree of correlation between any two event.When specifying an object event such as event 1, it may be determined that go out its in the event base His event degree of correlation with the event 1 respectively, alternatively, can determine to be greater than threshold with the degree of correlation of the event 1 in the event base The list of thing of value.

The embodiment of the present invention by obtain User action log, according to first event in event base it is corresponding it is multiple link, The corresponding multiple links of second event and the User action log, calculate the degree of correlation of the first event He the second event, or Person, according to the corresponding multiple search keys of first event, the corresponding multiple search keys of second event and the user behavior Log calculates the degree of correlation of the first event He the second event, passes through the search of different user and clicks behavior and calculates event Between the degree of correlation, improve the computational accuracy of the event degree of correlation.

Fig. 3 be another embodiment of the present invention provides event relatedness computation method flow diagram.In the base of above-described embodiment It is described according to the corresponding multiple links of the first event, the corresponding multiple links of the second event and the user on plinth User behaviors log calculates the degree of correlation of the first event and the second event, specifically comprises the following steps:

Each link in step 301, the corresponding multiple links of the traversal first event, described in currently traversing The corresponding link of first event it is corresponding with the second event it is multiple link in each link respectively constitute one and link pair.

For example, the corresponding multiple links of first event constitute the first lists of links, the first lists of links be expressed as URL1, URL2, URL3 }, the corresponding multiple links of second event constitute the second lists of links, the second lists of links be expressed as URL3, URL4}.Each link in the first lists of links { URL1, URL2, URL3 } is traversed, by the link currently traversed and the second chain It connects each link in list { URL3, URL4 } and respectively constitutes a link pair, to obtain following multiple links pair: (URL1, URL3), (URL1, URL4), (URL2, URL3), (URL2, URL4), (URL3, URL3), (URL3, URL4).

Step 302, according to each link to occurring in the different record information in the User action log Number calculates the degree of correlation of the first event and the second event.

Further statistics (URL1, URL3), (URL1, URL4), (URL2, URL3), (URL2, URL4), (URL3, URL4) Co-occurrence number in the User action log respectively, specifically, can the statistical result according to shown in table 2 as above, inquiry (URL1, URL3), (URL1, URL4), (URL2, URL3), (URL2, URL4), (URL3, URL4) are respectively in the user behavior day Co-occurrence number in will.Further, according to (URL1, URL3), (URL1, URL4), (URL2, URL3), (URL2, URL4), (URL3, URL4) co-occurrence number in the User action log respectively, it is related to the second event to calculate the first event Degree.

Optionally, the link is described to the number occurred in the different record information in the User action log It include the number of the record information of the link pair in User action log.As shown in table 2, being total to (URL1, URL2) is linked Occurrence number is 2.For example, including the Session of user 1, the Session of user 2 and user 3 in User action log Session.Wherein, the Session of user 1 is expressed as (session_1, query1, URL1, query2, URL2), user's 2 Session is expressed as (session_2, query1, URL2, query2, URL3, query3, URL1), the Session table of user 3 It is shown as (session_3, query2, URL2, query3, URL3, query4, URL4).As it can be seen that including in the Session of user 1 (URL1, URL2) includes (URL1, URL2) in the Session of user 2, do not include in the Session of user 3 (URL1, It URL2), i.e., include (URL1, URL2) there are two Session in the User action log.So link is to (URL1, URL2) The number of Session in co-occurrence number namely User action log including (URL1, URL2).The co-occurrence of other links pair Number similarly in this, no longer repeats one by one.

Optionally, described to be occurred according to each link to different record in information in the User action log Number, calculate the degree of correlation of the first event and the second event, including following several feasible implementations:

One kind is feasible to be achieved in that: by each link to the different record letters in the User action log The number occurred in breath carries out addition calculating, obtains the degree of correlation of the first event and the second event.

For example, by (URL1, URL3), (URL1, URL4), (URL2, URL3), (URL2, URL4), (URL3, URL4) point Co-occurrence number not in the User action log carries out addition calculating, and it is related to the second event to obtain the first event Degree.

Another kind is feasible to be achieved in that: being remembered according to each link to the difference in the User action log The number occurred in record information calculates each link and records the probability occurred in information at one；According to the link To recording at one, the probability occurred in information, first event is corresponding described in the link pair is linked at a record information Second event described in the probability of middle appearance and the link pair is corresponding be linked in a record information occur it is general Rate calculates the point mutual information of the corresponding link of the first event link corresponding with the second event of the link pair PMI；The point of the corresponding link of the first event of each link pair link corresponding with the second event is mutual Information PMI carries out addition calculating, obtains the degree of correlation of the first event and the second event.

For example, for (URL1, URL3), as shown in table 2, the co-occurrence of (URL1, URL3) in the User action log Number is 1, and it includes 3 record information which, which has altogether, then (URL1, URL3) occurs in a record information Probability is 1/3.URL1 is appeared in the Session of the Session of user 1, user 2, i.e. the number that URL1 occurs is 2, URL1 It is 2/3 that the probability occurred in information is recorded at one.URL3 is appeared in the Session of user 2 and the Session of user 3, i.e., The number that URL3 occurs is that the probability that 2, URL3 occurs in a record information is 2/3.(URL1, URL3) is recorded at one The probability occurred in information is denoted as P (URL1&URL3), and URL1 is recorded the probability occurred in information at one and is denoted as P (URL1), URL3 is recorded into the probability occurred in information at one and is denoted as P (URL3), optionally, each URL goes out in a record information Existing probability is not 0.The point mutual information of URL1 and URL3 can be calculated according to P (URL1&URL3), P (URL1) and P (URL3) (Pointwise Mutual Information, PMI) i.e. PMI (URL1, URL3), PMI (URL1, URL3) can be according to following public affairs Formula is calculated:

PMI (URL1, URL3)=log2 (P (URL1&URL3)/(P (URL1) * P (URL3)))

Can similarly calculate PMI (URL1, URL4), PMI (URL2, URL3), PMI (URL2, URL4), PMI (URL3, URL4), by PMI (URL1, URL3), PMI (URL1, URL4), PMI (URL2, URL3), PMI (URL2, URL4), PMI (URL3, URL4 addition calculating) is carried out, the degree of correlation of the first event He the second event is obtained.

Without loss of generality, for example, the corresponding multiple links of first event constitute the first lists of links, the first lists of links table It is shown as { A1, A2 ..., An }, the corresponding multiple links of second event constitute the second lists of links, and the second lists of links is expressed as {B1,B2,…,Bn}.Traverse each link in the first lists of links { A1, A2 ..., An }, by the link currently traversed with Each link in second lists of links { B1, B2 ..., Bn } respectively constitutes a link pair, to obtain following multiple links It is right: (A1, B1), (A1, B2) ..., (An, B1) ..., (An, Bn).According to (A1, B1), (A1, B2) ..., (An, B1) ..., A kind of method that (An, Bn) calculates the degree of correlation of the first event and the second event is: by (A1, B1), (A1, B2) ..., (An, B1) ..., the co-occurrence number of (An, Bn) respectively in User action log carry out addition calculating, obtain the first event and The degree of correlation of the second event.Another method is: calculating PMI (A1, B1) according to formula as above, PMI (A1, B2) ..., PMI (An, B1) ..., PMI (An, Bn), further by PMI (A1, B1), PMI (A1, B2) ..., PMI (An, B1) ..., PMI (An, Bn addition calculating) is carried out, the degree of correlation of the first event He the second event is obtained.

The embodiment of the present invention will be traversed currently by traversing each link in the corresponding multiple links of the first event To the first event it is corresponding link it is corresponding with the second event it is multiple link in each link respectively constitute one A link pair, according to each link to the number occurred in the different record information in the User action log, meter The degree of correlation for calculating the first event and the second event further improves the computational accuracy of the event degree of correlation.

Fig. 4 be another embodiment of the present invention provides event relatedness computation method flow diagram.In the base of above-described embodiment It is described crucial according to the corresponding multiple search keys of the first event, the corresponding multiple search of the second event on plinth Word and the User action log, calculate the degree of correlation of the first event and the second event, specifically comprise the following steps:

Each search key in step 401, the corresponding multiple search keys of the traversal first event, will be current It is every in the corresponding search key of the first event multiple search keys corresponding with the second event traversed A search key respectively constitutes a search key pair.

For example, the corresponding multiple search keys of first event constitute the first search key list, the first search is crucial Word list is expressed as { query1, query2, query3 }, and the corresponding multiple search keys of second event constitute the second search and close Keyword list, the second search key list are expressed as { query2, query3, query4 }.Traverse the first search key column Each search key in table { query1, query2, query3 } searches for the search key currently traversed and second Each search key in lists of keywords { query2, query3, query4 } respectively constitutes a search key pair, from And obtain following multiple search keys pair: (query1, query2), (query1, query3), (query1, query4), (query2, query3), (query2, query4), (query3, query4).

Step 402 records in information the difference in the User action log according to each described search keyword The number of appearance calculates the degree of correlation of the first event and the second event.

Further statistics (query1, query2), (query1, query3), (query1, query4), (query2, Query3), (query2, query4), (query3, query4) the co-occurrence number in the User action log respectively, specifically , can according to statistical result as listed in Table 1, inquire (query1, query2), (query1, query3), (query1, Query4), (query2, query3), (query2, query4), (query3, query4) are respectively in the User action log Co-occurrence number.Further, according to (query1, query2), (query1, query3), (query1, query4), The co-occurrence in the User action log is secondary respectively by (query2, query3), (query2, query4), (query3, query4) Number, calculates the degree of correlation of the first event He the second event.

Optionally, described search keyword is to the number occurred in the different record information in the User action log For the number of the record information in the User action log including described search keyword pair.As shown in table 1, search key Co-occurrence number to (query1, query2) is 2.For example, including the Session of user 1, user 2 in User action log The Session of Session and user 3.Wherein, the Session of user 1 be expressed as (session_1, query1, URL1, Query2, URL2), the Session of user 2 be expressed as (session_2, query1, URL2, query2, URL3, query3, URL1), the Session of user 3 is expressed as (session_3, query2, URL2, query3, URL3, query4, URL4).It can See, include in the Session of user 1 (query1, query2), include in the Session of user 2 (query1, query2), uses It does not include (query1, query2) including in the Session at family 3 there are two Session that is, in the User action log (query1, query2).So the co-occurrence number namely user behavior day of search key to (query1, query2) The number of Session in will including (query1, query2).The co-occurrence number of other search keys pair is similarly in this, no It repeats one by one again.

In addition, described record in information the difference in the User action log according to each described search keyword The number of appearance calculates the degree of correlation of the first event and the second event, including following several feasible implementations:

One kind is feasible to be achieved in that: by each described search keyword to the difference in the User action log The number occurred in record information carries out addition calculating, obtains the degree of correlation of the first event and the second event.

For example, by (query1, query2), (query1, query3), (query1, query4), (query2, Query3), the co-occurrence number in the User action log carries out phase respectively by (query2, query4), (query3, query4) Add calculating, obtains the degree of correlation of the first event He the second event.

Another kind is feasible to be achieved in that: according to each described search keyword in the User action log The number occurred in difference record information, calculates each described search keyword and records the probability occurred in information at one； According to described search keyword to recording the probability occurred in information, first event described in described search keyword pair at one Corresponding search key records second event described in the probability and described search keyword pair occurred in information at one Corresponding search key records the probability occurred in information at one, calculates first thing of described search keyword centering The point mutual information PMI of the corresponding search key of part and the corresponding search key of the second event；By each described search The corresponding search key of the first event of keyword centering and the point of the corresponding search key of the second event are mutual Information PMI carries out addition calculating, obtains the degree of correlation of the first event and the second event.

For example, as shown in table 1, (query1, query2) is in the User action log for (query1, query2) In co-occurrence number be 2, the User action log have altogether include 3 record information, then (query1, query2) one record The probability occurred in information is 2/3.Query1 is appeared in the Session of user 1 and the Session of user 2, i.e., query1 goes out Existing number is that the probability that 2, query1 occurs in a record information is 2/3.Query2 appear in user 1 Session, In the Session of user 2 and the Session of user 3, i.e. the number that query2 occurs is 3, query2 in a record information The probability of appearance is 3/3.(query1, query2) is recorded into the probability occurred in information at one and is denoted as P (query1& Query2), query1 is recorded into the probability occurred in information at one and is denoted as P (query1), query2 is believed in a record The probability occurred in breath is denoted as P (query2), and optionally, it is not 0 that each query, which records the probability occurred in information at one,. The point mutual information PMI of query1 and query2 can be calculated according to P (query1&query2), P (query1) and P (query2) That is PMI (query1, query2), PMI (query1, query2) can be calculated according to the following formula:

PMI (query1, query2)=log2 (P (query1&query2)/(P (query1) * P (query2)))

Can similarly calculate PMI (query1, query3), PMI (query1, query4), PMI (query2, query3), PMI (query2, query4), PMI (query3, query4), by PMI (query1, query2), PMI (query1, Query3), PMI (query1, query4), PMI (query2, query3), PMI (query2, query4), PMI (query3, Query4 addition calculating) is carried out, the degree of correlation of the first event He the second event is obtained.

Without loss of generality, for example, the corresponding multiple search keys of first event constitute the first search key list, the One search key list is expressed as { A1, A2 ..., An }, and the corresponding multiple search keys of second event constitute the second search Lists of keywords, the second search key list are expressed as { B1, B2 ..., Bn }.Traverse the first search key list A1, A2 ..., An in each search key, by the search key currently traversed and the second search key list B1, B2 ..., Bn in each search key respectively constitute a search key pair, to obtain following multiple search keys Word pair: (A1, B1), (A1, B2) ..., (An, B1) ..., (An, Bn).According to (A1, B1), (A1, B2) ..., (An, B1) ..., A kind of method that (An, Bn) calculates the degree of correlation of the first event and the second event is: by (A1, B1), (A1, B2) ..., (An, B1) ..., the co-occurrence number of (An, Bn) respectively in User action log carry out addition calculating, obtain the first event and The degree of correlation of the second event.Another method is: calculating PMI (A1, B1) according to formula as above, PMI (A1, B2) ..., PMI (An, B1) ..., PMI (An, Bn), further by PMI (A1, B1), PMI (A1, B2) ..., PMI (An, B1) ..., PMI (An, Bn addition calculating) is carried out, the degree of correlation of the first event He the second event is obtained.

The embodiment of the present invention is crucial by traversing each search in the corresponding multiple search keys of the first event Word, the corresponding search key of the first event currently traversed multiple search corresponding with the second event are crucial Each search key in word respectively constitutes a search key pair, according to each described search keyword in the use The number occurred in different record information in the user behaviors log of family, it is related to the second event to calculate the first event Degree, further improves the computational accuracy of the event degree of correlation.

Fig. 5 is the structural schematic diagram of event relatedness computation device provided in an embodiment of the present invention；Fig. 6 is that the present invention is implemented The structural schematic diagram for the event relatedness computation device that example provides.Event relatedness computation device provided in an embodiment of the present invention can To execute the process flow of event relatedness computation embodiment of the method offer, as shown in figure 5, event relatedness computation device 50 wraps It includes: obtaining module 51, the first determining module 52, the second determining module 53, the first computing module 54；Alternatively, as shown in fig. 6, thing Part relatedness computation device 50 includes: to obtain module 51, the first determining module 52, the second determining module 53, the second computing module 55.Wherein, module 51 is obtained for obtaining User action log, and the User action log includes multiple record information, described The search behavior of the corresponding user of each record information in multiple record information, the record information includes at least one At least one link that a search key and the user click；First determining module 52 is used for according to the first thing in event base The corresponding multiple links of part and the User action log, determine the corresponding multiple search keys of the first event, described First event corresponds to multiple first information, one a pair of the corresponding multiple links of first event and the multiple first information It answers, each search key in the corresponding multiple search keys of the first event is for searching for and clicking the multiple the The first information of at least one of one information；Second determining module 53 is used for corresponding more according to second event in the event base A link and the User action log, determine the corresponding multiple search keys of the second event, the second event pair Multiple second information are answered, the corresponding multiple links of second event and the multiple second information correspond, and described second Each search key in the corresponding multiple search keys of event is for searching for and clicking in the multiple second information At least one second information；First computing module 54 is used for according to the corresponding multiple links of the first event, second thing The corresponding multiple links of part and the User action log, calculate the degree of correlation of the first event and the second event；The Two computing modules 55 are used to multiple search according to the corresponding multiple search keys of the first event, the second event are corresponding Rope keyword and the User action log, calculate the degree of correlation of the first event and the second event.

Optionally, the first determining module 52 is specifically used for: corresponding multiple according to first event described in the event base Each link in link obtains from the User action log and links at least one corresponding search key with described； By at least one corresponding search key of each of multiple links link, it is determined as the first event Corresponding multiple search keys.

Optionally, the second determining module 53 is specifically used for: corresponding multiple according to second event described in the event base Each link in link obtains from the User action log and links at least one corresponding search key with described； By at least one corresponding search key of each of multiple links link, it is determined as the second event Corresponding multiple search keys.

Optionally, the first computing module 54 includes: the first Traversal Unit 541 and the first computing unit 542；First traversal is single Member 541 is for traversing each link in the corresponding multiple links of the first event, first thing that will currently traverse The corresponding link of part it is corresponding with the second event it is multiple link in each link respectively constitute one and link pair；First meter Unit 542 is calculated to be used to record the difference in the User action log time occurred in information according to each link Number, calculates the degree of correlation of the first event and the second event.

Optionally, the first computing unit 542 is specifically used for: by each link in the User action log The number occurred in difference record information carries out addition calculating, obtains the degree of correlation of the first event and the second event.

Optionally, the first computing unit 542 is specifically used for: according to each link in the User action log Different record information in the number that occurs, calculate each link and record the probability occurred in information at one；According to To recording at one, the probability occurred in information, first event is corresponding described in the link pair is linked at one for the link Second event described in the probability occurred in record information and the link pair is corresponding to be linked in a record information Existing probability calculates the point of the corresponding link of the first event link corresponding with the second event of the link pair Mutual information PMI；By the corresponding link of the first event of each link pair link corresponding with the second event Point mutual information PMI carry out addition calculating, obtain the degree of correlation of the first event and the second event.

Optionally, the second computing module 55 includes: the second Traversal Unit 551 and the second computing unit 552；Second traversal is single Member 551 will be traversed currently for traversing each search key in the corresponding multiple search keys of the first event The corresponding search key of the first event multiple search keys corresponding with the second event in each search Keyword respectively constitutes a search key pair；Second computing unit 552 be used for according to each described search keyword to The number occurred in different record information in the User action log calculates the first event and the second event The degree of correlation.

Optionally, the second computing unit 552 is specifically used for: by each described search keyword in the user behavior day The number occurred in different record information in will carries out addition calculating, obtains the phase of the first event and the second event Guan Du.

Optionally, the second computing unit 552 is specifically used for: according to each described search keyword in the user behavior The number occurred in different record information in log, calculates each described search keyword and occurs in a record information Probability；It is recorded described in the probability occurred in information, described search keyword pair according to described search keyword at one The corresponding search key of first event is recorded at one described in the probability and described search keyword pair occurred in information The corresponding search key of second event records the probability occurred in information at one, calculates the institute of described search keyword centering State the point mutual information PMI of the corresponding search key of first event and the corresponding search key of the second event；It will be each The corresponding search key of the first event of described search keyword centering and the corresponding search of the second event are crucial The point mutual information PMI of word carries out addition calculating, obtains the degree of correlation of the first event and the second event.

Fig. 5 and the event relatedness computation device of embodiment illustrated in fig. 6 can be used for executing the technology of above method embodiment Scheme, it is similar that the realization principle and technical effect are similar, and details are not described herein again.

Fig. 7 is the structural schematic diagram of equipment provided in an embodiment of the present invention.Equipment provided in an embodiment of the present invention can be held The process flow that part relatedness computation embodiment of the method provides is acted, as shown in fig. 7, equipment 70 includes: memory 71, processor 72, computer program；Wherein, computer program is stored in memory 71, and is configured as being executed by processor 72 to realize Event relatedness computation method as described above.

The equipment of embodiment illustrated in fig. 7 can be used for executing the technical solution of above method embodiment, realization principle and skill Art effect is similar, and details are not described herein again.

In addition, the embodiment of the present invention also provides a kind of computer readable storage medium, it is stored thereon with computer program, institute Computer program is stated to be executed by processor to realize event relatedness computation method described in above-described embodiment.

In several embodiments provided by the present invention, it should be understood that disclosed device and method can pass through it Its mode is realized.For example, the apparatus embodiments described above are merely exemplary, for example, the division of the unit, only Only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components can be tied Another system is closed or is desirably integrated into, or some features can be ignored or not executed.Another point, it is shown or discussed Mutual coupling, direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING or logical of device or unit Letter connection can be electrical property, mechanical or other forms.

The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme 's.

It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list Member both can take the form of hardware realization, can also realize in the form of hardware adds SFU software functional unit.

The above-mentioned integrated unit being realized in the form of SFU software functional unit can store and computer-readable deposit at one In storage media.Above-mentioned SFU software functional unit is stored in a storage medium, including some instructions are used so that a computer It is each that equipment (can be personal computer, server or the network equipment etc.) or processor (processor) execute the present invention The part steps of embodiment the method.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (Read- Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic or disk etc. it is various It can store the medium of program code.

Those skilled in the art can be understood that, for convenience and simplicity of description, only with above-mentioned each functional module Division progress for example, in practical application, can according to need and above-mentioned function distribution is complete by different functional modules At the internal structure of device being divided into different functional modules, to complete all or part of the functions described above.On The specific work process for stating the device of description, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations；To the greatest extent Pipe present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: its according to So be possible to modify the technical solutions described in the foregoing embodiments, or to some or all of the technical features into Row equivalent replacement；And these are modified or replaceed, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution The range of scheme.

Claims

1. a kind of event relatedness computation method characterized by comprising

User action log is obtained, the User action log includes multiple record information, every in the multiple record information The search behavior of the corresponding user of a record information, the record information include at least one search key and described At least one link that user clicks；

According to the corresponding multiple links of first event in event base and the User action log, determine that the first event is corresponding Multiple search keys, the first event corresponds to multiple first information, the corresponding multiple links of the first event and institute It states multiple first information to correspond, each search key in the corresponding multiple search keys of the first event is used for It searches for and clicks the first information of at least one of the multiple first information；

According to the corresponding multiple links of second event in the event base and the User action log, the second event is determined Corresponding multiple search keys, the second event correspond to multiple second information, the corresponding multiple links of the second event It is corresponded with the multiple second information, each search key in the corresponding multiple search keys of the second event For searching for and clicking the second information of at least one of the multiple second information；

According to the corresponding multiple links of the first event, the corresponding multiple links of the second event and the user behavior day Will calculates the degree of correlation of the first event and the second event；Alternatively,

According to the corresponding multiple search keys of the first event, the corresponding multiple search keys of the second event and institute User action log is stated, the degree of correlation of the first event and the second event is calculated.

2. the method according to claim 1, wherein described according to the corresponding multiple chains of first event in event base Connect with the User action log, determine the corresponding multiple search keys of the first event, comprising:

According to each link in the corresponding multiple links of first event described in the event base, from the User action log Middle acquisition links at least one corresponding search key with described；

By at least one corresponding search key of each of multiple links link, it is determined as described first The corresponding multiple search keys of event.

3. the method according to claim 1, wherein described corresponding more according to second event in the event base A link and the User action log, determine the corresponding multiple search keys of the second event, comprising:

According to each link in the corresponding multiple links of second event described in the event base, from the User action log Middle acquisition links at least one corresponding search key with described；

By at least one corresponding search key of each of multiple links link, it is determined as described second The corresponding multiple search keys of event.

4. method according to claim 1-3, which is characterized in that described corresponding more according to the first event A link, the corresponding multiple links of the second event and the User action log, calculate the first event and described the The degree of correlation of two events, comprising:

Each link in the corresponding multiple links of the first event is traversed, the first event currently traversed is corresponding Link it is corresponding with the second event it is multiple link in each link respectively constitute one and link pair；

According to each link to the number occurred in the different record information in the User action log, described in calculating The degree of correlation of first event and the second event.

5. according to the method described in claim 4, it is characterized in that, it is described according to it is each it is described link in the user behavior The number occurred in different record information in log, calculates the degree of correlation of the first event and the second event, comprising:

Each link is subjected to addition meter to the number occurred in the different record information in the User action log It calculates, obtains the degree of correlation of the first event and the second event.

6. according to the method described in claim 4, it is characterized in that, it is described according to it is each it is described link in the user behavior The number occurred in different record information in log, calculates the degree of correlation of the first event and the second event, comprising:

According to each link to the number occurred in the different record information in the User action log, calculate each The link records the probability occurred in information at one；

According to the link to recording the probability occurred in information, the corresponding chain of first event described in the link pair at one It connects and records in information described in the probability that occurs and the link pair that second event is corresponding to be linked at a record at one The probability occurred in information, the corresponding link of the first event for calculating the link pair are corresponding with the second event The point mutual information PMI of link；

The point of the corresponding link of the first event of each link pair link corresponding with the second event is mutual Information PMI carries out addition calculating, obtains the degree of correlation of the first event and the second event.

7. according to the described in any item methods of claim 4-6, which is characterized in that the link is in the User action log In different record information in be in the User action log the include link pair of the number that occurs record information it is a Number.

8. method according to claim 1-3, which is characterized in that described corresponding more according to the first event The corresponding multiple search keys of a search key, the second event and the User action log calculate described first The degree of correlation of event and the second event, comprising:

Each search key in the corresponding multiple search keys of the first event is traversed, described in currently traversing Each search key in the corresponding search key of first event multiple search keys corresponding with the second event Respectively constitute a search key pair；

According to each described search keyword to the number occurred in the different record information in the User action log, meter Calculate the degree of correlation of the first event and the second event.

9. according to the method described in claim 8, it is characterized in that, it is described according to each described search keyword in the use The number occurred in different record information in the user behaviors log of family, it is related to the second event to calculate the first event Degree, comprising:

Each described search keyword carries out the number occurred in the different record information in the User action log It is added and calculates, obtain the degree of correlation of the first event and the second event.

10. according to the method described in claim 8, it is characterized in that, it is described according to each described search keyword to described The number occurred in different record information in User action log, it is related to the second event to calculate the first event Degree, comprising:

According to each described search keyword to the number occurred in the different record information in the User action log, meter It calculates each described search keyword and records the probability occurred in information at one；

It is recorded first described in the probability occurred in information, described search keyword pair according to described search keyword at one The corresponding search key of event is recorded at one second described in the probability and described search keyword pair occurred in information The corresponding search key of event records the probability that occurs in information at one, calculates described the of described search keyword centering The point mutual information PMI of the corresponding search key of one event and the corresponding search key of the second event；

The corresponding search key of the first event of each described search keyword centering and the second event is corresponding The point mutual information PMI of search key carry out addition calculating, obtain the degree of correlation of the first event and the second event.

11. according to the described in any item methods of claim 8-10, which is characterized in that described search keyword is in the user The number occurred in different record information in user behaviors log is in the User action log including described search keyword pair Record information number.

12. a kind of event relatedness computation device characterized by comprising

Module is obtained, for obtaining User action log, the User action log includes multiple record information, the multiple note The search behavior of the corresponding user of each record information in information is recorded, the record information includes at least one search At least one link that keyword and the user click；

First determining module is used for according to the corresponding multiple links of first event in event base and the User action log, really Determine the corresponding multiple search keys of the first event, the first event corresponds to multiple first information, the first event Corresponding multiple links and the multiple first information correspond, in the corresponding multiple search keys of the first event Each search key is for searching for and clicking the first information of at least one of the multiple first information；

Second determining module, for according to the corresponding multiple links of second event in the event base and the user behavior day Will, determines the corresponding multiple search keys of the second event, and the second event corresponds to multiple second information, described second The corresponding multiple links of event and the multiple second information correspond, the corresponding multiple search keys of the second event In each search key for searching for and click the second information of at least one of the multiple second information；

First computing module, for according to the corresponding multiple links of the first event, the corresponding multiple chains of the second event Connect with the User action log, calculate the degree of correlation of the first event and the second event；Alternatively,

Second computing module, for corresponding according to the corresponding multiple search keys of the first event, the second event Multiple search keys and the User action log, calculate the degree of correlation of the first event and the second event.

13. event relatedness computation device according to claim 12, which is characterized in that the first determining module is specifically used In:

14. event relatedness computation device according to claim 12, which is characterized in that the second determining module is specifically used In:

15. the described in any item event relatedness computation devices of 2-14 according to claim 1, which is characterized in that first meter Calculating module includes:

First Traversal Unit will be traversed currently for traversing each link in the corresponding multiple links of the first event The first event it is corresponding link it is corresponding with the second event it is multiple link in each link respectively constitute one Link pair；

First computing unit, for recording in information to the difference in the User action log according to each link Existing number calculates the degree of correlation of the first event and the second event.

16. event relatedness computation device according to claim 15, which is characterized in that first computing unit is specific For:

17. event relatedness computation device according to claim 15, which is characterized in that first computing unit is specific For:

18. the described in any item event relatedness computation devices of 2-14 according to claim 1, which is characterized in that second meter Calculating module includes:

Second Traversal Unit, for traversing each search key in the corresponding multiple search keys of the first event, The corresponding search key of the first event multiple search keys corresponding with the second event that will currently traverse In each search key respectively constitute a search key pair；

Second computing unit, for being believed according to each described search keyword the different records in the User action log The number occurred in breath calculates the degree of correlation of the first event and the second event.

19. the event relatedness computation device according to claim 18, which is characterized in that described second calculates list Member is specifically used for:

20. the event relatedness computation device according to claim 18, which is characterized in that described second calculates list Member is specifically used for:

21. a kind of equipment characterized by comprising

Memory；

Processor；And

Computer program；

Wherein, the computer program stores in the memory, and is configured as being executed by the processor to realize such as Any method in claim 1-11.

22. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program Such as claim 1-11 described in any item methods are realized when being executed by processor.