CN103793390B - Querying condition similarity determines method, Object Query method and relevant apparatus - Google Patents
Querying condition similarity determines method, Object Query method and relevant apparatus Download PDFInfo
- Publication number
- CN103793390B CN103793390B CN201210421325.8A CN201210421325A CN103793390B CN 103793390 B CN103793390 B CN 103793390B CN 201210421325 A CN201210421325 A CN 201210421325A CN 103793390 B CN103793390 B CN 103793390B
- Authority
- CN
- China
- Prior art keywords
- querying condition
- condition
- mrow
- component
- querying
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
Abstract
This application discloses a kind of querying condition similarities to determine method, Object Query method and relevant apparatus, including：Inquiry click logs are obtained, inquiry click logs include each querying condition and the corresponding object clicked on；And for each querying condition in each querying condition, determine that the querying condition corresponds to the number that each object clicked on is clicked under the querying condition；And according to definite number, the click vector of each querying condition is determined, wherein, the component of vector is clicked on as object, the situation that component value characterization object is clicked under the querying condition；And for each two querying condition for meeting preset condition in each querying condition, according to the click vector of click the second querying condition of vector sum of wherein the first querying condition, determine the similarity between the first querying condition and the second querying condition.Using scheme provided by the embodiments of the present application, the accuracy of the similarity between definite different querying conditions is improved.
Description
Technical field
This application involves Internet technical field and field of computer technology more particularly to a kind of querying condition similarity are true
Determine method, Object Query method and relevant apparatus.
Background technology
In existing Internet technology, the information of some objects can be generally issued on website, for logging in the website
Family browses and the further post-treatment operations for being directed to special object.For example, by taking e-commerce website as an example, object tool
Body can be the product of seller user's issue, and the information of object can be specifically the description information of product, the attribute information of product,
And purchase information of product etc., log in e-commerce website user can by browsing the various information of release product,
The details of the product is solved, and can further perform collection, buy or recommend the processing operations such as other users；With society
Exemplified by area website, object can be specifically the model of community users issue, and the information of object can be specifically the description letter of model
Breath, content information of model etc., the various information for the model that the browsing user of login community website can be issued by browsing,
The details of the model is solved, and can further perform collection, money order receipt to be signed and returned to the sender or recommend the processing operations such as other users.
At present, website can according to its setting means on the page to user's show object, such as carried out according to different classifications
Show, and the function of Object Query can be provided a user, i.e., after querying condition input by user is received, from offer
Object complete or collected works in inquire the object for meeting the querying condition, and query result is presented to user, so that user is further
The operations such as clicked on, browsed to the object that query result includes, wherein, querying condition can be searching keyword.
However, during above-mentioned Object Query, often due to the speech habits of different user, in fact it could happen that user is defeated
The query intention expressed by querying condition entered is indefinite, is not allowed so as to cause the object for meeting the querying condition inquired
Really, not comprehensively.
To solve the above-mentioned problems, the prior art is proposed based on the similarity between each querying condition, determines inquiry item
The similar querying condition of part, and according to the scheme of definite similar querying condition progress Object Query, that is, determine and input
The similar querying condition of querying condition, and Object Query is carried out according to the similar querying condition, the object that inquiry is obtained is made
For a part for this query result, so as to carry out Object Query compared to the querying condition only according to input, inquiry knot is improved
The accuracy of fruit and comprehensive.
At present, it is whether similar according to the content of text of each querying condition in the similarity between determining querying condition
It is determined, for example, centre word is extracted from the content of text of each querying condition, and between the centre word according to each querying condition
Whether identical and near justice, determine the similarity between different querying conditions, similarity met to the querying condition of preset condition,
As similar querying condition.
However, determining the scheme of querying condition similarity based on the content of text of querying condition, it is only capable of determining limited
Similar querying condition, and it is that user intentionally gets that can not ensure based on the object that the similar querying condition determined inquires,
Also it is possible to due to centre word extraction error and the error for lexical comprehension, cause definite similar querying condition with
Similarity essence between the querying condition of input is not high, so as to cause finally inquire about obtained object still inaccuracy with not
Comprehensively.
The content of the invention
In view of this, the embodiment of the present application provides a kind of querying condition similarity and determines method, Object Query method and phase
Device is closed, for solving the problems, such as to determine inaccuracy for the similarity between different querying conditions in the prior art.
The embodiment of the present application is achieved through the following technical solutions：
The embodiment of the present application provides a kind of querying condition similarity and determines method, including：
Inquiry click logs are obtained, the inquiry click logs include each querying condition and the corresponding object clicked on；
For each querying condition in each querying condition, determine that the querying condition corresponds to each object clicked on and looked at this
The number being clicked under the conditions of inquiry；
According to definite number, determine that the click of each querying condition is vectorial, wherein, the component of vector is clicked on as object,
The situation that component value characterization object is clicked under the querying condition；
For each two querying condition for meeting preset condition in each querying condition, according to the point of wherein the first querying condition
The click vector of the second querying condition of vector sum is hit, determines the phase between first querying condition and second querying condition
Like degree.
The embodiment of the present application additionally provides a kind of Object Query method, including：
Obtain the current queries condition of input；
Obtain the similar querying condition that the predetermined similarity to the current queries condition reaches similarity threshold；
Meet the object of the current queries condition in the full Integrated query of object and meet the similar inquiry obtained
The object of condition；
Wherein, determine that reaching the similar of similarity threshold to the similarity of the current queries condition looks into the following way
Inquiry condition：
Inquiry click logs are obtained, the inquiry click logs include the current queries condition and other each inquiry items
Part and the corresponding object clicked on；
For each querying condition in the current queries condition and other each querying conditions, the querying condition pair is determined
The number that each object that should be clicked on is clicked under the querying condition；
According to definite number, determine that the click of each querying condition is vectorial, wherein, the component of vector is clicked on as object,
The situation that component value characterization object is clicked under the querying condition；
For each inquiry item to be determined for meeting preset condition in other each querying conditions with the current queries condition
Part according to the click for the clicking on the vector sum querying condition to be determined vector of the current queries condition, determines described currently to look into
Similarity between inquiry condition and the querying condition to be determined；
From each querying condition to be determined, determine to reach looking into for similarity threshold with the similarity of the current queries condition
Inquiry condition, the similar querying condition as the current queries condition.
The embodiment of the present application additionally provides a kind of querying condition similarity determining device, including：
Log acquisition unit, for obtain inquire about click logs, it is described inquiry click logs include each querying condition and
The corresponding object clicked on；
Number determination unit for being directed to each querying condition in each querying condition, determines the querying condition corresponding points
The number that each object hit is clicked under the querying condition；
Vector determination unit for the definite number of basis, determines that the click of each querying condition is vectorial, wherein, it clicks on
The component of vector is object, and component value characterizes the situation that object is clicked under the querying condition；
Similarity determining unit, for being directed to each two querying condition for meeting preset condition in each querying condition, according to
Wherein the first querying condition click the second querying condition of vector sum click vector, determine first querying condition with it is described
Similarity between second querying condition.
The embodiment of the present application additionally provides a kind of Object Query device, including：
Input unit, for obtaining the current queries condition of input；
Condition acquiring unit reaches similarity threshold for obtaining predetermined and current queries condition similarity
The similar querying condition of value；
Query unit obtains for meeting the object of the current queries condition in the full Integrated query of object and meeting
The similar querying condition object；
Similarity determining unit, for determining to reach similar to the similarity of the current queries condition in the following way
Spend the similar querying condition of threshold value：
Inquiry click logs are obtained, the inquiry click logs include the current queries condition and other each inquiry items
Part and the corresponding object clicked on；And each querying condition in the current queries condition and other each querying conditions is directed to, really
The fixed querying condition corresponds to the number that each object clicked on is clicked under the querying condition；And according to definite number, really
The click vector of fixed each querying condition, wherein, the component of vector is clicked on as object, and component value characterizes object in the querying condition
Under situation about being clicked；And it is treated really for each of preset condition is met in other each querying conditions with the current queries condition
Determine querying condition, according to the click vector for clicking on the vector sum querying condition to be determined of the current queries condition, determine institute
State the similarity between current queries condition and the querying condition to be determined；And from each querying condition to be determined, determine with
The similarity of the current queries condition reaches the querying condition of similarity threshold, is looked into as the similar of current queries condition
Inquiry condition.
In above-mentioned at least one technical solution provided by the embodiments of the present application, the similarity between querying condition is determined
When, the inquiry click logs for including each querying condition and the corresponding object clicked on are obtained first, and in each querying condition
Each querying condition determines that the querying condition corresponds to the number that each object clicked on is clicked under the querying condition, then
According to definite number, determine that the click of each querying condition is vectorial, wherein, the component of vector is clicked on as object, component value table
The situation that object is clicked under the querying condition has been levied, has then been looked into for each two for meeting preset condition in each querying condition
Inquiry condition according to the click vector of click the second querying condition of vector sum of wherein the first querying condition, determines the first inquiry item
Similarity between part and the second querying condition.During due in said program with reference to being inquired about based on each querying condition
To the number information that the object that inquiry obtains is clicked on, and the number information can be characterized when being inquired about using querying condition
Desired obtained object, so the similarity between two querying conditions is determined based on the number information, it can be more accurately
In the actual similarity degree of the object this aspect intentionally got for user between two querying conditions of reaction, that is, improve really
The accuracy of similarity between fixed different querying conditions.
Other features and advantage will illustrate in the following description, also, partly become from specification
It obtains it is clear that being understood by implementing the application.The purpose of the application and other advantages can be by the explanations write
Specifically noted structure is realized and obtained in book, claims and attached drawing.
Description of the drawings
Attached drawing is used for providing further understanding of the present application, and a part for constitution instruction, implements with the application
Example for explaining the application, does not form the limitation to the application together.In the accompanying drawings：
Fig. 1 is the flow chart that querying condition similarity provided by the embodiments of the present application determines method；
Fig. 2 is each querying condition and each object and the schematic diagram of number of clicks in the embodiment of the present application；
Fig. 3 is according to the similarity for clicking on definite first querying condition of vector and the second querying condition in the embodiment of the present application
Process chart；
Fig. 4 is the flow chart of Object Query method provided by the embodiments of the present application；
Fig. 5 is the structure diagram of querying condition similarity determining device provided by the embodiments of the present application；
Fig. 6 is the structure diagram of Object Query device provided by the embodiments of the present application.
Specific embodiment
In order to provide the implementation for improving the accuracy that the similarity between different querying conditions determines, the application is implemented
Example provides a kind of querying condition similarity and determines method, Object Query method and relevant apparatus, which can apply
In the process that querying condition is used to carry out Object Query, a kind of method was both can be implemented as, a kind of device can also be embodied as.With
Lower combination Figure of description illustrates the preferred embodiment of the application, it should be understood that preferred embodiment described herein
The application is only used for describing and explaining, is not used to limit the application.And in the case where there is no conflict, the implementation in the application
Feature in example and embodiment can be mutually combined.
The embodiment of the present application provides a kind of querying condition similarity and determines method, as shown in Figure 1, including：
Step 101 obtains inquiry click logs, and inquiry click logs include each querying condition and corresponding pair clicked on
As.
Step 102, for each querying condition in each querying condition, determine the querying condition correspond to click on it is each right
As the number being clicked under the querying condition.
The definite number of step 103, basis determines the click vector of each querying condition, wherein, click on vectorial component
Situation about being clicked for object, component value characterization object under the querying condition.
Step 104, for each two querying condition for meeting preset condition in each querying condition, according to the wherein first inquiry
The click vector of click the second querying condition of vector sum of condition, determines the phase between the first querying condition and the second querying condition
Like degree.
At present, for click, browsing and processing operation of the user to the object issued on Website page, can record corresponding
Access log, one of which access log are inquiry click logs, and inquiry click logs include user and carry out Object Query institute
Using querying condition and the record clicked on for the object showed corresponding with the querying condition, i.e. record uses one
A querying condition carries out the object clicked on after Object Query.
Due to including each querying condition and the corresponding object clicked in inquiry click logs, so in above-mentioned steps 102
Each querying condition can be directed to, determines that the querying condition corresponds to time that each object clicked on is clicked under the querying condition
Number can be specifically the inquiry click logs that behavioral statistics are clicked in acquisition for the inquiry in set period of time, and based on acquisition
Inquiry click logs, determine that each querying condition corresponds to the number that each object clicked on is clicked under the querying condition.
After determined number, you can by above-mentioned steps 103, according to definite number, to determine each querying condition
Vector is clicked on, the following two kinds mode specifically may be employed：
First way：For each querying condition, the component for determining the click vector of the querying condition is the inquiry item
Part corresponds to the object clicked on, and for convenience of description, the corresponding object clicked on subsequently is referred to as the first object；
Determine the number that the component value of component is clicked for the component the first object of correspondence under the querying condition.
Using the first way, the click vector of each querying condition can be quickly determined, hence into subsequently looking into
The definite step of inquiry condition similarity.
However, using the first way, it may be limited due to the information content in acquired inquiry click logs, it leads
It causes to correspond to the object clicked on using a querying condition in practice, not recorded in the inquiry click logs of acquisition, so that
It obtains the information and is not used to determining for subsequent query condition similarity, to solve this problem, also proposed in the embodiment of the present application
The following preferred second way：
The second way：For each querying condition, the component for determining the click vector of the querying condition is the inquiry item
Part corresponds to the first object clicked on and the corresponding other querying conditions for clicking on the first object are corresponded in the object clicked on, except the
Object outside an object for convenience of description, subsequently by the object in addition to the first object, is referred to as the second object；
That is, when the querying condition corresponds to first object clicked on and another querying condition corresponding points
The object hit, and another querying condition is corresponded in the object clicked on, and is additionally included in the querying condition and is corresponded to what is clicked on
The object being not present in first object, then the object is i.e. as corresponding second object of the querying condition；
For example, example as shown in Figure 2, wherein, querying condition includes Q1, Q2, Q3 and Q4, object include AD1, AD2,
AD3 and AD4；
It is AD3 that Q1, which corresponds to the object clicked on,；
It is AD2 and AD3 that Q2, which corresponds to the object clicked on,；
It is AD1, AD2 and AD4 that Q3, which corresponds to the object clicked on,；
It is AD1 and AD2 that Q4, which corresponds to the object clicked on,.
It is AD3 that then Q1, which corresponds to the first object clicked on, and AD3 is that Q2 corresponds to the first object clicked on, and Q2 corresponds to what is clicked on
First object further includes AD2, then AD2 is corresponding second objects of Q1；Correspondingly, Q2 correspond to the first object clicked on for AD2 and
Corresponding second object of AD3, Q2 is AD1 and AD4；Q3 corresponds to the first object clicked on as AD1, AD2 and AD4, Q3 corresponding the
Two objects are AD3；The first object that Q4 corresponds to click is AD1 and corresponding second object of AD2, Q4 is AD3 and AD4.
After determining that querying condition corresponds to the first object clicked on and corresponding second object, for each first pair
As, the component value for determining the first object respective components is the number that first object is clicked under the querying condition, for
Each second object determines the component value of the second object respective components using equation below：
Wherein, Q be the querying condition, biFor i-th of second objects in all second objects, W (Q, bi) it is the inquiry item
B in the click vector of partiThe component value of respective components, Q (bi) it is corresponding click biQuerying condition set, Q (qi) it is to correspond to
Clicking on object includes the set of querying condition of at least one first object and at least one second object, qiFor Q (qi) in
I-th of querying condition, W (q, bi) it is biThe number being clicked under querying condition q, Sum (q) are that querying condition q corresponds to what is clicked on
The number and value that each object is clicked under querying condition q, DecayFactor are decay factor, for example, 0.85 can be set to
Deng, the set for the object that A (Q) clicks on for querying condition Q correspondences, A (qi) it is querying condition qiThe set of the corresponding object clicked on, W
(Q,ax) it is object axThe number being clicked under querying condition Q, W (qi,ax) it is object axIn querying condition qiUnder be clicked
Number, Sum (ax) it is object axRespectively object a is clicked on correspondingxEach querying condition under the number and value that is clicked.
Still by taking Fig. 2 as an example, for convenience of description, the number for being subsequently clicked object under querying condition referred to as inquires about item
Part is to the number of clicks of object, then in Fig. 2, the number of clicks of Q1 to AD3 is 132456, the number of clicks point of Q2 to AD2 and AD3
Not Wei 100 and 123456, the number of clicks of Q3 to AD1, AD2 and AD4 are respectively the click of 20,10 and 10, Q4 to AD1 and AD2
Number is respectively 10 and 100.
The determination process of the component value of its second object respective components in the click vector of Q1 is described in detail as follows below：
It is AD3 that Q1, which corresponds to the first object clicked on, and corresponding second object is AD2, for Q1, b1For AD2；Then corresponding points
Hitting object includes at least one first object（AD3）With at least one second object（AD2）Querying condition set Q (qi)
Including Q2, for Q1, q1For Q2.
A (Q1) includes AD3, A (q1) it is A (Q2), including AD2 and AD3, then for q1, axFor AD3, using above-mentioned formula meter
Calculate Nod (q1) process it is as follows：
For b1（AD2）, Q (b1) it is Q (AD2), including Q2, Q3 and Q4, Q (qi) include Q2, then for b1, q Q2 adopt
W (Q1, b are calculated with above-mentioned formula1) process it is as follows：
So that it is determined that the click vector of querying condition Q1 includes AD2 respective components and AD3 respective components, component value are respectively
43.95 with 132456.
The determination process of the component value of its second object respective components in the click vector of Q2 is described in detail as follows below：
It is AD2 and AD3 that Q2, which corresponds to the first object clicked on, and corresponding second object is AD1 and AD4, for Q2, b1For
AD1, b2For AD4；Then corresponding object of clicking on includes at least one first object（One of AD2 and AD3）With at least one second
Object（One of AD1 and AD4）Querying condition set Q (qi) include Q3 and Q4, for Q2, q1For Q3, q2For Q4.
A (Q2) includes AD2 and AD3, A (q1) it is A (Q3), including AD1, AD2 and AD4, then for q1, axFor AD2, use
Above-mentioned formula calculates Nod (q1) process it is as follows：
A (Q2) includes AD2 and AD3, A (q2) it is A (Q4), including AD1 and AD2, then for q2, axFor AD2, use is above-mentioned
Formula calculates Nod (q2) process it is as follows：
For b1（AD1）, Q (b1) it is Q (AD1), including Q3 and Q4, Q (qi) include Q3 and Q4, then for b1, q for Q3 and
Q4 calculates W (Q2, b using above-mentioned formula1) process it is as follows：
For b2（AD4）, Q (b2) it is Q (AD4), including Q3, Q (qi) include Q3 and Q4, then for b2, q Q3, in use
It states formula and calculates W (Q2, b2) process it is as follows：
So that it is determined that querying condition Q2 click vector include AD1 respective components, AD2 respective components, AD3 respective components and
AD4 respective components, component value are respectively 5.703,100,123456 and 1.012.
The determination process of the component value of its second object respective components in the click vector of Q3 is described in detail as follows below：
It is AD1, AD2 and AD4 that Q3, which corresponds to the first object clicked on, and corresponding second object is AD3, for Q3, b1For
AD3；Then corresponding object of clicking on includes at least one first object（AD1, AD2 and AD4）With at least one second object（AD3）
Querying condition set Q (qi) include Q2, for Q3, q1For Q2.
A (Q3) includes AD1, AD2 and AD4, A (q1) it is A (Q2), including AD2 and AD3, then for q1, axFor AD2, use
Above-mentioned formula calculates Nod (q1) process it is as follows：
For b1（AD3）, Q (b1) it is Q (AD3), including Q1 and Q2, Q (qi) include Q2, then for b1, q Q2, in use
It states formula and calculates W (Q3, b1) process it is as follows：
So that it is determined that querying condition Q3 click vector include AD1 respective components, AD2 respective components, AD3 respective components and
AD4 respective components, component value are respectively 20,10,4.0443391 and 10.
The determination process of the component value of its second object respective components in the click vector of Q4 is described in detail as follows below：
It is AD1 and AD2 that Q4, which corresponds to the first object clicked on, and corresponding second object is AD3 and AD4, for Q4, b1For
AD3, b2For AD4；Then corresponding object of clicking on includes at least one first object（One of AD1 and AD2）With at least one second
Object（One of AD3 and AD4）Querying condition set Q (qi) include Q2 and Q3, for Q4, q1For Q2, q2For Q3.
A (Q4) includes AD1 and AD2, A (q1) it is A (Q2), including AD2 and AD3, then for q1, axFor AD2, use is above-mentioned
Formula calculates Nod (q1) process it is as follows：
A (Q4) includes AD1 and AD2, A (q2) it is A (Q3), including AD1, AD2 and AD4, then for q2, axFor AD1 and AD2,
Nod (q are calculated using above-mentioned formula2) process it is as follows：
For b1（AD3）, Q (b1) it is Q (AD3), including Q1 and Q2, Q (qi) include Q2 and Q3, then for b1, q Q2 adopt
W (Q4, b are calculated with above-mentioned formula1) process it is as follows：
For b2（AD4）, Q (b2) it is Q (AD4), including Q3, Q (qi) include Q2 and Q3, then for b2, q Q3, in use
It states formula and calculates W (Q4, b2) process it is as follows：
So that it is determined that querying condition Q4 click vector include AD1 respective components, AD2 respective components, AD3 respective components and
AD4 respective components, component value are respectively 10,100,40.4439 and 2.4285775.
The click vector of querying condition is determined using the above-mentioned second way, is the equal of the point determined in first way
On the basis of the amount of hitting to, vector extensions have been carried out, i.e., have expanded the second object respective components according to the above-mentioned second way, and really
The component value of fixed second object respective components, so that the component dimension higher of the click vector of identified querying condition,
So that subsequently definite querying condition similarity is more accurate.
In the embodiment of the present application, for the querying condition determined using above-mentioned first way or the above-mentioned second way
Vector is clicked on, the component value clicked in vector can also be directed to and be further modified processing, for example, can be small by component value
0 is arranged in the final component value of the component of setting component threshold value, to disclosure satisfy that the accuracy of follow-up definite similarity
On the premise of it is required that, calculation amount is reduced, improves treatment effeciency.Wherein, component threshold value can flexibly be set according to actual needs
It puts, is no longer described in detail herein.
After the click vector of each querying condition is determined, you can to enter above-mentioned steps 104, for each querying condition
The middle each two querying condition for meeting preset condition, according to click the second querying condition of vector sum of wherein the first querying condition
Vector is clicked on, determines the similarity between the first querying condition and the second querying condition.
Specifically, when in two querying conditions each corresponding object clicked on there are during at least one same object, can be with
Determine that two querying conditions meet preset condition；Centre word can also be carried out according to the content of text of each querying condition to carry
It takes, and two querying conditions that two querying conditions with identical central word are determined as meeting preset condition.
For two querying conditions for meeting preset condition, in the click vector sum second of known wherein the first querying condition
In the case of the click vector of querying condition, various algorithms of the prior art specifically may be employed, determine the first querying condition
With the similarity between the second querying condition, in the embodiment of the present application, propose that the mode of similarity is identified below, as shown in figure 3,
Specifically include following processing step：
Step 301, the click for click the second querying condition of vector sum for obtaining the first querying condition are vectorial, and determine first
The component that the click vector of querying condition includes and the component that the click vector of the second querying condition includes.
Step 302 judges the component of click vector of the first querying condition and point for clicking on vector of the second querying condition
Whether amount is identical, if identical, 303 is entered step, if it is not the same, entering step 304.
Wherein, it is identical to represent that two quantity for clicking on the component that vector each includes are identical, and the corresponding object of component
It is identical.
Step 303, the click vector and the second querying condition that the first querying condition is determined using default similarity algorithm
The similarity between vector is clicked on, and using result as the similarity between the first querying condition and second querying condition.
Wherein, cosine-algorithm may be employed in default similarity algorithm, and formula is as follows：
Wherein, S1,2For the similarity between the first querying condition and the second querying condition, v1 is the point of the first querying condition
The amount of hitting to, v2 be the second querying condition click vector, n be v1 and v2 component dimension, xiFor the component of i-th of component in v1
Value, yiFor the component value of i-th of component in v2, the corresponding object of i-th of component object corresponding with i-th of component in v2 in v1
It is identical.
Step 304, the component for clicking on vector when the vectorial component of the click of the first querying condition and the second querying condition
There are it is different when, determine the component of the component of the converting vector of the first querying condition and the converting vector of the second querying condition.
Wherein, the component of the converting vector of the first querying condition, the component of the click vector including the first querying condition, with
And exist in the component of the click vector of the second querying condition and do not deposited in the component of the click vector of the first querying condition
Component, the component of the converting vector of the first querying condition is identical with the component of the converting vector of the second querying condition.
That is, determine point of the component of the click vector of the first querying condition and the click vector of the second querying condition
The set of amount, and remove the component of repetition, as a result as the first querying condition converting vector component and the second querying condition
Converting vector component.
Step 305, determine the first querying condition converting vector each component component value and the second querying condition
Converting vector each component component value, be specifically as follows：
For each component of the converting vector of the first querying condition, when the component is in the click vector of the first querying condition
In in the presence of, the component value of the component of the converting vector of the first querying condition for the click vector of the first querying condition this point
The component value of amount, otherwise, the component value of the component is 0.
For each component of the converting vector of the second querying condition, when the component is in the click vector of the second querying condition
In in the presence of, the component value of the component of the converting vector of the second querying condition for the click vector of the second querying condition this point
The component value of amount, otherwise, the component value of the component is 0.
Step 306, the converting vector that the first querying condition is determined using default similarity algorithm and the second querying condition
Similarity between converting vector, and using result as the similarity between the first querying condition and the second querying condition.
Wherein, cosine-algorithm may be employed in default similarity algorithm, and formula is as follows：
Wherein, S1,2For the similarity between the first querying condition and the second querying condition, v1 is turning for the first querying condition
Commutation amount, v2 be the second querying condition converting vector, n be v1 and v2 component dimension, xiFor the component of i-th of component in v1
Value, yiFor the component value of i-th of component in v2, the corresponding object of i-th of component object corresponding with i-th of component in v2 in v1
It is identical.
Further, after the similarity for each two querying condition for determining to meet preset condition, preservation determines
As a result, for subsequently use, specifically, can every record first two in preserve the first querying condition and second respectively
Querying condition, in Section 3 preserve similarity value, and according to similarity value from big to small be ranked sequentially every record.
Method is determined using above-mentioned querying condition similarity provided by the embodiments of the present application, due to being looked into reference to based on each
The number information that the object obtained when inquiry condition is inquired about to inquiry is clicked on, and the number information can characterize use and look into
Inquiry condition desired obtained object when being inquired about, so similar between two querying conditions are determined based on the number information
Degree, the reality that can more accurately react between two querying conditions in the object this aspect intentionally got for user are similar
Degree improves the accuracy of the similarity between definite different querying conditions.
Method is determined based on above-mentioned querying condition similarity provided by the embodiments of the present application, the embodiment of the present application also provides one
Kind Object Query method, as shown in figure 4, including：
Step 401, the current queries condition for obtaining input.
Step 402 obtains the similar inquiry that the predetermined similarity to current queries condition reaches similarity threshold
Condition.
Step 403 meets the object of current queries condition in the full Integrated query of object and meets the similar inquiry obtained
The object of condition；
Wherein, the similar inquiry item for reaching similarity threshold to the similarity of current queries condition is determined in the following way
Part, i.e., above-mentioned querying condition similarity provided by the embodiments of the present application determine method：
Inquiry click logs are obtained, inquiry click logs include current queries condition and other each querying conditions and correspondence
The object of click；
For each querying condition in current queries condition and other each querying conditions, the querying condition corresponding points are determined
The number that each object hit is clicked under the querying condition；
According to definite number, determine that the click of each querying condition is vectorial, wherein, the component of vector is clicked on as object,
The situation that component value characterization object is clicked under the querying condition；
For each querying condition to be determined for meeting preset condition in other each querying conditions with current queries condition, root
According to the click vector for clicking on the vector sum querying condition to be determined of current queries condition, determine that current queries condition is treated really with this
Determine the similarity between querying condition；
From each querying condition to be determined, the inquiry item for reaching similarity threshold with the similarity of current queries condition is determined
Part, the similar querying condition as current queries condition.
Specifically, in relation to clicking on the relevant treatments modes such as vector, preset condition and similarity algorithm, above-mentioned Fig. 1 can refer to
Relevant way into process flow shown in Fig. 3, is no longer described in detail herein.
Based on same inventive concept, method, phase are determined according to the querying condition similarity that the above embodiments of the present application provide
Ying Di, another embodiment of the application additionally provide a kind of querying condition similarity determining device, structure diagram such as Fig. 5 institutes
Show, specifically include：
Log acquisition unit 501 inquires about click logs for obtaining, and the inquiry click logs include each querying condition
And the corresponding object clicked on；
Number determination unit 502 for being directed to each querying condition in each querying condition, determines that the querying condition corresponds to
The number that each object clicked on is clicked under the querying condition；
Vector determination unit 503 for the definite number of basis, determines that the click of each querying condition is vectorial, wherein, point
The component for the amount of hitting to is object, and component value characterizes the situation that object is clicked under the querying condition；
Similarity determining unit 504, for being directed to each two querying condition for meeting preset condition in each querying condition, root
According to the click vector of click the second querying condition of vector sum of wherein the first querying condition, first querying condition and institute are determined
State the similarity between the second querying condition.
Further, vector determination unit 503 specifically for being directed to each querying condition, determine the point of the querying condition
The component for the amount of hitting to corresponds to the first object clicked on for the querying condition；And determine that the component value of component corresponds to first for the component
The number that object is clicked under the querying condition.
Further, vector determination unit 503 specifically for being directed to each querying condition, determine the point of the querying condition
The component for the amount of hitting to corresponds to the first object clicked on and the corresponding other inquiries for clicking on first object for the querying condition
Condition is corresponded in the object clicked on, the second object in addition to first object；And determine the first object respective components
Component value is the number that first object is clicked under the querying condition；
The component value of the second object respective components is determined using equation below：
Wherein, Q be the querying condition, biFor i-th of second objects in all second objects, W (Q, bi) it is the inquiry item
B in the click vector of partiThe component value of respective components, Q (bi) it is corresponding click biQuerying condition set, Q (qi) it is to correspond to
Clicking on object includes the set of all querying conditions of at least one first object and at least one second object, qiFor Q (qi)
In i-th of querying condition, W (q, bi) it is biThe number being clicked under querying condition q, Sum (q) are querying condition q corresponding points
The number and value that each object hit is clicked under querying condition q, DecayFactor are decay factor, and A (Q) is inquiry item
Part Q corresponds to the set for the object clicked on, A (qi) it is querying condition qiThe set of the corresponding object clicked on, W (Q, ax) it is object ax
The number being clicked under querying condition Q, W (qi,ax) it is object axIn querying condition qiUnder the number that is clicked, Sum (ax) be
Object axRespectively object a is clicked on correspondingxEach querying condition under the number and value that is clicked.
Further, similarity determining unit 504, specifically for determine to meet preset condition in the following way two
Querying condition：
When there are during at least one same object, determine that this two are looked into two querying conditions each corresponding object clicked on
Inquiry condition meets preset condition.
Further, similarity determining unit 504, specifically for the component and the of the click vector when the first querying condition
When the component of the click vector of two querying conditions is identical, the click of first querying condition is determined using default similarity algorithm
The vectorial similarity between the click vector of second querying condition, and using result as first querying condition and institute
State the similarity between the second querying condition；
When the vectorial component of the click of the first querying condition, there are different from the vectorial component of the click of the second querying condition
When, determine the component of the component of the converting vector of first querying condition and the converting vector of second querying condition,
In, the component of the converting vector of first querying condition, including first querying condition click vector component and
Exist in the component of the click vector of second querying condition and in the vectorial component of the click of first querying condition
In the component that is not present, the component of the converting vector of first querying condition and the converting vector of second querying condition
Component is identical；
For each component of the converting vector of first querying condition, when the component is in first querying condition
In the presence of clicking in vector, the component value of the component of the converting vector of first querying condition is first querying condition
Click vector the component component value, otherwise, be 0；
For each component of the converting vector of second querying condition, when the component is in second querying condition
In the presence of clicking in vector, the component value of the component of the converting vector of second querying condition is second querying condition
Click vector the component component value, otherwise, be 0；
The converting vector of first querying condition and second querying condition are determined using default similarity algorithm
Similarity between converting vector, and using result as similar between first querying condition and second querying condition
Degree.
Based on same inventive concept, according to the Object Query method that the above embodiments of the present application provide, correspondingly, the application
Another embodiment additionally provides a kind of Object Query device, and structure diagram is as shown in fig. 6, specifically include：
Input unit 601, for obtaining the current queries condition of input；
Condition acquiring unit 602 reaches similar for obtaining the predetermined similarity to the current queries condition
Spend the similar querying condition of threshold value；
Query unit 603 obtains for meeting the object of the current queries condition in the full Integrated query of object and meeting
The object of the similar querying condition taken；
Condition determining unit 604, for determining to reach phase with the similarity of the current queries condition in the following way
Like the similar querying condition of degree threshold value：
Inquiry click logs are obtained, the inquiry click logs include the current queries condition and other each inquiry items
Part and the corresponding object clicked on；And each querying condition in the current queries condition and other each querying conditions is directed to, really
The fixed querying condition corresponds to the number that each object clicked on is clicked under the querying condition；And according to definite number, really
The click vector of fixed each querying condition, wherein, the component of vector is clicked on as object, and component value characterizes object in the querying condition
Under situation about being clicked；And it is treated really for each of preset condition is met in other each querying conditions with the current queries condition
Determine querying condition, according to the click vector for clicking on the vector sum querying condition to be determined of the current queries condition, determine institute
State the similarity between current queries condition and the querying condition to be determined；And from each querying condition to be determined, determine with
The similarity of the current queries condition reaches the querying condition of similarity threshold, is looked into as the similar of current queries condition
Inquiry condition.
The function of above-mentioned each unit may correspond to the respective handling step in flow shown in Fig. 1 to Fig. 4, no longer superfluous herein
It states.
In conclusion scheme provided by the embodiments of the present application, including：Inquiry click logs are obtained, are inquired about in click logs
Including each querying condition and the corresponding object clicked on；And for each querying condition in each querying condition, determine the inquiry item
Part corresponds to the number that each object clicked on is clicked under the querying condition；And according to definite number, determine each inquiry
The click vector of condition, wherein, the component of vector is clicked on as object, and component value characterization object is clicked under the querying condition
Situation；And for each two querying condition for meeting preset condition in each querying condition, according to wherein the first querying condition
The click vector of the second querying condition of vector sum is clicked on, determines the similarity between the first querying condition and the second querying condition.
Using scheme provided by the embodiments of the present application, the accuracy of the similarity between definite different querying conditions is improved.
The device that embodiments herein is provided can be realized by computer program.Those skilled in the art should be able to
Understand, above-mentioned module dividing mode is only one kind in numerous module dividing modes, if being divided into other modules or not drawing
Sub-module, if device have the function of it is above-mentioned, all should be within the protection domain of the application.
The application is with reference to method, the equipment according to the embodiment of the present application（System）And the flow of computer program product
Figure and/or block diagram describe.It should be understood that it can be realized by computer program instructions every first-class in flowchart and/or the block diagram
The combination of flow and/or box in journey and/or box and flowchart and/or the block diagram.These computer programs can be provided
The processor of all-purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices is instructed to produce
A raw machine so that the instruction performed by computer or the processor of other programmable data processing devices is generated for real
The device for the function of being specified in present one flow of flow chart or one box of multiple flows and/or block diagram or multiple boxes.
These computer program instructions, which may also be stored in, can guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works so that the instruction generation being stored in the computer-readable memory includes referring to
Make the manufacture of device, the command device realize in one flow of flow chart or multiple flows and/or one box of block diagram or
The function of being specified in multiple boxes.
These computer program instructions can be also loaded into computer or other programmable data processing devices so that counted
Series of operation steps is performed on calculation machine or other programmable devices to generate computer implemented processing, so as in computer or
The instruction offer performed on other programmable devices is used to implement in one flow of flow chart or multiple flows and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Obviously, those skilled in the art can carry out the application essence of the various modification and variations without departing from the application
God and scope.In this way, if these modifications and variations of the application belong to the scope of the application claim and its equivalent technologies
Within, then the application is also intended to comprising including these modification and variations.
Claims (11)
1. a kind of querying condition similarity determines method, which is characterized in that including：
Inquiry click logs are obtained, the inquiry click logs include each querying condition and the corresponding object clicked on；
For each querying condition in each querying condition, determine that the querying condition corresponds to each object clicked in the inquiry item
The number being clicked under part；
According to definite number, determine that the click of each querying condition is vectorial, wherein, the component for clicking on vector is the querying condition
Corresponding the first object clicked on and the corresponding other querying conditions for clicking on first object are corresponded in the object clicked on, and are removed
The second object outside first object, the component values of the first object respective components are first object under the querying condition
The number being clicked, the component value of the second object respective components are included at least in corresponding object of clicking on based on second object
Number of clicks under the querying condition of one the first object and at least one second object and in corresponding second object of clicking on
What the number of clicks of querying condition obtained；
For in querying condition, each in the corresponding object clicked on there are each two querying condition of at least one same object,
According to wherein the first querying condition click the second querying condition of vector sum click vector, determine first querying condition with
Similarity between second querying condition.
2. the method as described in claim 1, which is characterized in that according to definite number, determine the click of each querying condition
Vector specifically includes：
For each querying condition, determine that the component of the click vector of the querying condition corresponds to click on first for the querying condition
Object；
Determine the number that the component value of component is clicked for the component the first object of correspondence under the querying condition.
3. the method as described in claim 1, which is characterized in that according to definite number, determine the click of each querying condition
Vector specifically includes：
For each querying condition, determine that the component of the click vector of the querying condition corresponds to click on first for the querying condition
Object and the corresponding other querying conditions for clicking on first object are corresponded in the object clicked on, except first object it
The second outer object；
The component value for determining the first object respective components is the number that first object is clicked under the querying condition；
The component value of the second object respective components is determined using equation below：
<mrow>
<mi>W</mi>
<mrow>
<mo>(</mo>
<mi>Q</mi>
<mo>,</mo>
<msub>
<mi>b</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<munderover>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>q</mi>
<mo>&Element;</mo>
<mi>Q</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>b</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</munder>
<mrow>
<mi>q</mi>
<mo>&Element;</mo>
<mi>Q</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>q</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<mi>q</mi>
</munderover>
<mi>N</mi>
<mi>o</mi>
<mi>d</mi>
<mo>(</mo>
<mi>q</mi>
<mo>)</mo>
<mfrac>
<mrow>
<mi>W</mi>
<mrow>
<mo>(</mo>
<mi>q</mi>
<mo>,</mo>
<msub>
<mi>b</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>S</mi>
<mi>u</mi>
<mi>m</mi>
<mrow>
<mo>(</mo>
<mi>q</mi>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mo>*</mo>
<mi>D</mi>
<mi>e</mi>
<mi>c</mi>
<mi>a</mi>
<mi>y</mi>
<mi>F</mi>
<mi>a</mi>
<mi>c</mi>
<mi>t</mi>
<mi>o</mi>
<mi> </mi>
<mi>r</mi>
<mo>;</mo>
</mrow>
<mrow>
<mi>N</mi>
<mi>o</mi>
<mi>d</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>q</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munderover>
<munder>
<mo>&Sigma;</mo>
<mrow>
<msub>
<mi>a</mi>
<mi>x</mi>
</msub>
<mo>&Element;</mo>
<mi>A</mi>
<mrow>
<mo>(</mo>
<mi>Q</mi>
<mo>)</mo>
</mrow>
</mrow>
</munder>
<mrow>
<msub>
<mi>a</mi>
<mi>x</mi>
</msub>
<mo>&Element;</mo>
<mi>A</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>q</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<msub>
<mi>a</mi>
<mi>x</mi>
</msub>
</munderover>
<mi>W</mi>
<mrow>
<mo>(</mo>
<mi>Q</mi>
<mo>,</mo>
<msub>
<mi>a</mi>
<mi>x</mi>
</msub>
<mo>)</mo>
</mrow>
<mfrac>
<mrow>
<mi>W</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>q</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<msub>
<mi>a</mi>
<mi>x</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>S</mi>
<mi>u</mi>
<mi>m</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>a</mi>
<mi>m</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mo>;</mo>
</mrow>
Wherein, Q be the querying condition, biFor i-th of second objects in all second objects, W (Q, bi) it is the querying condition
Click on b in vectoriThe component value of respective components, Q (bi) it is corresponding click biQuerying condition set, Q (qi) clicked on to be corresponding
Object includes the set of all querying conditions of at least one first object and at least one second object, qiFor Q (qi) in
I-th of querying condition, W (q, bi) it is biThe number being clicked under querying condition q, Sum (q) are that querying condition q corresponds to what is clicked on
The number and value that each object is clicked under querying condition q, DecayFacto r are decay factor, and A (Q) is querying condition Q
The set of the corresponding object clicked on, A (qi) it is querying condition qiThe set of the corresponding object clicked on, W (Q, ax) it is object axIt is looking into
The number being clicked under inquiry condition Q, W (qi,ax) it is object axIn querying condition qiUnder the number that is clicked, Sum (ax) it is object
axRespectively object a is clicked on correspondingxEach querying condition under the number and value that is clicked.
4. the method as described in claim 1, which is characterized in that looked into according to the click vector sum second of wherein the first querying condition
The click vector of inquiry condition, determines the similarity between first querying condition and second querying condition, specifically includes：
When the component of the click vector of the first querying condition is identical with the component of the click vector of the second querying condition, using pre-
If similarity algorithm is determined between the click vector of first querying condition and the click vector of second querying condition
Similarity, and using result as the similarity between first querying condition and second querying condition；
When the click vector of the first querying condition component and the second querying condition click on vectorial component there are it is different when, really
The component of the component of the converting vector of fixed first querying condition and the converting vector of second querying condition, wherein, institute
The component of the converting vector of the first querying condition is stated, vectorial component is clicked on and in institute including first querying condition
State the second querying condition click vector component in exist and first querying condition click vector component in not
Existing component, the component of the component of the converting vector of first querying condition and the converting vector of second querying condition
It is identical；
For each component of the converting vector of first querying condition, when the component is in the click of first querying condition
In the presence of in vector, the component value of the component of the converting vector of first querying condition is the point of first querying condition
Otherwise the component value of the component of the amount of hitting to, is 0；
For each component of the converting vector of second querying condition, when the component is in the click of second querying condition
In the presence of in vector, the component value of the component of the converting vector of second querying condition is the point of second querying condition
Otherwise the component value of the component of the amount of hitting to, is 0；
The converting vector of first querying condition and the conversion of second querying condition are determined using default similarity algorithm
Similarity between vector, and using result as the similarity between first querying condition and second querying condition.
A kind of 5. Object Query method, which is characterized in that including：
Obtain the current queries condition of input；
Obtain the similar querying condition that the predetermined similarity to the current queries condition reaches similarity threshold；
Meet the object of the current queries condition in the full Integrated query of object and meet the similar querying condition obtained
Object；
Wherein, the similar inquiry item for reaching similarity threshold to the similarity of the current queries condition is determined in the following way
Part：
Obtain inquiry click logs, it is described inquiry click logs include the current queries condition and other each querying conditions and
The corresponding object clicked on；
For each querying condition in the current queries condition and other each querying conditions, the querying condition corresponding points are determined
The number that each object hit is clicked under the querying condition；
According to definite number, determine that the click of each querying condition is vectorial, wherein, the component for clicking on vector is the querying condition
Corresponding the first object clicked on and the corresponding other querying conditions for clicking on first object are corresponded in the object clicked on, and are removed
The second object outside first object, the component values of the first object respective components are first object under the querying condition
The number being clicked, the component value of the second object respective components are included at least in corresponding object of clicking on based on second object
Number of clicks under the querying condition of one the first object and at least one second object and in corresponding second object of clicking on
What the number of clicks of querying condition obtained；
For in other each querying conditions, and there are at least one phases in the current queries the condition each object of corresponding click
With each querying condition to be determined of object, the vector sum querying condition to be determined is clicked on according to the current queries condition
Vector is clicked on, determines the similarity between the current queries condition and the querying condition to be determined；
From each querying condition to be determined, the inquiry item for reaching similarity threshold with the similarity of the current queries condition is determined
Part, the similar querying condition as the current queries condition.
6. a kind of querying condition similarity determining device, which is characterized in that including：
Log acquisition unit inquires about click logs for obtaining, and the inquiry click logs include each querying condition and correspondence
The object of click；
Number determination unit for being directed to each querying condition in each querying condition, determines that the querying condition corresponds to what is clicked on
The number that each object is clicked under the querying condition；
Vector determination unit for the definite number of basis, determines that the click of each querying condition is vectorial, wherein, click on vector
Component the first object clicked on and the corresponding other querying conditions pair for clicking on first object are corresponded to for the querying condition
In the object that should be clicked on, the second object in addition to first object, the component values of the first object respective components for this first
The number that object is clicked under the querying condition, the component value of the second object respective components is in correspondence based on second object
Click object includes the number of clicks under the querying condition of at least one first object and at least one second object, Yi Ji
What the number of clicks of the corresponding querying condition for clicking on the second object obtained；
Similarity determining unit, for being directed in querying condition, each there are at least one identical right in the corresponding object clicked on
The each two querying condition of elephant, it is vectorial according to the click of click the second querying condition of vector sum of wherein the first querying condition, really
Fixed similarity between first querying condition and second querying condition.
7. device as claimed in claim 6, which is characterized in that the vector determination unit, specifically for being directed to each inquiry
Condition determines that the component of the click vector of the querying condition corresponds to the first object clicked on for the querying condition；And determine component
Component value correspond to the number that is clicked under the querying condition of the first object for the component.
8. device as claimed in claim 7, which is characterized in that the vector determination unit, specifically for being directed to each inquiry
Condition determines that the component of the click vector of the querying condition corresponds to the first object clicked on and corresponding points for the querying condition
The other querying conditions for hitting first object are corresponded in the object clicked on, the second object in addition to first object；With
And determine the number that the component value of the first object respective components is clicked for first object under the querying condition；
The component value of the second object respective components is determined using equation below：
<mrow>
<mi>W</mi>
<mrow>
<mo>(</mo>
<mi>Q</mi>
<mo>,</mo>
<msub>
<mi>b</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<munderover>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>q</mi>
<mo>&Element;</mo>
<mi>Q</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>b</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</munder>
<mrow>
<mi>q</mi>
<mo>&Element;</mo>
<mi>Q</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>q</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<mi>q</mi>
</munderover>
<mi>N</mi>
<mi>o</mi>
<mi>d</mi>
<mo>(</mo>
<mi>q</mi>
<mo>)</mo>
<mfrac>
<mrow>
<mi>W</mi>
<mrow>
<mo>(</mo>
<mi>q</mi>
<mo>,</mo>
<msub>
<mi>b</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>S</mi>
<mi>u</mi>
<mi>m</mi>
<mrow>
<mo>(</mo>
<mi>q</mi>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mo>*</mo>
<mi>D</mi>
<mi>e</mi>
<mi>c</mi>
<mi>a</mi>
<mi>y</mi>
<mi>F</mi>
<mi>a</mi>
<mi>c</mi>
<mi>t</mi>
<mi>o</mi>
<mi> </mi>
<mi>r</mi>
<mo>;</mo>
</mrow>
<mrow>
<mi>N</mi>
<mi>o</mi>
<mi>d</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>q</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munderover>
<munder>
<mo>&Sigma;</mo>
<mrow>
<msub>
<mi>a</mi>
<mi>x</mi>
</msub>
<mo>&Element;</mo>
<mi>A</mi>
<mrow>
<mo>(</mo>
<mi>Q</mi>
<mo>)</mo>
</mrow>
</mrow>
</munder>
<mrow>
<msub>
<mi>a</mi>
<mi>x</mi>
</msub>
<mo>&Element;</mo>
<mi>A</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>q</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<msub>
<mi>a</mi>
<mi>x</mi>
</msub>
</munderover>
<mi>W</mi>
<mrow>
<mo>(</mo>
<mi>Q</mi>
<mo>,</mo>
<msub>
<mi>a</mi>
<mi>x</mi>
</msub>
<mo>)</mo>
</mrow>
<mfrac>
<mrow>
<mi>W</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>q</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<msub>
<mi>a</mi>
<mi>x</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>S</mi>
<mi>u</mi>
<mi>m</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>a</mi>
<mi>m</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mo>;</mo>
</mrow>
Wherein, Q be the querying condition, biFor i-th of second objects in all second objects, W (Q, bi) it is the querying condition
Click on b in vectoriThe component value of respective components, Q (bi) it is corresponding click biQuerying condition set, Q (qi) clicked on to be corresponding
Object includes the set of all querying conditions of at least one first object and at least one second object, qiFor Q (qi) in
I-th of querying condition, W (q, bi) it is biThe number being clicked under querying condition q, Sum (q) are that querying condition q corresponds to what is clicked on
The number and value that each object is clicked under querying condition q, DecayFacto r are decay factor, and A (Q) is querying condition Q
The set of the corresponding object clicked on, A (qi) it is querying condition qiThe set of the corresponding object clicked on, W (Q, ax) it is object axIt is looking into
The number being clicked under inquiry condition Q, W (qi,ax) it is object axIn querying condition qiUnder the number that is clicked, Sum (ax) it is object
axRespectively object a is clicked on correspondingxEach querying condition under the number and value that is clicked.
9. device as claimed in claim 7, which is characterized in that the similarity determining unit, specifically for using such as lower section
Formula determines two querying conditions for meeting preset condition：
When there are during at least one same object, determine this two inquiry items in two querying conditions each corresponding object clicked on
Part meets preset condition.
10. device as claimed in claim 7, which is characterized in that the similarity determining unit, specifically for when the first inquiry
It is true using default similarity algorithm when the component of the click vector of condition is identical with the component of the click vector of the second querying condition
Similarity between the click vector of fixed first querying condition and the click vector of second querying condition, and by result
As the similarity between first querying condition and second querying condition；
When the click vector of the first querying condition component and the second querying condition click on vectorial component there are it is different when, really
The component of the component of the converting vector of fixed first querying condition and the converting vector of second querying condition, wherein, institute
The component of the converting vector of the first querying condition is stated, vectorial component is clicked on and in institute including first querying condition
State the second querying condition click vector component in exist and first querying condition click vector component in not
Existing component, the component of the component of the converting vector of first querying condition and the converting vector of second querying condition
It is identical；
For each component of the converting vector of first querying condition, when the component is in the click of first querying condition
In the presence of in vector, the component value of the component of the converting vector of first querying condition is the point of first querying condition
Otherwise the component value of the component of the amount of hitting to, is 0；
For each component of the converting vector of second querying condition, when the component is in the click of second querying condition
In the presence of in vector, the component value of the component of the converting vector of second querying condition is the point of second querying condition
Otherwise the component value of the component of the amount of hitting to, is 0；
The converting vector of first querying condition and the conversion of second querying condition are determined using default similarity algorithm
Similarity between vector, and using result as the similarity between first querying condition and second querying condition.
11. a kind of Object Query device, which is characterized in that including：
Input unit, for obtaining the current queries condition of input；
Condition acquiring unit reaches similarity threshold for obtaining predetermined and current queries condition similarity
Similar querying condition；
Query unit, for meeting the object of the current queries condition in the full Integrated query of object and meeting the institute obtained
State the object of similar querying condition；
Condition determining unit, for determining to reach similarity threshold with the similarity of the current queries condition in the following way
Similar querying condition：
Obtain inquiry click logs, it is described inquiry click logs include the current queries condition and other each querying conditions and
The corresponding object clicked on；And for each querying condition in the current queries condition and other each querying conditions, determining should
Querying condition corresponds to the number that each object clicked on is clicked under the querying condition；And it according to definite number, determines every
The click vector of a querying condition, wherein, it clicks on the component of vector and corresponds to the first object and right clicked on for the querying condition
Other querying conditions that first object should be clicked on are corresponded in the object clicked on, the second couple in addition to first object
As the component value of, the first object respective components is the number that first object is clicked under the querying condition, the second object pair
The component value for answering component is to include at least one first object and at least one in corresponding object of clicking on based on second object
Number of clicks under the querying condition of second object and the number of clicks acquisition in the corresponding querying condition for clicking on the second object
's；
And in other each querying conditions, and there are at least one in the current queries the condition each object of corresponding click
Each querying condition to be determined of same object, according to the click vector sum of the current queries condition querying condition to be determined
Click vector, determine the similarity between the current queries condition and the querying condition to be determined；And from each to be determined
In querying condition, the querying condition for reaching similarity threshold with the similarity of the current queries condition is determined, work as described
The similar querying condition of preceding querying condition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210421325.8A CN103793390B (en) | 2012-10-29 | 2012-10-29 | Querying condition similarity determines method, Object Query method and relevant apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210421325.8A CN103793390B (en) | 2012-10-29 | 2012-10-29 | Querying condition similarity determines method, Object Query method and relevant apparatus |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103793390A CN103793390A (en) | 2014-05-14 |
| CN103793390B true CN103793390B (en) | 2018-05-29 |
Family
ID=50669073
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210421325.8A Active CN103793390B (en) | 2012-10-29 | 2012-10-29 | Querying condition similarity determines method, Object Query method and relevant apparatus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103793390B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106682022A (en) * | 2015-11-10 | 2017-05-17 | 中国电信股份有限公司 | Log query method and device |
| CN107609152B (en) * | 2017-09-22 | 2021-03-09 | 百度在线网络技术（北京）有限公司 | Method and apparatus for expanding query expressions |
| CN108269122B (en) * | 2017-12-29 | 2021-08-06 | 阿里巴巴（中国）有限公司 | Advertisement similarity processing method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101206674A (en) * | 2007-12-25 | 2008-06-25 | 北京科文书业信息技术有限公司 | Enhancement type related search system and method using commercial articles as medium |
| CN101241512A (en) * | 2008-03-10 | 2008-08-13 | 北京搜狗科技发展有限公司 | Search method for redefining enquiry word and device therefor |
| CN101685448A (en) * | 2008-09-28 | 2010-03-31 | 国际商业机器公司 | Method and device for establishing association between query operation of user and search result |
-
2012
- 2012-10-29 CN CN201210421325.8A patent/CN103793390B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101206674A (en) * | 2007-12-25 | 2008-06-25 | 北京科文书业信息技术有限公司 | Enhancement type related search system and method using commercial articles as medium |
| CN101241512A (en) * | 2008-03-10 | 2008-08-13 | 北京搜狗科技发展有限公司 | Search method for redefining enquiry word and device therefor |
| CN101685448A (en) * | 2008-09-28 | 2010-03-31 | 国际商业机器公司 | Method and device for establishing association between query operation of user and search result |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103793390A (en) | 2014-05-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP5513624B2 (en) | Retrieving information based on general query attributes | |
| US8346765B2 (en) | Generating ranked search results using linear and nonlinear ranking models | |
| CN104504003B (en) | The searching method and device of diagram data | |
| Rekabsaz et al. | Exploration of a threshold for similarity based on uncertainty in word embedding | |
| CN104391978B (en) | Web page storage processing method and processing device for browser | |
| CN103077254B (en) | Webpage acquisition methods and device | |
| EP2430575A1 (en) | Search method, apparatus and system | |
| CN109189990A (en) | A kind of generation method of search term, device and electronic equipment | |
| CN106503175A (en) | The inquiry of Similar Text, problem extended method, device and robot | |
| US20130339369A1 (en) | Search Method and Apparatus | |
| CN105528422A (en) | Focused crawler processing method and apparatus | |
| CN106776559B (en) | Text semantic similarity calculation method and device | |
| WO2011011063A2 (en) | Method and system for document indexing and data querying | |
| CN103793390B (en) | Querying condition similarity determines method, Object Query method and relevant apparatus | |
| CN105930513A (en) | Browser history record sorting method and apparatus | |
| CN104123285B (en) | The air navigation aid and device of search result | |
| CN104102662A (en) | Method and device for determining interest and preference similarity of users | |
| CN103605744B (en) | The analysis method and device of site search engine data on flows | |
| CN107153697A (en) | Product search method and device in a kind of commodity transaction website | |
| CN108549631A (en) | Noun dictionary extracting method, electronic device and computer readable storage medium | |
| CN108388556A (en) | The method for digging and system of similar entity | |
| CN104765747A (en) | Webpage processing method and device | |
| CN103235783B (en) | A kind of method and apparatus for determining optimal search result | |
| WO2016176932A1 (en) | Method for searching for keyword on information display page | |
| Sahoo et al. | On the study of GRBF and polynomial kernel based support vector machine in web logs |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |