CN109446402A - A kind of searching method and device - Google Patents
A kind of searching method and device Download PDFInfo
- Publication number
- CN109446402A CN109446402A CN201710757313.5A CN201710757313A CN109446402A CN 109446402 A CN109446402 A CN 109446402A CN 201710757313 A CN201710757313 A CN 201710757313A CN 109446402 A CN109446402 A CN 109446402A
- Authority
- CN
- China
- Prior art keywords
- user
- historical behavior
- class
- ranking score
- search result
- 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.)
- Granted
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
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/3332—Query translation
- G06F16/3334—Selection or weighting of terms from queries, including natural language queries
-
- 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
- G06F16/955—Retrieval from the web using information identifiers, e.g. uniform resource locators [URL]
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
This application provides a kind of searching method and devices, during determining search result according to search key, from the extension object with incidence relation of object that historical behavior occurs with user, determine object relevant to search key, as a part of search result, therefore, search result can be closer to the behavioural habits of user, for a user, more accurate.
Description
Technical field
This application involves electronic information field more particularly to a kind of searching methods and device.
Background technique
Search engine is the most common function in website.After user inputs keyword in a search engine, search engine foundation
Keyword query is ranked up display to relevant search result, and to search result.For example, the search of e-commerce website is drawn
After holding up the keyword for receiving user's input, merchandise news relevant to keyword is inquired, and be ranked up to merchandise news,
Each merchandise news is shown to user according still further to ranking results.
However existing searching method, search result is exported only in accordance with keyword, without considering other factors, institute
To be unable to get more accurate search result for user oriented.
Summary of the invention
This application provides a kind of searching method and devices, it is therefore intended that it is more accurate how solution obtains user oriented
The problem of search result.
To achieve the goals above, this application provides following technical schemes:
A kind of searching method, comprising:
Search key according to user determines that the first class object, first class object are and described search keyword phase
The object of pass；
Based on the historical behavior of the user, the historical behavior object of the user is determined；
The determining extension object with the historical behavior object with incidence relation；
Determine the second class object relevant to the keyword in the extension object；
Integrated ordered to search result progress, described search result includes first class object and second class pair
As.
Optionally, the historical behavior based on the user determines that the historical behavior object of the user includes:
The historical behavior object of the user is obtained from the historical behavior data of the user；
There is the extension object of incidence relation to include: for the determination and the historical behavior object
Calculate behavioral similarity sim (i, j) between any one object i and the seed object j of any one user and/
Or sim (i, j；S, t, p), wherein sim (i, j) indicates that the user has the sum of the number of behavior, sim (i, j to i and j simultaneously；s,
T, p) indicate that the user has the sum of the number of behavior p to i and j simultaneously under s scene in t time range；
The extension object is obtained according to the similitude between each object and the historical behavior object of each user, it is described
Similitude includes at least the behavioral similarity.
Optionally, it is described to search result carry out it is integrated ordered before, further includes:
If the quantity of second class object is less than preset value, increase first class object in described search result
In accounting.
Optionally, it is described search result is carried out integrated ordered include:
The ranking score of described search result is calculated, second class object has similar ranking score and conventional sequence point
Number, described to have the conventional ranking score with first class object, the similar ranking score and conventional ranking score are not
Together.
Optionally, phase of the similar ranking score based on second class object with the historical behavior object of the user
Determine that the seed weight is according to classification belonging to second class object, user to described second like degree and seed weight
The time that the behavior type of class object and behavior occur determines.
Optionally, the similar ranking score is the product of the similarity and the seed weight.
Optionally, the similar ranking score be based further on the user historical behavior object and second class pair
The price difference of elephant.
Optionally, it is described to search result carry out it is integrated ordered after, further includes:
Described search result is shown according to the ranking score.
A kind of searcher, comprising:
First determining module determines that the first class object, first class object are for the search key according to user
Object relevant to described search keyword；
Second determining module determines the historical behavior object of the user for the historical behavior based on the user；
Third determining module, for the determining extension object with the historical behavior object with incidence relation；
4th determining module, for determining the second class object relevant to the keyword in the extension object；
Sorting module, for search result carry out it is integrated ordered, described search result include first class object and
Second class object.
Optionally, second determining module is specifically used for:
The historical behavior object of the user is obtained from the historical behavior data of the user；
The third determining module is specifically used for:
Calculate behavioral similarity sim (i, j) between any one object i and the seed object j of any one user and/
Or sim (i, j；S, t, p), wherein sim (i, j) indicates that the user has the sum of the number of behavior, sim (i, j to i and j simultaneously；s,
T, p) indicate that the user has the sum of the number of behavior p to i and j simultaneously under s scene in t time range；
The extension object is obtained according to the similitude between each object and the historical behavior object of each user, it is described
Similitude includes at least the behavioral similarity.
Optionally, further includes:
Control module is used for before the sorting module is integrated ordered to search result progress, if second class
The quantity of object is less than preset value, then increases accounting of first class object in described search result.
Optionally, the sorting module is specifically used for:
The ranking score of described search result is calculated, second class object has similar ranking score and conventional sequence point
Number, described to have the conventional ranking score with first class object, the similar ranking score and conventional ranking score are not
Together.
Optionally, phase of the similar ranking score based on second class object with the historical behavior object of the user
Determine that the seed weight is according to classification belonging to second class object, user to described second like degree and seed weight
The time that the behavior type of class object and behavior occur determines.
Optionally, the product of similarity and the seed weight described in the similar ranking score.
Optionally, the similar ranking score be based further on the user historical behavior object and second class pair
The price difference of elephant.
Optionally, further includes:
Display module, for showing described search result according to the ranking score.
A kind of searching method, comprising:
Historical behavior based on user determines the historical behavior object of the user；
The determining extension object with the historical behavior object with incidence relation；
Determine result object relevant to search key in the extension object；
The result object is ranked up.
Searching method and device described herein, according to search key determine search result during, from
User occurs in the extension object with incidence relation of the object of historical behavior, determines object relevant to search key,
As a part of search result, therefore, search result can be closer to the behavioural habits of user, for a user, more
Accurately.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of application for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is a kind of flow chart of searching method disclosed in the embodiment of the present application；
Fig. 2 (a) -2 (c) is the page display effect comparison of searching method disclosed in the embodiment of the present application and the prior art
Figure；
Fig. 3 is the flow chart that the method for analogical object model is established disclosed in the embodiment of the present application；
Fig. 4 is the flow chart of another searching method disclosed in the embodiment of the present application；
Fig. 5 is the structural schematic diagram of searcher disclosed in the embodiment of the present application.
Specific embodiment
Searching method disclosed in the embodiment of the present application can apply the server in website (such as e-commerce website)
On.The server is for running website, and after the search engine of website receives search key, server is not only in accordance with pass
Keyword provides search result, and the also historical behavior information according to the user of input keyword provides search result, searches to improve
Accuracy of the hitch fruit towards the user.
Below in conjunction with the attached drawing in the embodiment of the present application, technical solutions in the embodiments of the present application carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of embodiments of the present application, instead of all the embodiments.It is based on
Embodiment in the application, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall in the protection scope of this application.
Fig. 1 is a kind of searching method disclosed in the embodiment of the present application, comprising the following steps:
S101: receiving the search key of user's input, and search for object relevant to search key, and referred to as first
Class object.
By taking e-commerce website as an example, user inputs search key " sport footwear " in the search engine of website, then takes
Business device provides the information of relevant commodity " sport footwear " according to search key.Relevant object is obtained according to search key
Mode may refer to the prior art, and which is not described herein again.
S102: the historical behavior based on the user determines the historical behavior object of the user.
Wherein, the historical behavior object of user is the object that historical behavior occurred for user.
By taking e-commerce website as an example, in general, e-commerce website can identify user's according to the registration information of user
Identity information identifies that the user of input keyword is user A.
The history that the commodity of historical behavior (behavior includes but is not limited to collect, click, buying) are user A occurs for user A
Behavior commodity (object).
S103: the determining extension object with the historical behavior object of user with incidence relation.
The historical behavior object for extending object and user has an incidence relation, incidence relation can for similarity relation or
With same alike result or user's history behavior having the same etc..Example is connected, belongs to phase with the historical behavior commodity of user A
Commodity with brand are associated articles.
S104: the second class object relevant to search key in extension object is determined.
S105: integrated ordered to search result progress, search result includes the first class object and the second class object.
Specifically, can respectively from the first class object and the second class object selecting object as search result, wherein search
In hitch fruit, the quantity of the first class object and the second class object meets certain accounting.
In the present embodiment, accounting can be preset fixed value, can also be according to the first class object and the second class pair
The quantity of elephant adjusts.Specifically, increasing the accounting of the first class object if the quantity of the second class object is less than preset value.Example
Such as, for a kind of new commodity, the historical behavior commodity of user may be not present, the second class object is also just not present, in this feelings
Under condition, the quantity of the second class object is zero, then the accounting of the first class object is adjusted to 1.Conversely, in another extreme case
Under, it can be only using the second class object as object to be presented, in this case, it is possible to skip S101.
It is integrated ordered to refer to, the first class object and the second class object are ranked up as a whole, rather than by the first kind
Object order again sorts the second class object.That is, the first class object and the second class object are ranked up together.
Below by taking Fig. 2 (a)-Fig. 2 (c) as an example, to searching method shown in FIG. 1 and the effect after the searching method is used
Fruit is illustrated:
User clicks eye cream shown in Fig. 2 (a) list top in the items list shown in Fig. 2 (a), so, server
Eye cream shown in backstage record Fig. 2 (a) is the historical behavior commodity of user, and in all commodity in website, inquiry and Fig. 2
(a) the similar commodity of eye cream shown in record after the completion of inquiry on backstage.
Fig. 2 (b) is after user inputs " eye cream " in a search engine, and server searches for obtained commodity according to " eye cream ".
Fig. 2 (b) is also the search result that existing searching method shows user.
Fig. 2 (c) is the search result that searching method shown in FIG. 1 obtains, the presentation process of search result are as follows: server base
In keyword " eye cream ", search obtains commodity shown in Fig. 2 (b) from the entire service of website, also, is based on keyword " eye
Frost " is searched for from the commodity similar to eye cream shown in Fig. 2 (a) of backstage record, obtains search result.It is obtained again from twice
Search result in respectively select a part of commodity, formed final search result to user show.
It include being selected from the similar commodity of eye cream shown in Fig. 2 (a) as shown in Fig. 2 (c), in final search result
A part of commodity (the second class object, the commodity that the arrow of eye cream shown in Fig. 2 (a) is directed toward), and searched for from Fig. 2 (b)
To commodity in a part of commodity (the first class object, the commodity that the arrow of eye cream shown in Fig. 2 (b) is directed toward) for selecting, the
The quantity of one class object and the second class object is at certain accounting.
It can be seen that in search result other than the object searched out according to search key from process shown in FIG. 1, also
Including extension object similar with user's generation object of historical behavior, therefore, search result can be closer to the behavior of user
Habit, it is for a user, more accurate.
Specifically, the specific implementation process of S102 is as shown in figure 3, include following so that incidence relation is similarity relation as an example
Step:
S301: the historical behavior object of each user is obtained.
As previously mentioned, the seed object of a user refers to that the object of historical behavior occurs for the user.Server can be first
The historical behavior data of each user are obtained from the history data of website, it optionally, can be to the history of each user
Behavioral data is filtered, then the historical behavior object of each user is filtered out from filtered data.
The historical behavior data of user indicate the historical behavior that user generates object.Such as user's collection, click or purchase
Buy commodity.That is, a historical behavior data include user, object and behavioural information.
From the aspect of user, object and behavioural information three, by taking e-commerce website as an example, specific filter type packet
Include but be not limited to it is following any one:
1, the historical behavior data for belonging to the user of blacklist are filtered out, to prevent hacker from obtaining seed by cheating
Commodity.
2, the same user multiple behavior of the user to the same object (in such as one day) within a preset period of time.
3, filter out time of the act less than the first preset time value (such as 1 second) and or be greater than the second time value (such as
360 seconds) user's history behavioral data.
For example, user browsed the time of the details page of certain commodity less than 1 second, then it is considered as invalid clicks or user clicks
Lose interest in completely afterwards, therefore, such historical behavior data are considered as noise.Alternatively, the residence time after being clicked to user
Greater than 360 seconds, it may be possible to which user leaves invalid browsing time caused by being not turned off the page, therefore, such historical behavior number
According to being also considered as noise.
4, the historical behavior data that user generates oneself object are filtered out, for example, the commodity that user clicks are for oneself
Commodity then need to filter out this user's history behavioral data.
5, the data that behavior number is more than default value are filtered out.For example, to filter out the commodity that hits are more than 10000
User's history behavioral data.The reason is that, the similarity of this commodity and most of commodity is all very high, therefore will affect other
Commodity enter analogical object library.
S302: the synthesis phase on website between each object being currently included and the historical behavior object of each user is calculated
Like degree.
Shown in behavior similarity such as formula (1):
if au,i！=0&&au,j！=0, co_action=1else=0
Wherein, whether sim (i, j) indicates the behavior similarity of object i and j, have simultaneously to object (i, j) to all users
The number of behavior is summed.au,iIndicate whether user u has behavior to object i, being is 1, otherwise is 0；co_action(au,i,au,j)
Indicate whether user u has behavior simultaneously to commodity i and commodity j, being is 1, otherwise is 0.
Further, in practice, as clicked and purchase, cost that user pays is different for different behaviors, thus data can
Reliability and importance are also different.Different scenes such as the behavioral data under recommending and searching for, it may have difference.Joint act occurs
In time range how long, also there is different influences for the identification of similarity, for example, clicking jointly before the same day and one month
A possibility that commodity crossed relevant property, is smaller.Taking into account the above factors, in the present embodiment, partitive behavior type, row
For the time, behavior scene carries out similarity calculation, such as formula (2):
if uu,i；s,t,p！=0&&uu,j；s,t,p！=0, co_action=1else=0
uu,i；s,t,pIndicate whether the t time range under s scene has p behavior to user u to object i；co_action
(uu,i；s,t,p,uu,j；s,t,p) indicate whether user u has p behavior type simultaneously in t time range under s scene to object i and j,
Being is 1, otherwise is 0.sim(i,j；S, t, p) indicate time of the object i and j under s scene in t time range under p kind behavior type
The sum of number.
By the common two kind behavior type on e-commerce website: clicking and (be by collecting commodities merger for buying
Click, add purchase merger for purchase), available click-click, click-purchase, 3 kinds of behavior combinations of purchase-purchase.Again with complete
For network data and search contextual data and 1 day and 3 days two kinds of time ranges, according to formula (2), object i and j total can be obtained
To 3x2x2 kind similarity.
In the present embodiment, similarity shown in formula (1) and formula (2) is referred to as to the behavior similarity of object.Actually answering
In, formula (1) can be used and/or formula (2) obtains the behavior similarity of object.
It, can be with the content similarity of computing object other than the behavior similarity of object.The content similarity master of object
It to include the similarity of the image and/or text between object.The calculation of the content similarity of object may refer to existing skill
Art, which is not described herein again.
Comprehensive similarity can be obtained with Behavior-based control similarity and content similarity, it is each as what is be currently included on website
Similarity between a object and the historical behavior object of each user.
S303: according to comprehensive similarity obtained above, extension object is determined.
Specifically, the object that similarity can be met to threshold value is determined as extending object.
It can be seen that search every time from the step of Fig. 3 to be both needed to execute S301-S303, in order to mitigate the line in search process
Upper calculating pressure can optionally be predicted according to the principle training pattern of S301-S303 using trained model offline
The extension object of each historical behavior object, specific:
Using the comprehensive similarity of above-mentioned acquisition as the input of Logic Regression Models, the training logistic regression under different scenes
Model obtains analogical object model.
Specifically, by taking e-commerce website as an example, using the similarity input logic regression model of above-mentioned acquisition as feature,
The feature that addition commodity popularity point etc. represents commercial quality is trained together.To obtain with the consistent set of metadata of similar data of scene demand, use
Sample under search is trained, i.e., in search, the similar quotient released according to the historical behavior commodity of user is showed to user
Product if user clicks or purchase is positive sample, otherwise are negative sample.Use positive sample and negative sample training Logic Regression Models.
Optionally, the quality of existing tools assessment model can be used.
It should be noted that training process can carry out before search, after model training is good, can predict offline every
The analogical object of a historical behavior object calculates pressure on the line in search process to mitigate.
Fig. 4 is another searching method disclosed in the embodiment of the present application, compared with method shown in FIG. 1, in Fig. 4, respectively
It gives a mark to the first class object and the second class object, to obtain more accurate searching order.
In Fig. 4 the following steps are included:
S401: the search key of user's input is received.
S402: search object relevant to search key, referred to as the first class object.
S403: it obtains in the extension object that there is incidence relation with the historical behavior object of user, with search key phase
The analogical object of pass, referred to as the second class object.
Specifically, extension object can be obtained according to pre-set analogical object model.
S404: object to be presented is selected from the first class object and the second class object respectively, wherein object to be presented
In, the quantity of the first class object and the second class object meets certain accounting.
S405: the similar ranking score of the second class object in object to be presented is calculated.
In the present embodiment, the similar ranking score of the second class object is calculated according to formula (3):
Score=sseed(cate,type,time)*ssim (3)
Wherein, Score indicates ranking score, SseedIt is object classification cate, behavior type type and time of the act time
Corresponding seed weight (can be that corresponding seed power is arranged in different object classifications, different behavior types and time of the act in advance
Weight, for example, the seed weight of women's dress, buying behavior in one month is 1, women's dress, collection behavior in one month seed power
Weight is 0.5) SsimIt is the good similarity of off-line calculation, i.e. comprehensive similarity obtained in S302.
Specifically, SseedCalculation are as follows: behavior type and user couple with different object classification, user to object
The time of the act of object learns inhomogeneity different behavior type difference behaviors now out as feature, training Logic Regression Models
The importance of time, i.e., S hereseed。
In the present embodiment, it is contemplated that the similarity between seed weight and object.Different behavior types, the object of time of the act
With different importance.Seed weight can be according to user to the behavior type (such as clicking, purchase) of the object, and behavior occurs
Time determine, just need to buy one simultaneously as different classifications are influenced the long-time such as difference, such as household electrical appliances by the time
It is secondary, dress ornament etc. by seasonal effect, variation can than very fast, so, classification belonging to object is also the factor of determining seed weight
One of.
Other than formula (3), in e-commerce website, foundation of the price as marking can also be added, such as formula (4):
Score=sseed*ssim+α*gapprice (4)
Wherein, gappriceIt is the price difference of seed object and analogical object, α indicates the parameter of regulation price, when α is timing
Expression above mentions the analogical object higher than seed object price, is negative then conversely, Score indicates final similar sequence point.α can be with
It presets, can also be determined by Q-Learning model learning with artificial experience according to demand.
In practical applications, can take the circumstances into consideration to use formula (3) or formula (4).
S406: the conventional ranking score of the first class object and the second class object in object to be presented is calculated.
The concrete mode for calculating conventional ranking score may refer to the prior art, for example, according to commodity one month it
Interior sales volume marking, sales volume more balloon score are higher.Which is not described herein again.
S407: displaying is ranked up to object to be presented according to similar ranking score and conventional ranking score.
It should be noted that can integrate two class scores for the second class object with two class ranking scores and obtain one
A final score, for example, two class scores are made even office's value, alternatively, being averaged again after first multiplied by weight summation.
The embodiment of the present application also discloses a kind of searching method, comprising the following steps:
1, based on the historical behavior of user, the historical behavior object of the user is determined.
2, the determining extension object with the historical behavior object with incidence relation.
3, result object relevant to search key in the extension object is determined.
The specific implementation of first three step may refer to above-described embodiment, and which is not described herein again.
4, the result object is ranked up.
The mode of sequence can be with are as follows: is ranked up using similitude sequence point, alternatively, being arranged using conventional sequence point
Sequence.
Searching method described in the present embodiment only executes S403 and S405 in searching method shown in Fig. 4, i.e., will only go through
Search library of the history analogical object as keyword.
Fig. 5 is a kind of searcher disclosed in the embodiment of the present application, comprising: the first determining module, the second determining module, the
Three determining modules, the 4th determining module and sorting module.
Wherein, the first determining module is used to determine the first class object, the first kind pair according to the search key of user
As for object relevant to described search keyword.Second determining module is used for the historical behavior based on the user, determines institute
State the historical behavior object of user.Third determining module is for the determining extension with the historical behavior object with incidence relation
Object.4th determining module is for determining the second class object relevant to the keyword in the extension object.Sorting module
Integrated ordered for carrying out to search result, described search result includes first class object and second class object.
Optionally, device shown in fig. 5 can also include: control module, if the quantity for second class object
Less than preset value, then increase accounting of first class object in described search result.And display module, for according to
Ranking score shows search result.
The above modules realize the concrete mode of respective function, may refer to above method embodiment, no longer superfluous here
It states.
Searcher shown in fig. 5 can be set on the server of website (such as e-commerce website).In website
After search engine receives search key, described device does not provide search result only in accordance with search key, also according to input
The historical behavior information of the user of search key provides search result, to improve search result towards the accurate of the user
Property.
If function described in the embodiment of the present application method is realized in the form of SFU software functional unit and as independent production
Product when selling or using, can store in a storage medium readable by a compute device.Base
In such understanding, the part of the embodiment of the present application the part that contributes to existing technology or the technical solution
Can be embodied in the form of software products, which is stored in a storage medium, including some instructions to
So that a calculating equipment (can be personal computer, server, mobile computing device or the network equipment etc.) executes this Shen
Please each embodiment the method all or part of the steps.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only deposits
Reservoir (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or
The various media that can store program code such as CD.
Each embodiment in this specification is described in a progressive manner, the highlights of each of the examples are with it is other
The difference of embodiment, same or similar part may refer to each other between each embodiment.
The foregoing description of the disclosed embodiments makes professional and technical personnel in the field can be realized or use the application.
Various modifications to these embodiments will be readily apparent to those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the application.Therefore, the application
It is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase one
The widest scope of cause.
Claims (17)
1. a kind of searching method characterized by comprising
Search key according to user determines that the first class object, first class object are relevant to described search keyword
Object；
Based on the historical behavior of the user, the historical behavior object of the user is determined；
The determining extension object with the historical behavior object with incidence relation；
Determine the second class object relevant to the keyword in the extension object；
Integrated ordered to search result progress, described search result includes first class object and second class object.
2. the method according to claim 1, wherein the historical behavior based on the user, determine described in
The historical behavior object of user includes:
The historical behavior object of the user is obtained from the historical behavior data of the user；
There is the extension object of incidence relation to include: for the determination and the historical behavior object
Calculate behavioral similarity sim (i, j) between any one object i and the seed object j of any one user and/or
sim(i,j；S, t, p), wherein sim (i, j) indicates that the user has the sum of the number of behavior, sim (i, j to i and j simultaneously；s,t,
P) indicate that the user has the sum of the number of behavior p to i and j simultaneously under s scene in t time range；
The extension object is obtained according to the similitude between each object and the historical behavior object of each user, it is described similar
Property include at least the behavioral similarity.
3. method according to claim 1 or 2, which is characterized in that it is described to search result carry out it is integrated ordered before,
Further include:
If the quantity of second class object is less than preset value, increase first class object in described search result
Accounting.
4. method according to claim 1 or 2, which is characterized in that it is described search result is carried out integrated ordered include:
The ranking score of described search result is calculated, second class object has similar ranking score and conventional ranking score,
First class object has the conventional ranking score, and the similar ranking score is different with the conventional ranking score.
5. according to the method described in claim 4, it is characterized in that, the similar ranking score be based on second class object with
The similarity and seed weight of the historical behavior object of the user determines that the seed weight is according to second class pair
As the time that the behavior type of second class object and behavior occur for affiliated classification, user determines.
6. according to the method described in claim 5, it is characterized in that, the similar ranking score is the similarity and described kind
The product of sub- weight.
7. according to the method described in claim 6, it is characterized in that, the similar ranking score is based further on the user's
The price difference of historical behavior object and second class object.
8. according to the method described in claim 4, it is characterized in that, it is described to search result carry out it is integrated ordered after, also
Include:
Described search result is shown according to the ranking score.
9. a kind of searcher characterized by comprising
First determining module determines the first class object for the search key according to user, first class object for institute
State the relevant object of search key；
Second determining module determines the historical behavior object of the user for the historical behavior based on the user；
Third determining module, for the determining extension object with the historical behavior object with incidence relation；
4th determining module, for determining the second class object relevant to the keyword in the extension object；
Sorting module, integrated ordered for carrying out to search result, described search result includes first class object and described
Second class object.
10. device according to claim 9, which is characterized in that second determining module is specifically used for:
The historical behavior object of the user is obtained from the historical behavior data of the user；
The third determining module is specifically used for:
Calculate behavioral similarity sim (i, j) between any one object i and the seed object j of any one user and/or
sim(i,j；S, t, p), wherein sim (i, j) indicates that the user has the sum of the number of behavior, sim (i, j to i and j simultaneously；s,t,
P) indicate that the user has the sum of the number of behavior p to i and j simultaneously under s scene in t time range；
The extension object is obtained according to the similitude between each object and the historical behavior object of each user, it is described similar
Property include at least the behavioral similarity.
11. device according to claim 9 or 10, which is characterized in that further include:
Control module is used for before the sorting module is integrated ordered to search result progress, if second class object
Quantity be less than preset value, then increase accounting of first class object in described search result.
12. device according to claim 9 or 10, which is characterized in that the sorting module is specifically used for:
The ranking score of described search result is calculated, second class object has similar ranking score and conventional ranking score,
First class object has the conventional ranking score, and the similar ranking score is different with conventional ranking score.
13. device according to claim 12, which is characterized in that the similar ranking score is based on second class object
Determine that the seed weight is according to second class pair with the similarity and seed weight of the historical behavior object of the user
As the time that the behavior type of second class object and behavior occur for affiliated classification, user determines.
14. device according to claim 13, which is characterized in that the similar ranking score be the similarity with it is described
The product of seed weight.
15. device according to claim 14, which is characterized in that the similar ranking score is based further on the user
Historical behavior object and second class object price difference.
16. device according to claim 12, which is characterized in that further include:
Display module, for showing described search result according to the ranking score.
17. a kind of searching method characterized by comprising
Historical behavior based on user determines the historical behavior object of the user；
The determining extension object with the historical behavior object with incidence relation；
Determine result object relevant to search key in the extension object；
The result object is ranked up.
Priority Applications (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710757313.5A CN109446402B (en) | 2017-08-29 | 2017-08-29 | Searching method and device |
| TW107119974A TW201913415A (en) | 2017-08-29 | 2018-06-11 | Search method and apparatus |
| US16/115,324 US20190065611A1 (en) | 2017-08-29 | 2018-08-28 | Search method and apparatus |
| PCT/US2018/048387 WO2019046329A1 (en) | 2017-08-29 | 2018-08-28 | Search method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710757313.5A CN109446402B (en) | 2017-08-29 | 2017-08-29 | Searching method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109446402A true CN109446402A (en) | 2019-03-08 |
| CN109446402B CN109446402B (en) | 2022-04-01 |
Family
ID=65437406
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710757313.5A Active CN109446402B (en) | 2017-08-29 | 2017-08-29 | Searching method and device |
Country Status (4)
| Country | Link |
|---|---|
| US (1) | US20190065611A1 (en) |
| CN (1) | CN109446402B (en) |
| TW (1) | TW201913415A (en) |
| WO (1) | WO2019046329A1 (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111475725A (en) * | 2020-04-01 | 2020-07-31 | 百度在线网络技术（北京）有限公司 | Method, apparatus, device, and computer-readable storage medium for searching content |
| CN112040250A (en) * | 2020-07-21 | 2020-12-04 | 拉扎斯网络科技（上海）有限公司 | Information processing method, information processing device, storage medium and electronic equipment |
| WO2021017306A1 (en) * | 2019-07-30 | 2021-02-04 | 平安科技（深圳）有限公司 | Personalized search method, system, and device employing user portrait, and storage medium |
| CN113763005A (en) * | 2020-09-23 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Picture advertisement pushing method, electronic equipment and computer readable storage medium |
| CN113761345A (en) * | 2021-02-09 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Search processing method, apparatus, device and storage medium |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR102373486B1 (en) * | 2021-08-09 | 2022-03-14 | 쿠팡 주식회사 | Method for providing brand information and apparatus for the same |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070088692A1 (en) * | 2003-09-30 | 2007-04-19 | Google Inc. | Document scoring based on query analysis |
| CN101751422A (en) * | 2008-12-08 | 2010-06-23 | 北京摩软科技有限公司 | Method, mobile terminal and server for carrying out intelligent search at mobile terminal |
| CN101908184A (en) * | 2009-06-04 | 2010-12-08 | 维鹏信息技术(上海)有限公司 | Control method and system for distributing information through multiple associated terminals |
| CN102254028A (en) * | 2011-07-22 | 2011-11-23 | 青岛理工大学 | A personalized product recommendation method and system integrating attribute and structure similarity |
| CN102479366A (en) * | 2010-11-25 | 2012-05-30 | 阿里巴巴集团控股有限公司 | Commodity recommendation method and system |
| CN103793388A (en) * | 2012-10-29 | 2014-05-14 | 阿里巴巴集团控股有限公司 | Method and device for search result sorting |
| CN104102328A (en) * | 2013-04-01 | 2014-10-15 | 联想(北京)有限公司 | Information processing method and information processing equipment |
| CN104731830A (en) * | 2013-12-24 | 2015-06-24 | 腾讯科技（深圳）有限公司 | Recommendation method, recommendation device and server |
| CN104794135A (en) * | 2014-01-21 | 2015-07-22 | 阿里巴巴集团控股有限公司 | Method and device for carrying out sorting on search results |
| CN106446210A (en) * | 2016-09-30 | 2017-02-22 | 四川九洲电器集团有限责任公司 | Multimedia file searching method, server and client |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9020835B2 (en) * | 2012-07-13 | 2015-04-28 | Facebook, Inc. | Search-powered connection targeting |
| US9292515B1 (en) * | 2013-03-15 | 2016-03-22 | Google Inc. | Using follow-on search behavior to measure the effectiveness of online video ads |
| US20140280174A1 (en) * | 2013-03-16 | 2014-09-18 | Elan Bitan | Interactive user-controlled search direction for retrieved information in an information search system |
| CN104636402B (en) * | 2013-11-13 | 2018-05-01 | 阿里巴巴集团控股有限公司 | A kind of classification of business object, search, method for pushing and system |
| GB2545548A (en) * | 2014-03-29 | 2017-06-21 | Camelot Uk Bidco Ltd | Improved method, system and software for searching, identifying, retrieving and presenting electronic documents |
| KR101670412B1 (en) * | 2015-04-03 | 2016-10-31 | 경북대학교 산학협력단 | Mornitoring system for near miss in workplace and Mornitoring method using thereof |
| CN104933172A (en) * | 2015-06-30 | 2015-09-23 | 百度在线网络技术（北京）有限公司 | Information pushing method and device based on user searching behavior |
-
2017
- 2017-08-29 CN CN201710757313.5A patent/CN109446402B/en active Active
-
2018
- 2018-06-11 TW TW107119974A patent/TW201913415A/en unknown
- 2018-08-28 WO PCT/US2018/048387 patent/WO2019046329A1/en active Application Filing
- 2018-08-28 US US16/115,324 patent/US20190065611A1/en not_active Abandoned
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070088692A1 (en) * | 2003-09-30 | 2007-04-19 | Google Inc. | Document scoring based on query analysis |
| CN101751422A (en) * | 2008-12-08 | 2010-06-23 | 北京摩软科技有限公司 | Method, mobile terminal and server for carrying out intelligent search at mobile terminal |
| CN101908184A (en) * | 2009-06-04 | 2010-12-08 | 维鹏信息技术(上海)有限公司 | Control method and system for distributing information through multiple associated terminals |
| CN102479366A (en) * | 2010-11-25 | 2012-05-30 | 阿里巴巴集团控股有限公司 | Commodity recommendation method and system |
| CN102254028A (en) * | 2011-07-22 | 2011-11-23 | 青岛理工大学 | A personalized product recommendation method and system integrating attribute and structure similarity |
| CN103793388A (en) * | 2012-10-29 | 2014-05-14 | 阿里巴巴集团控股有限公司 | Method and device for search result sorting |
| CN104102328A (en) * | 2013-04-01 | 2014-10-15 | 联想(北京)有限公司 | Information processing method and information processing equipment |
| CN104731830A (en) * | 2013-12-24 | 2015-06-24 | 腾讯科技（深圳）有限公司 | Recommendation method, recommendation device and server |
| CN104794135A (en) * | 2014-01-21 | 2015-07-22 | 阿里巴巴集团控股有限公司 | Method and device for carrying out sorting on search results |
| CN106446210A (en) * | 2016-09-30 | 2017-02-22 | 四川九洲电器集团有限责任公司 | Multimedia file searching method, server and client |
Non-Patent Citations (1)
| Title |
|---|
| 刘欢: "电子商务环境下个性化信息推荐服务效果调查研究", 《中国优秀硕士学位论文全文数据库 经济与管理科学辑》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021017306A1 (en) * | 2019-07-30 | 2021-02-04 | 平安科技（深圳）有限公司 | Personalized search method, system, and device employing user portrait, and storage medium |
| CN111475725A (en) * | 2020-04-01 | 2020-07-31 | 百度在线网络技术（北京）有限公司 | Method, apparatus, device, and computer-readable storage medium for searching content |
| CN111475725B (en) * | 2020-04-01 | 2023-11-07 | 百度在线网络技术（北京）有限公司 | Method, apparatus, device and computer readable storage medium for searching content |
| CN112040250A (en) * | 2020-07-21 | 2020-12-04 | 拉扎斯网络科技（上海）有限公司 | Information processing method, information processing device, storage medium and electronic equipment |
| CN113763005A (en) * | 2020-09-23 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Picture advertisement pushing method, electronic equipment and computer readable storage medium |
| CN113761345A (en) * | 2021-02-09 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Search processing method, apparatus, device and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| US20190065611A1 (en) | 2019-02-28 |
| TW201913415A (en) | 2019-04-01 |
| WO2019046329A1 (en) | 2019-03-07 |
| CN109446402B (en) | 2022-04-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20190018900A1 (en) | Method and Apparatus for Displaying Search Results | |
| US8631003B2 (en) | Query identification and association | |
| US9706008B2 (en) | Method and system for efficient matching of user profiles with audience segments | |
| CN104391999B (en) | Information recommendation method and device | |
| CN102760124B (en) | Pushing method and system for recommended data | |
| CN109446402A (en) | A kind of searching method and device | |
| CN104679771B (en) | A kind of individuation data searching method and device | |
| TWI512653B (en) | Information providing method and apparatus, method and apparatus for determining the degree of comprehensive relevance | |
| US20140229281A1 (en) | Taxonomy based targeted search advertising | |
| CN103793388A (en) | Method and device for search result sorting | |
| KR20100114860A (en) | Touchpoint customization system | |
| US20140258002A1 (en) | Semantic model based targeted search advertising | |
| CN108509466A (en) | A kind of information recommendation method and device | |
| CN109993583B (en) | Information pushing method and device, storage medium and electronic device | |
| CN105469263A (en) | Commodity recommendation method and device | |
| CN110020128A (en) | A kind of search result ordering method and device | |
| CN108153791A (en) | A kind of resource recommendation method and relevant apparatus | |
| CN106326318B (en) | Searching method and device | |
| CN103383702A (en) | Method and system for recommending personalized news based on ranking of votes of users | |
| CN103744904B (en) | A kind of method and device that information is provided | |
| CN106919611A (en) | Product information method for pushing and device | |
| CN106156351A (en) | Multimedia resource recommendation information generates method and device | |
| CN106447364A (en) | Information recommending method and information recommending device | |
| CN106919609A (en) | Product information method for pushing and device | |
| CN106897412A (en) | A kind of method and apparatus for recommending associated application based on intended application |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |