CN100593168C - Variable personalization of search results in a search engine - Google Patents
Variable personalization of search results in a search engine Download PDFInfo
- Publication number
- CN100593168C CN100593168C CN200580014339A CN200580014339A CN100593168C CN 100593168 C CN100593168 C CN 100593168C CN 200580014339 A CN200580014339 A CN 200580014339A CN 200580014339 A CN200580014339 A CN 200580014339A CN 100593168 C CN100593168 C CN 100593168C
- Authority
- CN
- China
- Prior art keywords
- document
- user
- group
- interest
- theme
- 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
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
A search engine provides personalized rankings of search results. A user interest profile identifies topics of interest to a user. Each topic is associated with one or more sites, and a boost value, which can be used to augment an information retrieval score of any document from the site. Search results from any search are provided to the user, with a variable control of the ranking of the results. The results can be ranked by their unboosted information retrieval score, thus reflecting no personalization, or by their fully or partially boosted information retrieval scores. This allows the user to selectively control how their interests affect the ranking of the documents.
Description
The related application cross reference
The application's case according to 35U.S.C. § 119 (e) advocate based on filed an application on March 29th, 2004 the 60/557th, file an application in No. 783 application cases and on Dec 1st, 2004 the 11/002nd, the right of No. 474 application cases, above-mentioned application case all is incorporated herein with way of reference.The application's case also is incorporated herein the 10/646th, No. 331 (on August 22nd, 2003 filed an application) and the 10/641st, No. 489 (on August 14th, 2003 filed an application) U.S. patent application case with way of reference; All above-mentioned application cases are all owned together with the application's case.
Technical field
The application's case relates to the on-line search engine by and large, and in particular to the personalization to Search Results.
Background technology
The present art of on-line search engine technique can be in response to very advanced aspect the ability of the document (for example webpage or its some part, image or the like) of query term in its retrieval.The particular document that search engine fast searching now and query term are complementary.Yet, though current search engine usually returns the document that does not reflect user's basic interest exactly corresponding to the ad hoc inquiry item at all.Therefore, (one of them is very interested in physical culture for two different users, and another is interested in politics) in a given inquiry, will obtain identical result, for example " medicine inspection in the baseball ", even first user may to have implemented medicine inspection interested in understanding which team, then user may be to understanding with to carry out the relevant legislation of medicine inspection in physical culture interested.Therefore, need provide a kind of interest to make the mechanism and the method for Search Results personalization according to the user.
In addition, though the user may have applicable to one group of specific interest of handling its search inquiry, the user may not think always to make these interest influence Query Result.Therefore, mechanism and the method that the degree that is desirable to provide a kind of user of making and can influences his interest the result of a given inquiry is carried out variable adjustment.
Summary of the invention
The invention provides and make a user influence system and the mode that the degree of the representation as a result of a given inquiry is carried out variable adjustment his or her personal interest.The present invention has a plurality of embodiment, comprises a kind of system that is used to provide a search engine, a kind of method that is used to handle a search inquiry and a kind of user interface of described influence degree being carried out variable adjustment for the user.
From user's viewpoint, the invention provides a kind of mechanism that is used to make the Search Results personalization.Described user can stipulate that one group of interest (one group of theme of for example described user's interest or classification) is as an interest profile.Then, described user is committed to a search engine with a search inquiry.The described search engine pair document relevant with described inquiry retrieves and provides these documents as Search Results.At first (though not certainly all like this) arranged described Search Results under the situation of not considering user interest.This makes the user can see the correlativity of described document and not influenced by his personal interest.The user can control arrangement to document in the search result set by variable mode.In a preferred embodiment, this is undertaken by direct manipulation figure control element (a kind of personalized control).This graphic elements can have different diagrammatic representation forms, and it is expressed as a slide sign in one embodiment.Adjust described control element just can interest profile according to him in expressed user interest adjust described result's degree of alignment.In response to the change in location of described control element, described Search Results is rearranged, thereby make described user can see how his interest influences the ordering to the result.
In one embodiment, the arrangement of Search Results can work to increase the correlativity score of the document that is associated with subject catalogue.Each theme in the described catalogue all is considered to the document (for example website) relevant with described theme and is associated with one group.For each document (being similarly the website), all provide the lifting values of the degree of correlation of a described website of indication and described theme.All pages in the one given website all obtain the lifting values of this website.Described lifting values can be regarded as the proportionality factor of the correlativity of a described website of expression and described theme.During query processing, return this group document and score and do not consider that it is contained in the described subject catalogue.In addition, determine the independent score of each document, the basic correlativity score of wherein said document is to adjust by its lifting values (if there is) that is associated, for example by described correlativity score be multiply by described lifting values, thus the correlativity of described document of bi-directional scaling and described inquiry.Generally speaking, those documents that only belong to listed website in the described catalogue just can have a correlativity score that promotes, and other document in the described Search Results will can not have the correlativity score of described lifting.Therefore, according to the correlativity score of the lifting of document document is rearranged can change with the position of personalization control-position of described personalized control from have no effect (wherein said lifting values is had no effect to described correlativity score) to influence (wherein lifting values 100% ground with each document is applied to described correlativity score) fully, reach the intermediate application level of any amount.Therefore, the user can come immediately the document in the described Search Results to be rearranged by the position of adjusting described control.
One system embodiment comprise one be used for to the user provide the search engine of the website that the subject catalogue, of theme (classification) is used for searching in response to user's inquiry tape index and index, and an interest that is used for according to the user be that described Search Results is determined the personalization component that described Search Results at least a (and preferable a group) is arranged.One user interface embodiment comprise a subject catalogue with various themes, and wherein said user can select interest topic some graphics controllers (for example check box), and one allow the user to come described Search Results is arranged the graphics controller of adjusting according to his interest profile.
Next, graphic and diagram and technical information are set forth the present invention with reference to each.Describedly graphicly just describe various embodiments of the present invention for illustrational purpose.The those skilled in the art will easily know by hereinafter illustrating, reach the alternate embodiment of described structure, method and function shown in also can adopting, and this does not deviate from principle of the present invention.
Description of drawings
Fig. 1 one is used for the portal site of a search engine.
Fig. 2 is the webpage that is used to edit individual theme configuration file of the user.
Fig. 3 is another theme and subtopic webpage.
Fig. 4 is another subtopic webpage.
Fig. 5 shows the portal site of the search engine that has query term " stanford ".
Fig. 6 shows that the initial search result and about query term " stanford " is used for controlling the controller 608 of influence degree of described user's interest profile when described Search Results is arranged.
Fig. 7 is presented at the arrangement that described user has been adjusted to controller 608 Search Results shown in Figure 6 behind the centre position.
Fig. 8 is presented at the arrangement that described user has been adjusted to controller 608 Search Results shown in Figure 6 behind the maximum position.
Fig. 9 is presented at the subject web page after described user deletes " Computer " theme and selects one " Music " theme.
Figure 10 is presented at the arrangement of Search Results shown in Figure 6 after the change theme interest among Fig. 9.
Embodiment
Referring now to Fig. 1,, shows the inlet (or login page) 100 of a search engine among the figure.Described search engine comprise one wherein the user import the inquiry input field 102 of a search query term.One search button 106 is submitted to described search engine with described inquiry and is handled.Described user will also not be used to make the user profile of Search Results personalization to define to one, and this is shown a message when visiting described search engine first.Described user-accessible editor Edit Profile (configuration file) link 104 is so that create its personal interest.
Click editor's link 104 and just can visit the page shown in Fig. 2 200.This page shows the top layer theme 204 of a subject catalogue 202.Theme in the described catalogue can be (for example) and is being present in
Http:// dmoz.org/The place OpenDirectory Project (open catalogue plan) in defined theme or can be a certain other subject catalogue.(in this disclosure, term " theme " is interchangeable with " classification ", and " subtopic " is also interchangeable with " subclass ".) described top layer topic links can have one group of subtopic by arrow 206 indication.At any one time, with beginning a search by clicking StartSearching (beginning search) button 208 per family.Click once specific top layer theme and the user can be taken to a page that is associated with the subtopic of described theme.
The result of this step of the theme " Computer " on the page 200 has been clicked in Fig. 3 graphic extension wherein user.As shown in Figure 3, the subtopic 302 of theme " Computer " has been presented in the row of the right hand.What be close to each subtopic 302 is a check box 314.The user can select check box 314 to indicate him interested in a specific subtopic.Perhaps, the user can select check box 316 at whole subtopic itself " Computer (General) ".This indication user is interested in all subtopics of theme " Computer ".This user-selected group theme can be called user profile, personalized configuration file or interest profile.In the left side of Fig. 3, after the user has selected one or more interest topics, in interest hurdle 306, list these themes.What be close to each subject name is one to make the user can delete the Remove Links 308 of each theme respectively.The user also can delete all themes by selecting " Delete all preferences " link 310.
Described subtopic itself can have the additional subtopic layer of arbitrary quantity.For example, Fig. 4 graphic extension subtopic " Games " further has subtopic 402 " Board Games ", " Card Games " or the like.These recreation are close to the subtopic arrow 206 of " Games " theme label and present to the user in response to user's click.Equally, the user can select whole subclass by selecting any one check box in check box 404 or each subtopic check box 406.In one embodiment, except that top layer theme 204, also have two subtopic layers.This can provide enough the theme cover width and the size of space, and can not apply unnecessary computation complexity.
Again referring to Fig. 3, the user has indicated him interested in the computing machine usually.Can call searched page 100 once more by clicking Start Searching (beginning search) button 208, searched page 100 shows that no longer the request user creates the message of personalized configuration file.At this moment, the user imports a search inquiry in described search column, and inquiry described herein is " Stanford " shown in Fig. 5.Search engine is implemented search and Search Results is provided its index.Fig. 6 graphic extension initial search result page 600.In addition, Search Results 602 comprises that one group shows the page text segment 619 of the part of described document context together with a link 615 and of leading to the described page to the user by the document 604 of a URL 617 indication respectively.Near described page top is personalized controller 608.In the context of the application's case, " document " is interpreted as the medium of arbitrary type that can be indexed and be retrieved by a search engine, comprises webpage, image, multimedia file, text document, PDF or the like.Use generic term " document " and unexpectedly containing and limit the scope of the invention.
Personalized controller 608 makes the user come the arrangement of each document/page of Search Results is adjusted according to his interest (represented in the preferable personalized as described configuration file).Described personalized controller may be interpreted as the degree of control user's interest profile to the influence or the effect of the arrangement of Search Results.Controller 608 among the embodiment that it is a slide sign type controller that Fig. 6 is illustrated in its middle controller 608, controller 608 comprise a joystick 610 that moves to property dimension along the line between the maximal value 614 in minimum value 610.The controller 608 of this type is very directly perceived usually and be easy to the user and understand its mode of operation.Alternate embodiment comprises equally index dial type controllers that variable adjustment is provided, and some radio buttons that has a relevant predetermined adjustment amount respectively (for example five respectively with 0%, 20%, 40%, 80% and 100% radio button that is associated) between minimum and maximal value.The technician in user-interface design field will understand, and also can adopt the controller of other type to reach essentially identical effect.
In example shown in Figure 6, personalized controller 608 is set to its minimum value 612 always.Therefore, come Search Results is correspondingly arranged according to the correlativity of Search Results and inquiry and with the irrelevant information (for example PageRank) of inquiry, PageRank be disclosed in the name be called " the node aligning method in the linked database (Method for noderanking in a linked database) " the 6th, 285, in No. 999 United States Patent (USP)s, the full text of this patent and whole teaching content thereof are incorporated herein with way of reference.It should be noted that described Search Results is relevant with Stanford University (Stanford University) usually.
Referring now to Fig. 7,, be presented among the figure user slide sign 610 has been adjusted to one behind the centre position that approximately is in mid point between minimum value and the maximal value to the arrangement of Search Results 602.At this moment, each document 604 is rearranged, and those more relevant with user's interest (front be designated " Computers (General)) " have a higher order.In order to make the user distinguish the document more relevant with his interest, those documents are indicated by an individual icon 618 that is close to link text 615.Noticing, under the situation that slide sign 610 mediates, just is a personalization results every a document approximately, although and nonessential like this.Notice that further segment 619 texts have the text relevant with computing machine (for example " hacking (breaking in) "), even this term is not included in the initial search query.
Referring now to Fig. 8,, is presented at the arrangement that the user has been adjusted to slide sign 610 its maximum position 614 back Search Results 602 among the figure.At first, in this example, all documents 604 in the Search Results 602 all reflect user's personalized configuration file and explain with individual icon 618.Secondly, the document in the described Search Results is relevant with computing machine and Stanford university, thereby the reflection user is to the interest of computing machine, because it is relevant with inquiry " Stanford ".For example, preceding two results are relevant with " breaking in " at Stanford; Ensuing three documents are concrete relevant with the various computer science projects of Stanford.Therefore, the 3rd document is the Artificial Intelligence Laboratory (Artificial Intelligence Laboratory) at Stanford, the 4th document is the Center for the Study of Language and Information (language and Information Research Center) at Stanford, and the 5th document is the Database Group (database group) at Stanford.All the other documents are same relevant with " Computers " interest of user.Further notice, no longer coming across among the first eight result with linking of homepage Stanford university as first (for example maximally related) is the result when not personalized application (Fig. 6) and when only using medium personalization (Fig. 7) before this.
In Fig. 6-8 example illustrated, the user can dynamically adjust the arrangement of document in real time, and search engine need not to recomputate Search Results.This further helps the user can probe into described Search Results.
Shown in Fig. 6-8, an Edit Profile (editor's configuration file) link 616 can be at any time edited the user to his interest of user profile.Fig. 9 graphic extension user has been back to the subject catalogue page 200 shown in Fig. 2, deleted his existing C omputers theme interest and created " Music (music) " this different interest.Then, the user reselects Start Searching (beginning search) button 208.Search engine provides the last set result through upgrading, described last set result's document through upgrading will with before identical because do not change described query term as yet.Yet, for last search, the user has been positioned controller 610 personalized fully place, and therefore described search engine is used this setting and according to existing (and new) personalized configuration file (this moment is at " Music (music) ") document arranged.This greatest differences that in described Search Results, is produced of Figure 10 graphic extension.Herein, six results at top are all relevant with music and art, and before all are not in the top of described search result list about the result of computing machine.This shows how user's personalized configuration file can change the arrangement to one group of given document significantly.
In one embodiment of this invention, the personalization to Search Results is realized by following method.The technician in Design of Search Engine field can understand the basic structure and the operation thereof of the search engine that wherein can adopt the method easily.
1) when the user selects theme (classification), described theme (classification) is added in the trace file of tabulation of an interest IN who comprises him (wherein IN is the abbreviation of " interests (interest) ").Described interest can be expressed as text token (for example " computers (computing machine) ") or be expressed as adding index in the catalogue, and wherein each theme/subtopic all has a call number.When the described website of user capture, this interest list is retrieved from described trace file by system.Perhaps, described website can provide a user who makes the user register a user name and password therein to register the page, and user's interest is stored in the database at the search engine site place, and logins in response to the user it is retrieved.
2) when the user sends a search inquiry Q, just described tabulation IN is sent to a front-end web servers.
3) then, described front-end web servers is with two request (Q, empty) and (Q IN) issues a rear end search engine, and it is individually R_base (conventional) and R_in (personalized) results set that described rear end search engine returns two to described front-end web servers.
4) described basis inquiry (Q, sky) with in that not apply the search of under any personalized situation Q being carried out on the search engine identical, thereby R_base is the results set that described search engine will return in response to described inquiry usually.As indicated above, according to the correlativity of itself and described inquiry and for example any other factors such as PageRank come correspondingly described Search Results to be arranged, PageRank be disclosed in name be called " the 6th of the node aligning method in the linked database (Method for node ranking in a linked database); 285; in No. 999 United States Patent (USP)s, the full text of this patent and all the teaching content be incorporated herein with way of reference.
5) (Q IN) comprises user's interest list in Ge Xinghua inquiry.Described search engine is followed one and is similar to conventional processing controls flow process of inquiring about, but it is in activation one personalization component between functional period of keeping the score, with " lifting " document relevant with described interest IN set.
In one embodiment of this invention, use the mapping (each mapping of each the theme/subtopic in the described catalogue) of one group of following form to implement these liftings:
Website->promote
Thereby for example, the mapping that classification " Health (health) " has following form
Website->promote
It comprises following input:
nih.gov->5.8
cdc.gov-->7.9
med.stanford.edu->3.5
Nih.gov wherein, cdc.gov and med.Stanford.edu are for determining (with manual or automated manner) each website relevant with theme " Health ".In other words, for each theme in the described catalogue, all exist one group to determine the website relevant, and, all define a lifting values for each website in these websites with described theme.
Usually determine the lifting of website listed in the described subject catalogue by following mode:
A) produce one " site map ", each node of wherein said figure is website (being essentially the page on the same main frame), and according to from a website link to the page on another main frame page quantity come the edge between each node is weighted.Can use the figure of this same type to calculate all themes lifting mappings.
B) for each theme in the described catalogue, for example " Health " selects several its homepages to be shown in website among the OpenDirectory (open catalogue) as " initial " website S0.For example, for a university, as Stanford, initial website may be selected to arbitrary website with the .stanford.edu ending.
C) branch calculates for twice:
I) at first, discern one group by those websites among the S0 the website S1 of a large amount of links, wherein how to be fetched by the institute of the website among S0 chain in large quantities and be its assignment one weight according to each website among the S1.
Ii) then, discern those by those websites among the S1 the website S2 of a large amount of links, and it is weighted as S1.
D) promote website among the S2 with the weight that it was assigned to.
When personalized enquire has interest set ({ Health (health) for example, Music (music) }) time, then described personalization component is consulted mapping separately, and subsequently for the Query Result (being document) that is in these mappings, described personalization component promotes the IR score of these Query Results, so that it will obtain higher ordering.Described IR must be divided into the information retrieval score that described search engine provides for each document, and can with document relevance and for example other factors such as PageRank change.
For example, if http://www.med.stanford.edu/research/ is one of result of inquiry Q and the IN that comprises " Health ", then because of the described page is positioned on the website med.stanford.edu, described result's IR score will obtain one 3.5 times multiplication lifting.For a plurality of interest, the value in each correspondence mappings be multiply by together mutually obtain a combined effect.A kind ofly be used to described catalogue to promote website in the mapping to determine that the further details of the embodiment of described lifting values is set forth in the 10/646th, No. 331 U.S. patent application case, this patent application case is incorporated herein with way of reference.
6) described front-end web servers receives R_base and R_in.The R_base representative is set at 0 (or its minimum position 612) result 602 that the user saw when locating at personalized controller 608.The result 602 that the user was seen when R_in representative was in its maximum position 614 (promptly 10) at described slide sign.In one embodiment, described controller has 11 positions, but also can use any amount of position.Interpolation is carried out in the arrangement between 0 and 10 that is shown to the user; This means, by R_base and R_in, described personalization component is arranged (R_1...R_9) by each independent result is moved to the virtual route of position when controller 608 is in its maximum position from its position when controller 608 is in its minimum position along one in the middle of constructing.
7) how to provide the slide sign effect to the user:
Described front-end web servers is attached to R_base with the result who is not among the R_base among the R_in, to obtain a tabulation about all unique results (in R_base or R_in) that is called R_all.Then, described front-end web servers is that the continuous identifier of assignment as a result among the R_all is to serve as identifier.Then, described front-end web servers sends two items to user's browser:
A) each result's mapping among the R_all identifier->result }, it all comprises its identifier in described result for each document; And
B) N vector, its mark user should see the arrangement of (representing with identifier as a result), wherein N is the number of positions of slide sign, and each vector is all represented the sort result according to personalized degree.In this example, will have the position 0 to 10 of 11 vectors corresponding to controller 608.In one embodiment, vector 0 is [0,1,2,3,4,5,6,7,8,9,10], and its reflection has the result's of the local ident that begins with R_0 non-personalized the arrangement to assignment.Vector 10 be the ordering among the R_in as a result of fully personalization; For example, it may be [15,17,5,0,1,2,3,4,8,9,10], the document in the correlation map of document 15 and 17 (for example) subject catalogue that obtains promoting according to the lifting values in the mapping graph and the value that is associated with the position of controller 608 for its IR score wherein.Described N arrangement vector transferred to client just can make described client browser can recomputate described arrangement apace, rather than make server bear described calculating.
8) last, the Javascript code in the described results page makes user's browser use suitable in described 11 positional alignment vectors one to come described Search Results is resequenced in response to the user changes the position of controller 608.The further details of the embodiment of above-mentioned steps 6-8 is set forth in the 10/641st, No. 489 U. S. application case, and this application case is incorporated herein with way of reference.
Above elaborated the present invention particularly with reference to a possible embodiment.Be understood by those skilled in the art that the present invention also can be embodied as other embodiment.At first, the concrete name of each assembly, term, attribute, data structure or any other programming or the configuration aspects of capitalization all are not enforceable or Special Significance is arranged, and the mechanism that is used to make up the present invention or its feature can have different titles, form or agreement.In addition, described system can form by the combination structure by hardware and software as indicated above, or is implemented in the hardware element fully.In addition, described herein specific function of carrying out between various system components is only divided to exemplary, rather than enforceable; The function of being implemented by the individual system assembly also can be implemented by a plurality of assemblies, implements and also can be changed into by single component by the function that a plurality of assembly is implemented.
Some part of above-mentioned explanation provides feature of the present invention with algorithm and the symbolic representation form to the operation of information.These algorithmic descriptions and representation are the means that the technician of data processing field is used for most effectively the essence of its work being conveyed to others skilled in the art.These are interpreted as being implemented by computer program when operating in and being set forth with function or logical course.In addition, prove also that sometimes these operation arrangements being called module or coming appellation by the function title is easily, this does not lose its generality.
Unless concrete regulation is arranged according to the above description apparently in addition, otherwise should be appreciated that, this explanation in the whole text in, the argumentation that utilization such as " calculating " or " determining " or " identification " or similar term carry out is meant a computer system or step and process that similar computing electronics carried out, and described computer system or similar computing electronics are handled and conversion the data that are expressed as physics (electronics) amount in computer system memory or register or other this type of information storer, transmission or display device.
Some aspect of the present invention is included in process steps and the instruction that is described as algorithm pattern herein.It should be noted that process steps of the present invention and instruction can include in software, firmware or hardware, and when include in software, can download and reside on the employed different platform of real-time network operating system and can move from described different platform.
The invention still further relates to a kind of apparatus operating that is used to implement herein.This equipment can be at required purposes special configuration, and perhaps it can comprise a multi-purpose computer, and described multi-purpose computer is stored in one by one and can be started selectively by the computer program on the calculating readable media of computer access or reconfigure.This kind computer program can be stored in the computer-readable storage medium, the disk of (but being not limited to) any kind for example, comprise floppy disk, CD, CD-ROM, magnetooptical disc, ROM (read-only memory) (ROM), random-access memory (ram), EPROM, EEPROM, magnetic or optical card, or be suitable for the medium that store electrons is instructed and is coupled to any kind of a computer system bus respectively.The technician in integrated circuit (IC) design and Video Codec field should be appreciated that, the present invention can easily be made in various types of integrated circuit (comprising application specific integrated circuit (ASIC)) based on above-mentioned functions and structure explanation.In addition, the present invention can be contained in various types of video coding apparatus.
Algorithm that is provided and operation herein is not relevant inherently with arbitrary certain computer or miscellaneous equipment.Various general-purpose systems also can be with using according to the program of this paper teaching content, perhaps provablely can construct more special equipment easily and implement required method step.The those skilled in the art will easily know structure and the equivalent version that various these type systematics are required.In addition, the present invention is not set forth with reference to arbitrary certain programmed language.Should be appreciated that, can use various programming languages to implement described teaching content of the present invention herein, and any reference of language-specific all is intended to disclose implementation of the present invention and optimal mode.
At last, it should be noted that language as used in this specification is to choose for readable and guiding purpose in principle, and may not be to choose to be used for describing or to limit subject matter of the present invention.Therefore, disclosure of the present invention is intended to illustration but not limits the scope of the invention.
Claims (23)
1, a kind of be used for according to a user with the personalization as a result of a search inquiry by computer-implemented method, it comprises:
Retrieve an interest profile of a user, described interest profile comprises the data of a plurality of interest topics of representing described user, the document that each theme is relevant with described theme with one or more is associated, each described document is associated with a lifting values, and described lifting values is represented a degree of correlation of described document and described theme;
Receive the personalization input that a search inquiry and is represented the influence degree of a described interest profile from described user;
Each the information retrieval score that provides that is based upon in the described document in response to described search inquiry is arranged one group of document;
Described user is presented in the expression of described group document through arranging;
For described group document, use the described lifting values of the described document correspondence in the described group of document to adjust corresponding described information retrieval score through arranging;
Rearrange described group of document according to described information retrieval score through adjusting; And
Described group document through rearranging is presented to described user.
2, the method for claim 1 wherein presents described document through arranging and further comprises:
Be close to one and be contained in each document that the interest in described user's the described interest profile is associated and show an individual icon.
3, the method for claim 1, it further comprises:
Described personalized input is mapped directly to a manipulation of a controller.
4, method as claimed in claim 3, wherein said controller are a slide sign.
5, method as claimed in claim 3, wherein said controller are an index dial.
6, method as claimed in claim 3, wherein said controller are radio button.
7, the method for claim 1, wherein adjust described information retrieval score and further comprise:
Retrieve described user's described interest profile from a trace file that is associated with a webpage of described user's browsing.
8, the method for claim 1, wherein adjust described information retrieval score and further comprise:
Be coupled to the described user's of database retrieval of a search engine described interest profile from one with communication mode.
9, the method for claim 1, wherein said user's described interest profile comprises one group of text token, each text token is corresponding to an interest topic.
10, the method for claim 1, wherein said user's described interest profile comprise one group of subject index number, and each call number is corresponding to an interest topic.
11, the method for claim 1, it further comprises:
Determine each document whether be contained in one by one second web site url to first website in, the theme that described second website and is contained in described user's the described interest profile is associated.
12, method as claimed in claim 11, wherein said lifting values be based on one from described second website to the number of links of described first website.
13, the method for claim 1, it further comprises:
For each document that is associated with described theme in the described interest profile in the described group of document:
Determine that one comprises the website of described document;
For each and the theme that described website is associated, be a website and determine a lifting values; And
As a function of the described lifting values of described website, for described document is determined a lifting values.
14, the method for claim 1, it further comprises:
Arranging described group of document further comprises based on described search inquiry and carries out a search; And
Rearranging described group of document according to described information retrieval score through adjusting further comprises and dynamically rearranges described group of document and do not re-execute described search.
15, the method for claim 1 is wherein adjusted described information retrieval score according to a controller that is associated with personalization component by described lifting values.
16, method as claimed in claim 15, it further comprises:
Determine a plurality of arrangement vectors, its a plurality of positions with described controller are relevant, and each arranges the arrangement of vector representation one based on described group of document of the influence degree of described interest profile; And
Described group document through rearranging is presented to described user to be comprised based on one in the described a plurality of arrangement vectors of a choice of location of described controller.
17, the method for claim 1, it further comprises:
Receive the modification of a pair of described interest profile from described user, described modification comprises a kind of in following at least: a) at least one theme is added into described interest profile, and b) at least one theme deletes in described interest profile;
Determine one the 3rd arrangement of described group of document in response to described modified interest profile; And
With described the 3rd arrangement described document is presented to described user.
18, a kind of device that is used for the Search Results personalization, it comprises:
Be used for receiving the device of an interest profile of a user from described user, described interest profile comprises the data of expression for a plurality of interest topics of described user, the document that each theme is relevant with described theme with one or more is associated, each described document is associated with a lifting values, and described lifting values is represented a degree of correlation of described document and described theme;
Be used for receiving the device that on behalf of the personalization of the influence degree of a described interest profile, a search inquiry and import from described user;
Each information retrieval that provides that is used for being based upon in response to described search inquiry described document must assign to arrange the device of one group of document;
Be used for the device of presenting to described user through the group document of arranging with described;
For described group document, be used for using the described lifting values of the described document correspondence of described group of document to adjust the device of corresponding described information retrieval score through arranging;
Be used for rearranging the device of described group of document according to described information retrieval score through adjusting; And
Be used for device that described group document through rearranging is presented to described user.
19, device as claimed in claim 18, wherein said being used for further is configured to the device that described group document through arrangement is presented to described user:
Be close to one and be contained in each document that the interest in described user's the described interest profile is associated and show an individual icon.
20, device as claimed in claim 18, it further comprises:
Described personalized input is mapped directly to a manipulation of a controller.
21, device as claimed in claim 20, wherein said controller comprises a slide sign.
22, device as claimed in claim 20, wherein said controller comprises an index dial.
23, device as claimed in claim 20, wherein said controller comprises radio button.
Applications Claiming Priority (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US55778304P | 2004-03-29 | 2004-03-29 | |
US60/557,783 | 2004-03-29 | ||
US11/002,474 | 2004-12-01 |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2010100006457A Division CN101814079B (en) | 2004-03-29 | 2005-03-21 | Method and device for variable personalization of search results |
Publications (2)
Publication Number | Publication Date |
---|---|
CN101080709A CN101080709A (en) | 2007-11-28 |
CN100593168C true CN100593168C (en) | 2010-03-03 |
Family
ID=38907375
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN200580014339A Active CN100593168C (en) | 2004-03-29 | 2005-03-21 | Variable personalization of search results in a search engine |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN100593168C (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20230126493A1 (en) * | 2020-04-09 | 2023-04-27 | Siemens Energy Global GmbH & Co. KG | Monitoring an electrical-energy transmission system |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8150829B2 (en) * | 2008-04-11 | 2012-04-03 | Fujitsu Limited | Facilitating display of an interactive and dynamic cloud of terms related to one or more input terms |
US8655982B2 (en) * | 2009-12-28 | 2014-02-18 | Rakuten, Inc. | Information searching apparatus, number deciding method, information searching program, information searching system and recording medium |
CN101853308A (en) | 2010-06-11 | 2010-10-06 | 中兴通讯股份有限公司 | Method and application terminal for personalized meta-search |
CN105205139B (en) * | 2015-09-17 | 2019-06-14 | 罗旭斌 | A kind of personalization document retrieval method |
CN109299378B (en) * | 2018-10-26 | 2021-02-12 | Oppo广东移动通信有限公司 | Search result display method and device, terminal and storage medium |
US11863102B2 (en) * | 2019-03-04 | 2024-01-02 | Linak A/S | Linear actuator system and a method of setting up and programming power limit values for such an actuator system |
-
2005
- 2005-03-21 CN CN200580014339A patent/CN100593168C/en active Active
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20230126493A1 (en) * | 2020-04-09 | 2023-04-27 | Siemens Energy Global GmbH & Co. KG | Monitoring an electrical-energy transmission system |
Also Published As
Publication number | Publication date |
---|---|
CN101080709A (en) | 2007-11-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101814079B (en) | Method and device for variable personalization of search results | |
US11420059B1 (en) | Personalized network searching | |
US9940398B1 (en) | Customization of search results for search queries received from third party sites | |
US7810035B2 (en) | Browsing web content using predictive navigation links | |
CN101124609B (en) | Search systems and methods using in-line contextual queries | |
CN100593168C (en) | Variable personalization of search results in a search engine | |
WO2004013726A2 (en) | Method and apparatus for search ranking using human input and automated ranking | |
WO2004097569A2 (en) | Search engine supplemented with url's that provide access to the search results from predefined search queries | |
KR20080031928A (en) | Systems for and methods of finding relevant documents by analyzing tags | |
WO2001035270A1 (en) | Search dedicated web site and search method on internet | |
WO2008061133A1 (en) | Search result ranking based on attributes of search listing collections | |
Darie et al. | Searching the Catalog |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CP01 | Change in the name or title of a patent holder | ||
CP01 | Change in the name or title of a patent holder |
Address after: American California Patentee after: Google limited liability company Address before: American California Patentee before: Google Inc. |