CN1768340A - System and method for providing preferred language ordering of search results - Google Patents

System and method for providing preferred language ordering of search results Download PDF

Info

Publication number
CN1768340A
CN1768340A CN 200480009097 CN200480009097A CN1768340A CN 1768340 A CN1768340 A CN 1768340A CN 200480009097 CN200480009097 CN 200480009097 CN 200480009097 A CN200480009097 A CN 200480009097A CN 1768340 A CN1768340 A CN 1768340A
Authority
CN
China
Prior art keywords
search results
language
search
preferred
preferred language
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
Application number
CN 200480009097
Other languages
Chinese (zh)
Other versions
CN100543730C (en
Inventor
约翰·兰平
本·戈梅斯
瑞树·麦格拉斯
阿米特·辛哈厄
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Google LLC
Original Assignee
Google LLC
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Google LLC filed Critical Google LLC
Publication of CN1768340A publication Critical patent/CN1768340A/en
Application granted granted Critical
Publication of CN100543730C publication Critical patent/CN100543730C/en
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Images

Landscapes

  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

A system and method for providing preferred language (39) ordering (85) of search results (38) is described. A search query (36) describing potentially retrievable information (22) provided in a plurality of search result languages (37) is received. A search (36) is executed by evaluating the search query (36) against information characteristics maintained in a searchable data repository. At least one preferred language (39) applicable to search results (38) generated responsive to the executed search (36) is dynamically determined. At least some of the search results (38) are ordered in consideration of the at least one preferred language (39).

Description

Be used to provide the system and method for the preferred language ordering of Search Results
Cross reference to related application
Present application for patent is advocated the U.S. Provisional Patent Application case the 60/459th that is entitled as " System And Method For Providing Preferred Language Ordering OfSearch Results " of the application in 31 days March in 2003 under 35 USC § 119 (e), No. 339 right of priority, and advocate to submit in 3 days April in 2003 under 35 USC § 120 the 10/407th, the right of priority of No. 476 non-temporary patent application cases of the U.S., the disclosure of these patent application cases is incorporated this paper by reference into.
Technical field
The present invention relates generally to the information location, and the present invention relates to a kind of system and method that the preferred language ordering of Search Results is provided specifically.
Background technology
Though later 1960s can be traced back in the internet, the universal availability of individual's calculating and internet and acceptance have caused explosive growth and the unforeseen progress at the information sharing technical elements.Specifically, WWW (" World Wide Web ") allows the audient in the worldwide may have access to imponderable quantity of information with the store electrons form revolutionaryly, comprises with the archives form with the sound (audio frequency) and vision (image and the video) information of real-time form.In brief, the desktop access of World Wide Web information bank of nearly all language in the user of each networking provides in fact endless worldwide.
Message exchange on World Wide Web is with the client-server mode operation.Independent clients carries out the network content for global retrieval with the form of Web browser usually and presentation layer is used.Described Web browser sends the request message of network content for global to central www server, and described central www server is as data-carrier store and retrieve stored storehouse.Described www server analysis request message and in response message, return the network content for global of being asked.
Search engine is evolved along with the World Wide Web utilization rate that increases, so that the user can effectively and search with saving time and retrieve relevant network content for global.Along with the increase of the quantity and the type of network content for global, the sophistication of search engine and degree of accuracy need to improve too.In general, search engine makes great efforts to provide the result of E.B.B. for responding a search inquiry.Yet, judge that quality is difficult, because the correlativity of the network content for global of being retrieved is subjective and interest, knowledge and attitude that depend on the user in essence.
The existing method that search engine uses is based on search query word and the word that index goes out from global web page is complementary.More senior method use such as description in " TheAnatomy of a Large-Scale Hypertextual Search Engine " (1998) of S.Brin and L.Page and be presented to the 6th of Page in September 4 calendar year 2001,285, judge the importance of the network content for global of being retrieved based on the hyperlink structure analysis described in No. 999 United States Patent (USP)s, the disclosure of these documents is incorporated this paper into the form of quoting.
A typical search query scenario starts from submitting a natural language problem or independent key word to search engine.Search engine has the potential data repository of retrieving the information characteristics of network content for global to description and carries out search and identification candidate's Search Results.Search often can return thousands of or even result up to a million, so the common rank or the most promising result's that scores the subclass only of most search engine.Also can introduce the target search result, for example advertisement or thematic information content.Then, forward Search Results is dedicates the user to, is generally the form of network content for global title, hyperlink and other descriptive information (such as the text extract that extracts from Search Results).
It is used that search engine can be world user everywhere usually.Therefore, provide the part of high-quality Search Results to provide those Search Results with the acceptable language of request user.Acceptable language comprises language and other acceptable language of user's appointment.For example, the user of a French preference also can receive the Search Results of English.Acceptable language also can comprise relational language and dialect.For example, Portuguese Search Results is acceptable for the user of common preference Spanish.At last, acceptable language can comprise dead language, as ancient Greek's language or old english; Or pseudo-language, as Klingon (Klingon).Dead language and pseudo-language usually not searched engine support, but can reflect science, history or the personal interest of asking the user.
Now, most Web browsers, www server and relevant World Wide Web are used and are used HTTP(Hypertext Transport Protocol) to handle the web information exchange.HTTP is a kind of sessionless agreement and the state that does not keep any identification user preference usually, comprises language.Uniquely can be used for representing that the information of the acceptable language of user is to be independent of preference or the preference in search inquiry itself that each HTTP affairs keeps.At first, the preference that provides of user or on World Wide Web client computer or www server, specify.The client preference as the language of being accepted by Web browser, is to communicate by the request message head.The server end preference is specified by the search engine option and is independent of and uses cookie (it must be retrieved from the World Wide Web client computer) or keep by each HTTP affairs of logging program before carrying out search.
Though specifying aspect can received language effectively, the user is provided with language preference in practice seldom clearly.In addition, too tool is restricted usually for language preference, presents " all or do not have " (all-or-nothing) normal forms.The language preference function is served as the Search Results filtrator, and those Search Results under preferred language only are provided, and filters out those Search Results under relevant or alternative language.
Similarly, the default setting that appointment can received language is no matter all can make the Search Results that provides suitable further complicated at client-side or at server end.Usually, default setting can be mistake.For example, can specify English according to the Web browser option is the default language preference, but this may be not suitable for presenting Search Results for the user who does not have a good command of English.
Secondly, the preference based on inquiry derives from the word in a given search inquiry.Yet search query word can not be used for the assertion language preference reliably because of some reasons.The first, such as the proper noun of name, place name or name often and language independent and can not indicate well and present the required ideal language of Search Results.For example, a search engine can not be judged the acceptable language of being made up of proprietary name " Elvis " of search inquiry.The second, search inquiry often only is made up of several independent words, and especially when nominal key, this can not provide the enough contexts that can be used for the assertion language preference usually.Be similar to proprietary name, individual malapropism can or cause the language misunderstanding with language independent.For example, the search inquiry of being made up of word " WaldorfAstoria " can mislead search engine.
Therefore, need provide a kind of dynamic judgement to be used for presenting the method for the language preference of Search Results to the user.Preferably, this method will be held the acceptable preference of user and the language of preference not too, and be included in the relevant and alternative language in the described language preference.
Also need a kind of method that presents Search Results according to the user preference language with orderly fashion.Preferably, this method can sort or score Search Results with the Search Results of those preferred languages of preference, also holds the Search Results of those other Languages simultaneously.
Summary of the invention
The invention provides a kind of dynamically judge in response to a search inquiry preferred language and the system and method that Search Results is sorted.Judge the user preference and the language of preference not too based on the assessment of search inquiry, user interface and Search Results feature.Judge the search inquiry feature from the metadata of describing described search inquiry.Also use Internet protocol (IP) address of search inquiry metadata and client-side and server end preference and client computer to judge user interface features.Judge the Search Results feature based on the assessment of each Search Results.Based on this search engine be used to consider preferred language and (if selection) not too preference language and organize the method for Search Results, the Search Results that retrieves in response to search inquiry is sorted.The ordering of described Search Results is to utilize a predetermined shift factor (shiftingfactor) to realize or be assigned to the numerical score of each Search Results and reclassified these Search Results by a weighting factor and realize by adjusting one.
One embodiment provides a kind of system and method that is used for the Search Results ordering.At least one preferred language of the Search Results can be applicable to produce in response to the search that retrievable information is potentially carried out and to be provided in to provide with plural number kind Search Results language dynamically is provided.Consider sort at least some Search Results of these Search Results of described at least one preferred language.
Another embodiment provides a kind of system and method that is used to provide the preferred language ordering of Search Results.The search inquiry that is used to describe potential retrievable information that reception provides with plural number kind Search Results language.But search is by carrying out the assessment that is retained in the information characteristics in the search data thesaurus.Dynamically judge at least one preferred language of the Search Results that can be applicable to produce in response to performed search.Consider sort at least some Search Results of described Search Results of described at least one preferred language.
Another embodiment provides a kind of system and method that is used for the ordering of dynamic assertion language preference and Search Results.Receive a search inquiry request message, and from described search inquiry request message, analyze at least one in word and the attribute, the potential searchable network content for global that provides with plural number kind Search Results language with identification.But by carrying out search to remaining at least one in information characteristics assessment word in the search data thesaurus and the attribute, and Search Results produces in response to performed search.Judge at least one preferred language.The feature of at least one below the assessment: the user interface and the Search Results of search inquiry request message, transmission search inquiry request message.Select described at least one preferred language based on the feature of being assessed.Consider at least some the Search Results orderings of described at least one preferred language to Search Results.Search Results is rendered as the Search Results response message.
The those skilled in the art is from the other embodiments of the invention of will being not difficult to find of embodiment hereinafter, wherein is considered to realize that by graphic extension optimal mode of the present invention describes embodiments of the invention.Should be appreciated that the present invention can comprise other and different embodiment, and can make modification to its some details aspect tangible, and can not break away from spirit of the present invention and category various.Therefore, should to regard as be illustrative and nonrestrictive to the drawings and specific embodiments in itself.
Description of drawings
Fig. 1 is for showing the block scheme according to the system of the preferred language ordering that is used to provide Search Results of the present invention.
Fig. 2 is the functional-block diagram of the search engine of exploded view 1.
Fig. 3 is the process flow diagram flow chart that search inquiry is carried out and Search Results is handled of the search engine of exploded view 1.
Fig. 4 is the data structure diagram with a request message of the search engine reception of the mode exploded view 1 of example.
Fig. 5 is the data structure diagram with a response message of the search engine transmission of the mode exploded view 1 of example.
Fig. 6 is for showing the method flow diagram that is used to provide the preferred language ordering of Search Results according to of the present invention.
Fig. 7 is a process flow diagram of showing the routine of the preferred language of judging the method that is used for Fig. 6.
Fig. 8 is the process flow diagram of the function of the assessment search inquiry feature of showing the routine be used for Fig. 7.
Fig. 9 is the process flow diagram of the function of the assesses user interface characteristics of showing the routine be used for Fig. 7.
Figure 10 is the process flow diagram of the routine of the rank search result of showing the method be used for Fig. 6.
Figure 11 utilizes the process flow diagram of a shift factor rank search result with the routine of the routine that is used for Figure 10 for showing.
Figure 12 utilizes the process flow diagram of a weighting factor rank search result with the routine of the routine that is used for Figure 10 for showing.
Embodiment
Fig. 1 is for showing the block scheme according to the system 10 of the preferred language ordering that is used to provide Search Results of the present invention.A plurality of individual customer machines 12 are by such as internet or as the internet 13 of the communication network of other form of approving of one of ordinary skill in the art and docking with a server 11 with communication mode.By handle the network content for global request and via its separately the user 19 of other operation of client computer 12 operate individual customer machine 12.
In general, each client computer 12 can be can be connected to a network (such as internet 13) and can with the mutual any type of computing platform of application program.The exemplary embodiment of individual customer machine is including but not limited to its various arrangements and configuration of personal computer, digital assistants, " intelligence " cell phone and pager, lightweight client computer (lightweight client), workstation, " making mute " terminating machine that is docked to an application server and one of ordinary skill in the art's approval.Internet 13 comprise various topologys, dispose and be arranged in the enterprise interoperable the network interconnectivity assembly that couples arrangement, wide area network and LAN (Local Area Network) and be including but not limited to the equivalent network technology of conventional wired, wireless, satellite, optical fiber and one of ordinary skill in the art's approval.
For the network content for global exchange, and especially in order to handle search, each client computer 12 is carried out a Web browser 18 (" browser "), by it search inquiry is sent to a www server 20 of carrying out on server 11.The information of the form that is network content for global is usually described or discerned to each search inquiry, and it is potential searchable by www server 20.Usually the feature that is expressed as word (such as key word etc.) is provided search inquiry and such as the attribute of language, character code etc., it makes a search engine 21 (also carrying out on server 11) can discern and beam back Search Results.Described word and attribute are the form of the metadata of the data of composition data of description.Search inquiry, search inquiry feature and metadata as other style, form or the definition of one of ordinary skill in the art approval are feasible.
Beaming back Search Results presents with the form with network content for global title, hyperlink and other descriptive information (such as the text extract that extracts from Search Results) usually to browser 18.Server 11 keep one wherein keep network content for global 22 adhere to memory storage 15.Network content for global 22 also can remotely be retained on other www server (not shown), and described long-range www server directly or indirectly interconnects and preferably can be conducted interviews by each client computer 12 by internet 13.
The network content for global 22 of search engine 21 preferred identification optimum matching search query word is to provide high-quality Search Results, such as in " The Anatomy of aLarge-Scale Hypertextual Search Engine " (1998) of S.Brin and L.Page and in September 4 calendar year 2001, being presented to the 6th of Page, 285, described in No. 999 United States Patent (USP)s, its disclosure is incorporated this paper by reference into.When identification and matching network content for global 22,21 pairs of information characteristics of describing potential searchable network content for global of search engine are operated, as being described further with reference to figure 2 hereinafter.Note that except single processing environment, server 20 (comprising www server 20 and search engine 21) is provided functional also can be by lax or closely couple and distribute or parallel computation is disposed and provided.
Individual computer system (comprising server 11 and client computer 12) comprises the universal programming digital computing system, described calculation element is made up of following part: a CPU (central processing unit) (being respectively processor 13 and 16), random access memory (being respectively storer 14 and 17), nonvolatile secondary storage device 15 (such as hard drives or CD-ROM drive), network or wave point and peripheral unit (comprising user interface components, such as keyboard and display).Comprise that the program code of software program and data are written into RAM being carried out and to be handled by CPU, and bear results to show, export, to transmit or to store.
Fig. 2 is the functional-block diagram of the search engine 21 of exploded view 1.The process or the program of the source code that each assembly is a computer program, write with conventional programming language (such as the C++ programming language), and its be presented with as one single handle, distribute or parallel configuration in object code or syllabified code and carrying out by one or more CPU, as known in the art.The various embodiment of source code and object code and syllabified code can be kept on the medium of an embodied on computer readable or be implemented on the transfer medium in the carrier wave.
Search engine 21 is made up of five assemblies: analyzer 31, index 32, score indicator 33, language advancing device 34 and renderer 35.In simple terms, search engine 21 receives a search inquiry 36 by browser 18 communications from user 19, carries out search, produces Search Results 38, considers language preference rank search result 38, and sends the Search Results 37 after the ordering.Search inquiry 36 preferably is provided as the request message of HTTP compatibility and the Search Results 37 after the ordering preferably is provided as the response message (as further describing with reference to figure 4 and Fig. 5 respectively hereinafter) of HTTP compatibility, but as the request of other form of one of ordinary skill in the art's approval and response to exchange be feasible.
In more detail, analyzer 31 receives search inquiry 36.Each search inquiry 36 is described potential retrievable information, such as network content for global 22.Analyzer 31 is parsed into search inquiry 36 indivedual marks (token) then.Described mark comprises that the head value and of forming metadata 48 comprises the entity body of actual search inquiry.Metadata 48 is copied to language advancing device 34.
But index 32 is by carrying out search to the information characteristics assessment search inquiry 36 that remaines in the search data thesaurus 49.Information characteristics is an actual network content for global 22 or for describing the metadata (such as hyperlink) of word and the attribute be used to discern network content for global.But as the search data thesaurus 49 of other structure that the those skilled in the art approved and tissue for feasible.In case finish described search, index 32 produces last set result 38 by feature application that will appointment in search inquiry 36 to institute's canned data.But as the search data thesaurus 49 of other structure that the those skilled in the art approved and tissue for feasible.
Index 32 can discern thousands of potentially or even Search Results 38 up to a million, so usually only a subclass of Search Results 38 (usually between 100 to 10,000) be left most promising Search Results 38.Also can introduce target search result (not shown), such as advertisement or thematic information content.Most promising Search Results 38 is then pressed the degree rank or the score qualitatively of match search query terms.Search Results 38 can be with the numerical value score to reflect relative quality of match or goodness.Score indicator 33 is assigned to each Search Results 38 with the expression quality of match with a numerical score 41.
Language advancing device 34 is carried out two major functions.At first, language advancing device 34 is judged the one or more preferred language 39 of each search inquiry 36 and is judged the language 40 of its one or more not too preference according to circumstances, as being described further with reference to figure 7 hereinafter.In one embodiment, can use a speech selection device (not shown) to judge these preferred languages 39.The second, language advancing device 34 consideration preferred languages 39 and (if available) the not too language 40 of preference come rank search result 38, as further describing with reference to Figure 10 hereinafter.In one embodiment, these Search Results 38 can use a Search Results sorting unit (not shown) to sort.For the purpose of efficient, a subclass of the most promising Search Results 38 of language advancing device 34 preference orderings, common in the scope of 15 to 30 Search Results, but also can use other scope of approving as the those skilled in the art.
In described embodiment, score indicator 33 is assigned as those Search Results 38 with higher matching degree with respect to other Search Results 38 corresponding higher ranks.For example, if Spanish is a preferred language 39, those Spanish Search Results 38 have the matching degree more higher than the Search Results 38 of (for example) English so.Yet, those not too the Search Results 38 of the language of preference (such as, Portuguese) also can have the matching degree more higher than the Search Results 38 of English, but its matching degree is lower than the matching degree of Spanish Search Results 38.Perhaps, can use a counting arrangement method, score indicator 33 is assigned as those Search Results 38 with higher matching degree with respect to other Search Results 38 corresponding lower coupling ranks by this.Search result rank as other style, appointment or definition that one of ordinary skill in the art approved is feasible.
In another embodiment, Search Results 38 is scored to reflect relative quality of match or goodness by numerical value.Score indicator 33 assigns a numerical score 41 to give the relative mass of each Search Results 38 with the expression coupling, the relatively lower numerical score 41 reflection preferable qualities of high value mark 41.For example, if Spanish is a preferred language 39, those Spanish Search Results 38 can have higher numerical score 41 than the Search Results 38 of (for example) English so.Yet, those not too the Search Results 38 of the language of preference (such as Portuguese) also can have the numerical score 41 more higher than the Search Results 38 of English, but its numerical score 41 is lower than the numerical score 41 of Spanish Search Results 38.Perhaps, can use a counting point system, score indicator 33 assigns low numerical score 41 to reflect preferable qualities than higher numerical score 41 by this.The Search Results score of other style, appointment or the definition that can approve as the those skilled in the art is for feasible.
For judging the preferred language 39 and the language 40 of preference not too, language advancing device 34 is assessed search inquiry features (SQ character) 43, user interface features (UI character) 44 and Search Results feature (SR character) 45.Can judge search inquiry feature 43 from metadata 48.Can judge user interface features 44 from any available language preference 42 that metadata 48 is kept by server 11 (server end) with being independent of each search inquiry 36.Can judge Search Results feature 45 from Search Results 38.
Language advancing device 34 rank search results 38.In one embodiment, the Search Results 38 with numerical ordering does not sort with a shift factor 46, as being described further with reference to Figure 11 hereinafter.In described embodiment, be adopted as two (2.0) constant shift factor 46 with two positions of Search Results 38 degradations with the language of non-preference.When Search Results 38 sorted by reducing matching degree with search inquiry 36, this shift factor 38 was suitable for.For example, the Search Results 38 of language of non-preference that occurs in the 3rd position of Search Results 38 tabulation can be downgraded to the 5th position.Also can adopt the shift factor of other form.For example, shift factor 46 can shift to an earlier date the Search Results 38 of the language of non-preference.Equally, shift factor 46 can (for example) add, subtracts, takes advantage of by using, mark, remove and the mathematical function of logarithm factor or can be a formula or function and as one of ordinary skill in the art approved its various combination and permutation are come rank search result 38.
In another embodiment, when one adjusts the formula preference and is used for the scope of numerical score 41, utilize a weighting factor 47 sort the numerical value score Search Results 38, will further describe with reference to Figure 12 hereinafter.One of ordinary skill in the art will be appreciated that, these weighting factors 47 can be that a constant, use (for example) add, subtract, take advantage of, mark, remove and the mathematical function of logarithm factor, or its various combination and permutation that can be that a formula or function and those skilled in the art approved.In described embodiment, numerical score 41 increases along with the increase of quality of match.For example, higher numerical score 41 lower numerical score 41 reflect preferable qualities.Perhaps, as one of ordinary skill in the art approval, can adopt a system by this numerical score 41 can reduce along with the increase of quality of match.In described embodiment, be the preferred language 39 or the situation of the language 40 of preference not too according to Search Results 38, use the numerical score 41 of two weighting factors 47 to increase each Search Results 38.Search Results 38, one weighting factor W for a kind of preferred language 39 LPProvide by following formula:
w LP = { ∀ s 1 → n : s i = s i + 1 2 }
Search Results 38, one weighting factor W for a kind of language 40 of not too preference LPLProvide by following formula:
w LPL = { ∀ s 1 → n : s i = ( s i × 2 ) + 1 3 }
These weighting factors when numerical score 41 the most suitable in 0.0 and 1.0 scope and when approximately evenly distributing.In described embodiment, has a given score value S iSearch Results 38 when related with a kind of preferred language 39 than more shifting to an earlier date when related with a kind of language 40 of not too preference.Yet, approve as one of ordinary skill in the art, but the Search Results 38 phase inverse ratios related with a kind of language 40 of not too preference and a kind of preferred language 39 related Search Results 38 more shift to an earlier date.Perhaps, be a kind of preferred language 39 or the situation of the language 40 of preference not too according to Search Results 38, can utilize the numerical score 41 of weighting factor 47 to reduce each Search Results 38.Approve as one of ordinary skill in the art, also can adopt the numerical score of other scope and the weighting factor of form.
In another embodiment, can adjust shift factor 46 and weighting factor 47 to hold more unreliable or preference of dynamic language judgement reliably.For example, short search inquiry 36 or rare Search Results 37 can reduce the accuracy that the preference of dynamic language is judged because the content that works is less.Can loosen shift factor 46 and weighting factor 47 with more not obvious ground rank search result 37.Perhaps, a longer search inquiry 36 or tediously long Search Results 37 can increase accuracy, and can increase shift factor 46 and weighting factor 47 with rank search result 37 more significantly.
At last, renderer 35 is by the Search Results 37 of browser 18 after user 19 presents ordering.Usually, because the complete or collected works of the Search Results 37 after the ordering can surpass the available presentation space on the browser 18, so only need to present the part of the Search Results 37 after the ordering.Presenting is communication by means of Search Results 37 after the ordering of a Search Results response message.
Fig. 3 is process flow Figure 50 that search inquiry is carried out and Search Results is handled of the search engine 21 of exploded view 1.Described process flow comprises five Main Stage.At first, accept and analyze a search inquiry 52 (process 51) and forward it to next stage.One storage data repository is carried out search (process 53) and Search Results 54 is relayed to next stage.Judge user 19 preference and the language 56 of preference (process 55) not too based on metadata and available contexts.With preference and not too the language 56 of preference be relayed to next stage to be used for rank search result 54 (process 57).Stage in the end, pass on and present Search Results 58 (process 59) after the ordering.Judge preferred language and not too stage (process 57) of the stage of the language of preference (process 55) and rank search result by cutting out the quality that Search Results has been strengthened Search Results according to the user language preference of dynamic judgement.
The data structure diagram of the request message 60 that Fig. 4 receives for the search engine of showing in the mode of example by Fig. 1 21.Request message 60 is a HTTP compatible request message, such as at D.Gourley and E.Totty " HTTP; the Definitive Guide " the 3rd chapter, the 43-73 page or leaf, O ' Reillyand Assocs., described in the Sebastopol, CA (2002), its disclosure is incorporated this paper by reference into.Request message 60 is made up of three parts: start line 61, header 62 and entity body 63.Start line 61 identifications send to the HTTP method of the input data (such as " POST ") of search engine 21 from browser 18.Start line 61 also comprises request URL(uniform resource locator) (URL) 68 and a HTTP version identifier.Exemplary request URL "/search.cgi " identification one searching request.
Header 62 by zero or the title and the value of more MIME compatibilities be composed of, it provides a description the metadata 48 of feature of the interface of browser 18 and entity body 63 (meaning promptly, search inquiry itself).The compatible header of four MIME is following to be provided for judging those metadata 48 to the acceptable language of a request user:
(1) accept character set (64): expression can be accepted or the character set of preference by request browser 18, for example ISO-Latin-1, eight expansions to the ASCII that supports western European languages.ISO-Latin-1 is also referred to as ISO-8859-1.
(2) accept language (65): represent the language of the acceptable or preference of user 19 with the order of preference, and comprise a figure of merit q according to circumstances, for example, to the strong preference of French (fr) with to the weak preference of English (en).
(3) content type (66): describe the medium type (for example plain text (text/plain text)) of entity body 63 and the character set of the entity body 63 that is used to encode (for example, ISO-Latin-1) can be expressed as a parameter.
(4) content language (67): the natural language of expression statement entity body 63, for example French (fr).
Approve as one of ordinary skill in the art: use to be equal to or related protocol and the request message format with other form of the similar information of above-mentioned header 62 and other header and parameter is provided also is feasible.
The data structure diagram of the response message 70 that Fig. 5 sends for the search engine of showing in the mode of example by Fig. 1 21.Response message 70 is response messages of a HTTP compatibility, and such as D.Gourley and E.Totty, described in the Id., its disclosure is incorporated this paper by reference into.Response message 70 also is made up of three parts: start line 71, header 72 and entity body 73.Start line 71 comprises a HTTP identifier, responsive state sign indicating number and human readable rational phrase.
Header 72 is composed of by the title and the value of zero or more MIME compatibilities, and it provides a description the metadata of the feature of entity body 73 (anticipating promptly each Search Results).The compatible header of two MIME is following to be provided for judging those metadata to the acceptable language of a request user:
(1) content type (74): describe the medium type (for example plain text (text/plain text)) of entity body 73 and the character set of the entity body 73 that is used to encode (for example, ISO-Latin-1) can be expressed as a parameter.
(2) content language (75): the natural language of expression statement entity body 73, for example English (en).
Approve as one of ordinary skill in the art: use to be equal to or related protocol and the response message format with other form of the similar information of above-mentioned header 72 and other header and parameter is provided also is feasible.
Fig. 6 is the process flow diagram of showing according to the method 80 of a kind of preferred language ordering that is used to provide Search Results 38 of the present invention.This method 80 is described as a series of process operations or step, and they can be carried out by (for example) search engine 21 (as shown in Figure 1).
Accept a search inquiry 36 from a user 19, and it is parsed into metadata 48 (square frame 81).But, search for but on search data thesaurus 49, carry out, and produce Search Results 38 (square frame 82) by to remaining in the information characteristics assessment search inquiry 36 in the search data thesaurus 49.Because can be potential real estate give birth to thousands of or even Search Results 38 up to a million, so only the subclass (usually between 100 to 10,000) of Search Results 38 is left most promising Search Results 38.Also can introduce target search result (not shown), such as advertisement or thematic information content.Known in this technology, before Search Results 38 is offered browser 18, Search Results 38 can be defined the level temporarily and is " original " webpage (structural data or not structural data), can extract the metadata of the feature of describing each Search Results 38 from these " original " webpages.Sort qualitatively these Search Results 38 with match search inquiry 36 (square frames 83) so that a rank or score to be provided, comprise above the numerical score 41 of the reflection Search Results quality of describing with reference to figure 2.
So far, discerned and Search Results 38 has been carried out rank or score.Then, decision request user 19 preferred language 39 and the language 40 of preference (square frame 84) not too, this will further describe with reference to figure 7 hereinafter.Then, according to preferred language 39 and (if discerning) not too the matching degree of the language 40 of preference (square frame 85) come rank search result 38, this will further describe with reference to Figure 10 hereinafter.For the purpose of efficient, rank search result 38 most promising subclass (usually in the scope of 15 to 30 Search Results) only preferably, but it will be understood by one of ordinary skill in the art that and can use other scope.At last, present Search Results 37 after the ordering by browser 18.Usually, only need to present the part of the Search Results 37 after the ordering, because the complete or collected works of the Search Results 37 after the ordering can surpass the available presentation space on the browser 38.In described embodiment, search inquiry 36 is provided as the request message 60 of a HTTP compatibility and each Search Results as the response message 70 of a HTTP compatibility and provide back to user 19.
In another embodiment, storage preferred language 39 and not too the language 40 of preference carry out searching request 36 simultaneously in order to use in the future from same request user 19.Equally, can enable or forbid preferred language 39 and (if available) the presenting of language 40 of preference not too, thereby control presents and utilizes preferred language ordering and the ability of after sorted Search Results 37 by " trigger " that provide by user interface is provided.In another embodiment, before Search Results being presented to request user 19, with the Search Results 37 after the ordering respectively by preferred language 39 and (if available) not too the language 40 of preference flock together.Perhaps, can be such as by using adjacent lines in the form or unit to arrange the not too Search Results 37 after the ordering of the language 40 of preference of preferred language 39 and (if available), to present near the Search Results of those non-preferred languages.Stop described routine then.
Fig. 7 is used to judge the process flow diagram of preferred language 39 with the routine 90 of the method 80 that is used for Fig. 6 for showing.The purpose of this routine is based on search inquiry feature 43, user interface features 44 and Search Results feature 45 and comes decision request user 19 the preferred language 39 and the language 40 of any not too preference.
At first, based on metadata 48 assessment search inquiry features 43 (square frames of analyzing from search inquiry 36 91), this will further describe with reference to figure 8 hereinafter.If based on the language (square frame 92) of search inquiry feature 43 decidable appointments, select described appointed language so as preferred language 39 (square frame 93).
Next step, is if from search inquiry feature 43, can not judge appointed language (square frame 92), assesses user interface characteristics 44 (square frame 94) so, this will further describe with reference to figure 9 hereinafter.If the undefined English of user interface is as acceptable language (square frame 95), the language of selecting each non-English so is as a preferred language 39, and selects English as a language 40 of preference (square frame 96) not too.
Next step if English is provided as the acceptable language of user interface (square frame 95), assesses Search Results feature 45 (square frame 97) so.In described embodiment, Search Results 38 is with a kind of the providing in two kinds of forms.At first, Search Results 38 can be gathered into the set of " original " webpage, goes out language feature from this set decidable.The second, Search Results 38 can be organized into the metadata of describing various features, comprises language feature and corresponding to the web page contents of Search Results 38.Can be by dynamically to judge the main language of each Search Results 38 as the content analysis of the description in the 6th, 167, No. 369 United States Patent (USP)s that were presented to Schulze on Dec 26th, 2000, the disclosure of this patent is incorporated this paper into the form of quoting.If most of results of Search Results 38 are English (square frame 98), select English so as a preferred language 39 (square frame 99).Otherwise, do not adopt any preferred language 39 or the English 40 of preference (square frame 100) not too, and routine returns.
Selected the example ( square frame 93,96 and 99) of one or more preferred language 39 for each, relational language (if existence) can be chosen to be extra preferred language 39 (square frame 101) according to circumstances.Relational language comprises the language and the dialect on those shared one common bases, and the user who is proficient in a kind of this speech like sound by this can understand (may have slight difficulty) relational language.For example, a user who is proficient in Spanish often can understand the information that provides with Portuguese.To preferred language 39 be done after any interpolation, described routine returns.
Fig. 8 is for showing the process flow diagram of assessment search inquiry feature 43 with the function 110 of the routine 90 that is used for Fig. 7.The purpose of this function is based on from the header 62 of search inquiry request message 60 and analyzes the available metadata 48 of gained and judge any preferred language 39.
At first, retrieval is corresponding to any available metadata 48 (square frame 111) of the header 62 of search inquiry request message 60.Parameter by assessment content language header 67 is provided by the language of contents (square frame 112) that provides in entity body 63.If specify (square frame 113), return the language (square frame 114) of appointment so.Otherwise, judge the character set (square frame 115) of the content that in entity body 63, provides by assessing contents-types header 66.If be appointed as language specific character collection (square frame 116), return the language (square frame 117) of appointment so.For example, ISO-2022-JP is a Japanese specific character collection.Otherwise, do not have any language preference and return (square frame 118).
Fig. 9 is for showing the process flow diagram of assesses user interface characteristics 44 with the function 120 of the routine 90 that is used for Fig. 7.The purpose of this function is based on user interface features and judges any preferred language 40.
Can often be appointed as user option by search engine 21 generation user interfaces and representation language preference.Therefore, at first obtain any available language preference (square frame 122).The language preference of server end or directly keep in the cookie of client computer storage is perhaps kept by search engine 21 and uses cookie or logging program to come access to discern each request user 19 without peer.If available, obtain suitable speech recognition cookie or carry out the appropriate users logging program.Obtain the language preference of being stored and it is used to judge the language preference (square frame 122) of any server end.If specify (square frame 123), return the language (square frame 124) of appointment so.
The language that browser 18 can limit Search Results 38 is accepted as the language of client-side language preference.Therefore, obtain any available metadata 48 (square frame 125), and accept the language (square frame 126) that language head 65 judges that the user interface of browser 18 is accepted by assessment corresponding to the head 62 of a search inquiry request message 60.Accept the language except that specifying, accept language head 65 and can comprise the figure of merit q in 0.0 to 1.0 scope that shows the language preference degree.If specify (square frame 127), return appointed language (square frame 120), have simultaneously less than those language of 1.0 figure of merit q and be appointed as the not too language 40 of preference.
If what do not have to specify the user interface be used for browser 18 accepts language (square frame 127), accept the character set (square frame 128) that character set head 64 judges that the user interface of browser 18 is accepted by assessment so.If be appointed as appointed language character set (square frame 130), return appointed language (square frame 131) so.
Otherwise, can not come the assertion language preference by the language preference of server end or client-side.Yet, still can be based on judging a language preference at URL68 and the on-site available information of client computer.The location in the IP territory that judgement is discerned in URL68 (square frame 132).For example, be appointed as " Www.acme.at" URL68 have an IP territory " .at ", it represents an Australian IP territory.Yet specific I P territory (such as " .com ") is to such an extent as to use so extensively can not draw useful language preference conclusion.If URL68 provides a useful IP territory (square frame 133), judge the locale language (square frame 134) of being assigned in described IP territory so and it is returned (135).If useless (square frame 133) judges the location that sends the client 12 of search inquiry 36 from it by assessment client computer IP parameter (it can be separated out by the head part of the transmission control protocol that sends search inquiry 36 therein (TCP) package) so.Client computer IP parameter provides an IP address, and it is the 32-bit number address of being write as four numerals separating by the cycle.Can search by the domain name service (Domain Name Service) of the last numeral in the IP address and judge an IP territory.Equally, if its to drawing language preference useful (square frame 137), judge the language (square frame 138) in IP territory so and it returned (square frame 139).In described embodiment, the main language in the territory of appointment in the IP address of selected client computer 12.Otherwise,, return English (square frame 140) so if IP address (square frame 137) is not provided.
Figure 10 is for showing the process flow diagram of rank search result 38 with the routine 145 of the method 80 that is used for Fig. 6.The purpose of this routine is based on not too the rank or the sort method of search engine 21 uses of the language 40 of preference come rank search result 38 by considering preferred language 39 and (if available).
At first, judge the method for utilizing with rank or rank search result 38 by search engine 21 (square frame 146).Be applicable to numeral adjustment (square frame 147) if be assigned to the numerical score 41 of Search Results 38, utilize a numerical score to calculate so, utilize a weighting factor (47) to come rank search result 38 (square frame 149) whereby, these will further describe with reference to Figure 12 hereinafter.Otherwise, coming rank search result 38 (square frame 148) by utilizing a shift factor 26, these will further describe with reference to Figure 11 hereinafter.Described routine returns then.
Figure 11 utilizes the process flow diagram of a shift factor 46 rank search results 38 with the routine 150 of the routine 130 that is used for Fig. 9 for showing.The purpose of this routine be relevant to preferred language 39 or not too the Search Results 38 of the language 40 of preference the Search Results 38 of non-preferred language is shifted downwards.In described embodiment, be adopted as two (2.0) constant shift factor 46 with two positions of Search Results 38 degradations with non-preferred language.This shift factor 38 is suitable for when coming rank search result 38 by minimizing with the matching degree of search inquiry 36.For example, the Search Results 38 that occurs in the non-preferred language in the 3rd position of tabulation of Search Results 38 can be downgraded to the 5th position.Also can adopt the shift factor of other form.For example, shift factor 46 can shift to an earlier date the Search Results 38 of non-preferred language.Equally, one of ordinary skill in the art it should be understood that shift factor 46 can add by comprising, subtracts, takes advantage of, mark, remove and the mathematical function of logarithm factor and above various combination and permutation are come rank search result 38.
In one embodiment, described routine operation is less than or equal to a subclass of the sum of Search Results 38, but the those skilled in the art it should be understood that the subclass choice criteria that can use other form.One variable n is set to be shown to the quantity of user 19 Search Results, and a upper limit UL of preferred language 39 and not too a upper limit LPUL of preferred language all be set to the quantity n of Search Results 38 and the smaller value of 2n (square frame 151).Followingly then circulate rank search result 38 (square frame 153-166) with iterative processing.At first, an index j is set to upper limit UL (square frame 152) and is that (square frame 153) carry out handled in timing at index j.Search Results is retained in an array Result[] in, if Result[j] be not a preferred language Pref_Lang and a Result[j] not a language Less_Pref_Lang of preference (square frame 154) not too, the following shift factor 46 that utilizes makes Result[j] degradation.A smaller's (square frame 155) and a temporary variable temp that one variable target_pos is set to 2j and upper limit UL are set to Result[j] (square frame 156).Residue Search Results 38 by [target_pos] added 1 and the Result[j+1 that is shifted] promote (square frame 157) and with Result[target_pos] be set to temp (square frame 158).Upper limit UL be set to target_pos subtract 1 (square frame 159) and not too the upper limit LPUL of preference be set to the not too upper limit LPUL and the smaller among the upper limit UL (square frame 160) of preference.
If Result[j] be the language Less_Pref_Lang (square frame 161) of not too preference, Result[j so] shift factor 46 following degradations utilized.A smaller's (square frame 162) and a temporary variable temp that variable target_pos is set among the upper limit LPUL of the floor (floor) of 1.5j and preference not too are set to Result[j] (square frame 163).Residue Search Results 38 by [target_pos] added 1 and the Result[j+1 that is shifted] promote (square frame 164) and with Result[target_pos] be set to temp (square frame 165).Not too the upper limit of preference (LPUL) is set to target_pos and subtracts 1 (square frame 166).
At last, index j is set to j and subtracts 1 (square frame 167) and continue processing next loop iteration (square frame 168), and described thereafter routine returns.
Figure 12 utilizes the process flow diagram of a weighting factor 47 rank search results 38 with the routine 170 of the routine 130 that is used for Fig. 9 for showing.The purpose of this routine is to come rank search result 38 with deflection preferred language 39 or a kind of Search Results in the language 40 of preference not too by recomputating the numerical score 41 that is assigned to each Search Results 38.In one embodiment, described routine is to the subclass operation of the quantity that is less than or equal to Search Results 38, it equals to be shown to the quantity of the user's of multiply by a remainder (for example 2 or 3) Search Results 38, but those skilled in the art will realize that the subclass choice criteria that can use other form.
In described embodiment, numerical score 41 increases with the increase of quality of match.For example, higher numerical score 41 lower numerical score 41 reflect better quality.Perhaps, one of ordinary skill in the art it should be understood that and can adopt one to use the system that reduces numerical score 41 with the quality of match increase.In described embodiment, two weighting factors 47 are that being used for according to Search Results 38 is preferred languages 39 or not too situation of one of the language 40 of preference increases the numerical score 41 of each Search Results 38.For the Search Results 38 that is one of preferred language 39, provide a weighting factor W by equation (1) LP:
w LP = { ∀ s 1 → n : s i = s i + 1 2 } - - - ( 1 )
Search Results 38 for one of language 40 of preference not too provides a weighting factor W by equation (2) LPL:
w LPL = { ∀ s 1 → n : s i = ( s i × 2 ) + 1 3 } - - - ( 2 )
When numerical score 41 in 0.0 to 1.0 scope and when approximately evenly distributing, these weighting factors are optimum.In described embodiment, when related, has a given mark S with one of preferred language 39 i Search Results 38 shifted to an earlier date than when related, comparing with one of language 40 of preference not too.Yet one of ordinary skill in the art it should be understood that the related Search Results 38 of the related Search Results of one of language 40 with preference not too 38 opposite comparable and one of preferred language 39 shifts to an earlier date.Perhaps, can according to Search Results 38 for one of preferred language 39 or not too the situation of one of the language 40 of preference utilize the numerical score 41 of weighting factor 47 to reduce each Search Results 38.The those skilled in the art it should be understood that also can adopt the numerical score of other scope and the weighting factor of form.
Described routine is to a subclass operation that is less than or equal to the sum of Search Results 38, and followingly passes through iterative processing circulation (square frame 172-177) and recomputate numerical score 41.One variable n is set to multiply by the quantity (square frame 171) of the Search Results of a remainder (for example 2 or 3).Search Results 38 is retained in an array Result[] in.In iterative processing circulation (square frame 172-177), recomputate the numerical score 41 of each Search Results 38 by a variable I index.During each iteration (square frame 172), if Result[i] be a preferred language Pref_Lang (square frame 173), then with Score[i] be set to quantity Score[i] and adding 1 half (square frame 167), meaning is equation (1).Otherwise, if Result[i] be a language Less_Pref_Lang of preference (square frame 175) not too, then with Score[i] be set to quantity Score[i] 2 extraordinarily go up 1 1/3rd (square frames 176), meaning promptly, equation (2).Otherwise, do not need to do any numerical score adjustment.Handle and continue next iteration (177).After all iteration, to Search Results 38 rearrangements (square frame 178), described then routine returns.
Though the present invention has been carried out detail display and description with reference to its embodiment, it will be understood by one of ordinary skill in the art that and to make above-mentioned in the form and details therein and other variation and can not break away from spirit of the present invention and category.

Claims (74)

1. system (10) that is used for rank search result (38), it comprises:
One speech selection device (21), it judges that dynamically (84) can be applicable at least a preferred language (39) that produces and the Search Results (38) of (86) is provided with plural number kind Search Results language (37) in response to the search (36) that potential retrievable information (22) is carried out: an and Search Results sorting unit (34), it considers described at least a preferred language (39) and at least some Search Results in the described Search Results (38) that sorts.
2. system according to claim 1 (10), it further comprises:
One search inquiry tokenizer (43), it uses a search inquiry (36) of specifying described search (36) and judges (84) described at least a preferred languages (39).
3. system according to claim 2 (10), it further comprises:
Described search inquiry tokenizer (43), it is used for assessing at least a language (42) that is used for described search inquiry (36) and at least one character code (36) that is used for described search inquiry (36) at least one.
4. system according to claim 1 (10), it further comprises:
One user interface tokenizer (44), it uses user interface features (44) to judge (84) described at least a preferred languages (39).
5. system according to claim 4 (10), it further comprises:
Described user interface tokenizer (44), its assessment are appointed as one and are independent of one and specify at least a scheduled instruction (42), at least a language of being accepted by described user interface (44) (42), at least one character code (36) of being accepted by described user interface (44) of a storage preference (39) of the search inquiry (36) of described search (36) to submit the network address (36) of client applications (36) of described search inquiry (36) at least one with one to from it.
6. system according to claim 1 (10), it further comprises:
One Search Results characterizes device (45), and it uses at least some Search Results in the described Search Results (38) to judge (84) described at least a preferred languages (39).
7. system according to claim 6 (10), it further comprises:
Described Search Results characterizes device (45), and its assessment is used at least a language (39) of one or more described Search Results (38).
8. system according to claim 1 (10), it further comprises:
One index (32), the described Search Results of its rank (38); With described Search Results sorting unit (34), it is with respect to sort at least some Search Results in the described Search Results (38) of described at least a preferred language (39).
9. system according to claim 8 (10), its further comprise following at least one:
Described Search Results sorting unit (34), its each this type of Search Results (38) at least one position (39) of demoting with the language (40) outside the described at least a preferred language (39), and with each this type of Search Results (38) of described at least a preferred language (39) at least one position (39) in advance; With
Described Search Results sorting unit (34), it is each this type of Search Results (38) of the language (40) outside the described at least a preferred language (39) at least one position (39) in advance, and with each this type of Search Results (38) of described at least a preferred language (39) at least one position (39) of demoting.
10. system according to claim 1 (10), it further comprises:
One score indicator (33), it assigns a numerical score (41) to give described Search Results (38); With a Search Results sorting unit (34), it adjusts the described numerical score (41) of at least some Search Results in the described Search Results (38) of described at least a preferred language (39).
11. system according to claim 10 (10), its further comprise following at least one:
Described Search Results sorting unit (34), its increase is assigned to the described numerical score (41) of the described Search Results (38) of described at least a preferred language (39), and keeps the described numerical score (41) of the described Search Results (38) that is assigned to the language (40) outside the described at least a preferred language (39); With
Described Search Results sorting unit (34), its minimizing is assigned to the described numerical score (41) of the described Search Results (38) of described at least a preferred language (39), and keeps the described numerical score (41) of the described Search Results (38) that is assigned to the language (40) outside the described at least a preferred language (39).
12. system according to claim 1 (10), it further comprises:
Described Search Results sorting unit (34), its classification have described at least some Search Results in the described Search Results (38) of adjusted numerical score (41).
13. system according to claim 1 (10), it further comprises:
Described speech selection device (21), it selects the language (40) of at least a not too preference.
14. a method (80) that is used for rank search result (38), it comprises:
Judge that dynamically (84) can be applicable to produce (82) and at least a preferred language (39) that the Search Results (38) of (86) is provided with plural number kind Search Results language (37) in response to the search (36) that potential retrievable information (22) is carried out; With
Consider at least some Search Results in described at least a preferred language (39) ordering (83) described Search Results (37).
15. method according to claim 14 (80), it further comprises:
Use one to specify the search inquiry (36) of described search (36) to judge (84) described at least a preferred languages (39).
16. method according to claim 15 (80), it further comprises:
At least one of at least a language (42) that assessment is used for described search inquiry (36) and at least one character code (36) that is used for described search inquiry (36).
17. method according to claim 14 (80), it further comprises:
Use user interface features (44) to judge (84) described at least a preferred languages (39).
18. method according to claim 17 (80), it further comprises:
Assessment (84) is appointed as one and is independent of at least a scheduled instruction (42), at least a language of being accepted by described user interface (42), at least one character code (36) of being accepted by described user interface of a storage preference (39) of specifying the search inquiry (36) of described search (36)
Submit the network address (36) of a client applications (36) of described search inquiry (36) at least one with one to from it.
19. method according to claim 14 (80), it further comprises:
Use at least some Search Results in the described Search Results (38) to judge (84) described at least a preferred languages (39).
20. method according to claim 19 (80), it further comprises:
Assessment is used at least a language (42) of one or more described Search Results (38).
21. method according to claim 14 (80), it further comprises:
The described Search Results of rank (85) (38); With
With respect at least some Search Results in described at least a preferred language (39) ordering (85) described Search Results (36).
22. method according to claim 21 (80), its further comprise following at least one:
With each this type of Search Results (38) degradation (85) at least one position of the language (40) outside the described at least a preferred language (39), and each this type of Search Results (38) that will described at least a preferred language (39) shifts to an earlier date (85) at least one position; With
With each this type of Search Results (38) of the language (40) outside the described at least a preferred language (39) (85) at least one position in advance, and with each this type of Search Results (38) of described at least a preferred language (39) at least one position of demoting.
23. method according to claim 14 (80), it further comprises:
Assign (83) one numerical score (41) to give described Search Results (38); With
The described numerical score (41) of at least some results in the described Search Results (38) of adjustment (83) described at least a preferred languages (39).
24. method according to claim 23 (80), its further comprise following at least one:
Increase the described numerical score (41) that (83) are assigned to the described Search Results (38) of described at least a preferred language (39), and keep the described numerical score (41) that (83) are assigned to the described Search Results (38) of the language (42) outside the described at least a preferred language (39); And reduce the described numerical score (41) that (83) are assigned to the described Search Results (38) of described at least a preferred language (39), and keep the described numerical score (41) that (83) are assigned to the described Search Results (38) of the language (42) outside the described at least a preferred language (39).
25. method according to claim 14 (80), it further comprises:
Classification (83) has described at least some Search Results of the described Search Results (38) of adjusted numerical score (41).
26. method according to claim 14 (80), it further comprises:
Select the language (40) of (84) at least a not too preference.
27. the medium of an embodied on computer readable (30), its preservation is used to carry out the code of method according to claim 14 (80).
28. a system (10) that is used to provide preferred language (39) ordering of Search Results (38), it comprises:
One analyzer (31), the search inquiry (36) of the potential retrievable information (22) that provides with plural number kind Search Results language (37) is provided in its acceptance one;
One index (32), but it is by assessing described search inquiry (36) and carry out a search (36) according to remaining in information characteristics (49) in the search data thesaurus (49); With
One language advancing device (language promoter) (34), its dynamically (84) judge at least a preferred language (39) that can be applicable to the Search Results (38) that produces in response to performed search (36), and consider sort at least some Search Results of described Search Results (38) of described at least a preferred language (39).
29. system according to claim 28 (10), it further comprises:
A kind of search inquiry tokenizer (43), its judgement (84) one character codes (36) and at least one of a language (42) that is used for described search inquiry (36); With
One speech selection device (21), it is selected corresponding to described at least one at least a language (42) in a character code (36) and the language (42) as described at least a preferred language (39).
30. system according to claim 28 (10), it further comprises:
One user interface tokenizer (44), its judgement (84) one employed at least a language of user interface (44) (42) are to handle described search inquiry (36); With
One speech selection device (21), it selects the employed described at least a language of described user interface (44) (42) as described at least a preferred language (39).
31. system according to claim 30 (10), it further comprises:
Described user interface tokenizer (44), it judges the employed at least a language of (84) described user interface (44) (42) based on following at least one: submit to the described network address (36), of a client applications (36) of described search inquiry (36) can accept at least one and a scheduled instruction (42) of being appointed as a storage preference (39) that is independent of described search inquiry (36) the character set (49) of language (42) and described client applications (36) from it.
32. system according to claim 28 (10), it further comprises:
One user interface tokenizer (44), its judgement (84) one are used for handling the default language (42) of described search inquiry (36) by a user interface (44); With
One speech selection device (21), it selects a language (42) outside the described default language (39) as described at least a preferred language (39), and selects the language (40) of described default language (39) as at least a not too preference.
33. system according to claim 32 (10), it further comprises:
One Search Results characterizes device (21), at least a Search Results language (42) of at least one Search Results of its judgement (84) (38); With
One speech selection device (21), it selects described at least a Search Results language (42) as described at least a preferred language (39).
34. system according to claim 33 (10), wherein said at least a Search Results language (42) is used for most of Search Results of described Search Results (38).
35. system according to claim 28 (10), it further comprises:
One Search Results sorting unit (34), its language (42) based on this type of Search Results (38) mates the described Search Results (38) that sorts with one of described at least a preferred language (39).
36. system according to claim 28 (10), it further comprises:
One Search Results sorting unit (34), its matching degree described Search Results (38) that sorts based on the language (42) of this type of Search Results (38) and described at least a preferred language (39).
37. system according to claim 36 (10), it further comprises:
Described Search Results sorting unit (34), it utilizes the demote described Search Results (38) of the language (42) outside the described at least a preferred language (39) of a predetermined shift factor (46).
38. according to the described system of claim 37 (10), wherein said predetermined factor (46) equals two (2.0) substantially.
39. system according to claim 36 (10), it further comprises:
Described Search Results sorting unit (34), it utilizes a predetermined shift factor (46) to shift to an earlier date the described Search Results (38) of the language (42) outside the described at least a preferred language (39).
40. system according to claim 28 (10), it further comprises:
One score indicator (33), its according to the matching degree of described information characteristics (49) each described Search Results (38) that sorts.
41. system according to claim 28 (10), wherein said Search Results (38) is assigned a numerical score (41), and described system further comprises:
One Search Results sorting unit (34), its increase are assigned to the described numerical score (41) of at least some Search Results in the described Search Results (38) of described at least a preferred language (39).
42. according to the described system of claim 41 (10), wherein said numerical score (41) is adjusted according to following formula:
s i = s i + 1 2
S wherein iComprise the described numerical score (41) that is used for each Search Results i (38).
43. according to the described system of claim 41 (10), wherein said Search Results (38) is assigned a numerical score (41), described system further comprises:
One Search Results sorting unit (34), its minimizing are assigned to the described numerical score (41) of at least some Search Results in the described Search Results (38) of described at least a preferred language (39).
44. according to the described system of claim 41 (10), it further comprises:
One speech selection device (21), the language (40) of at least a not too preference of its judgement (84); With
Described Search Results sorting unit (34), its increase are assigned to the described numerical score (41) of at least some Search Results in the described Search Results (38) of language (40) of described at least a not too preference.
45., wherein adjust described numerical score (41) according to following formula according to the described system of claim 44 (10):
s i = ( s i × 2 ) + 1 3
S wherein iComprise the described numerical score (41) that is used for each Search Results i (38).
46. according to the described system of claim 41 (10), it further comprises:
One speech selection device (21), the language (40) of at least a not too preference of its judgement (84);
With
Described Search Results sorting unit (34), its minimizing are assigned to the described numerical score (41) of at least some Search Results in the described Search Results (38) of language (40) of described at least a not too preference.
47. system according to claim 28 (10), it further comprises: a renderer, it presents described Search Results (38).
48. according to the described system of claim 47 (10), it further comprises:
Described renderer, its carry out following at least one: the control that presents of at least some Search Results in the described Search Results (38) of each described at least a preferred language (39) is enabled, at least some Search Results in the described Search Results (38) of each described at least a preferred language (39) are flocked together, and before presenting to described user, at least some Search Results in the described Search Results (38) of each described at least a preferred language (39) are arranged near at least some Search Results in the described Search Results (38) of at least a language (42) outside the described at least a preferred language (39).
49. system according to claim 28 (10), it further comprises:
One speech selection device (21), it selects the language (40) of at least a not too preference.
50. system according to claim 28 (10), it further comprises:
One speech selection device (21), it is included in the one or more relational language (39) in the described at least a preferred language (39).
51. a method (80) that is used to provide the preferred language ordering (85) of Search Results (37), it comprises:
Receive (81) one search inquiries (36), potential retrievable information (22) that provides with plural number kind Search Results language (37) is provided for it;
But by assessing described search inquiry (36) and carry out (82) one search (36) to remaining in information characteristics (49) in the search data thesaurus (49);
Judge that dynamically (84) can be applicable at least a preferred language (39) of the Search Results (38) of generation in response to the search (36) of performed (82); With
Consider sort at least some Search Results in (85) described Search Results (38) of described at least a preferred language (39).
52. according to the described method of claim 51 (80), it further comprises:
Judge at least one of (84) one character codes (36) and a language (42) that is used for described search inquiry (36); With
Select (84) corresponding to described at least one at least a language (42) in a character code (36) and the language (42) as described at least a preferred language (39).
53. according to the described method of claim 51 (80), it further comprises:
Judge that (84) are used for handling at least a language (42) of (84) described search inquiry (36) by a user interface (44); With
Select (84) described user interface (44) employed described at least a language (42) as described at least a preferred language (39).
54. according to the described method of claim 53 (80), it further comprises:
Judge the employed described at least a language of (84) described user interface (44) (42) based on following at least one: submit to the described network address (36), of a client applications (36) of described search inquiry (36) to be accepted at least one and a scheduled instruction (42) of being appointed as a storage preference (39) that is independent of described search inquiry (36) the character set (49) of language (37) and described client applications (36) from it.
55. according to the described method of claim 51 (80), it further comprises:
Judge that (84) are used for handling a default language (39) of described search inquiry (36) by a user interface (44);
Language (39) outside selection (84) described default language (39) is as described at least a preferred language (39); With
Select the language (40) of (84) described default language (39) as at least a not too preference.
56. according to the described method of claim 51 (80), it further comprises:
Judge that (84) are used at least a Search Results language (42) of at least one Search Results (37);
With
Select (84) described at least a Search Results language (42) as described at least a preferred language (39).
57. according to the described method of claim 56 (80), wherein said at least a Search Results language (42) is used for most of Search Results of described Search Results (38).
58. according to the described method of claim 51 (80), it further comprises:
Language (42) based on this type of Search Results (38) mates (85) the described Search Results (38) that sorts with one of described at least a preferred language (39).
59. according to the described method of claim 51 (80), it further comprises:
Based on the matching degree of the language (42) of this type of Search Results (38) and described at least a preferred language (39) (85) the described Search Results (38) that sorts.
60. according to the described method of claim 59 (80), it further comprises:
Utilize the demote described Search Results (38) of the language (42) outside (83) the described at least a preferred languages (39) of a predetermined shift factor (46).
61. according to the described method of claim 60 (80), wherein said predetermined factor equals two (2.0) substantially.
62. according to the described method of claim 59 (80), it further comprises:
Utilize a predetermined shift factor (46) to shift to an earlier date the described Search Results (38) of the language (42) outside (83) the described at least a preferred languages (39).
63. according to the described method of claim 51 (80), it further comprises:
According to (85) each the described Search Results (38) that sorts of the matching degree with described information characteristics.
64. according to the described method of claim 63 (80), wherein said Search Results (38) is assigned a numerical score (41), described method further comprises:
Increase the described numerical score (41) of at least some Search Results in the described Search Results (38) that (83) be assigned to described at least a preferred language (39).
65., wherein adjust described numerical score (41) according to following formula according to the described method of claim 64 (80):
s i = s i + 1 2
S wherein iComprise the described numerical score (41) that is used for each this Search Results i (38).
66. according to the described method of claim 64 (80), wherein said Search Results (38) is assigned a numerical score (41), described method further comprises:
Reduce the described numerical score (41) of at least some Search Results in the described Search Results (38) that (83) be assigned to described at least a preferred language (39).
67. according to the described method of claim 64 (80), it further comprises:
Judge the language (40) of (84) at least a not too preference; With
Increase the described numerical score (41) of at least some Search Results in the described Search Results (38) of language (40) that (84) be assigned to described at least a not too preference.
68. according to the described method of claim 67 (80), wherein said numerical score (41) is adjusted according to following formula:
s i = ( s i × 2 ) + 1 3
S wherein iComprise the described numerical score (41) that is used for each this Search Results i (38).
69. according to the described method of claim 64 (80), it further comprises:
Judge the language (40) of (84) at least a not too preference; And reduce the described numerical score (41) of at least some Search Results in the described Search Results (38) of language (40) that (84) be assigned to described at least a not too preference.
70. according to the described method of claim 51 (80), it further comprises:
Present (86) described Search Results (38).
71. according to the described method of claim 70 (80), it further comprises:
Carry out (86) following at least one: the control that presents of at least some Search Results in the described Search Results (38) of each described at least a preferred language (39) is enabled (86), at least some Search Results in the described Search Results (38) of each described at least a preferred language (39) are flocked together and before presenting to described user at least some Search Results in the described Search Results (38) of each described at least a preferred language (39) is arranged near at least some Search Results in the described Search Results (38) of at least a language (42) outside the described at least a preferred language (39).
72. according to the described method of claim 51 (80), it further comprises:
Select the language (40) of (84) at least a not too preference.
73. according to the described method of claim 51 (80), it further comprises:
Comprise (84) one or more relational language (39) in described at least a preferred language (39).
74. the medium of an embodied on computer readable (30), its preservation are used for carrying out the code according to the described method of claim 51 (80).
CNB2004800090977A 2003-03-31 2004-03-31 Be used to provide the system and method for the preferred language ordering of Search Results Expired - Lifetime CN100543730C (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
US45933903P 2003-03-31 2003-03-31
US60/459,339 2003-03-31
US10/407,476 2003-04-03

Publications (2)

Publication Number Publication Date
CN1768340A true CN1768340A (en) 2006-05-03
CN100543730C CN100543730C (en) 2009-09-23

Family

ID=36743351

Family Applications (1)

Application Number Title Priority Date Filing Date
CNB2004800090977A Expired - Lifetime CN100543730C (en) 2003-03-31 2004-03-31 Be used to provide the system and method for the preferred language ordering of Search Results

Country Status (1)

Country Link
CN (1) CN100543730C (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009124480A1 (en) * 2008-04-11 2009-10-15 华为技术有限公司 Search method, device and system
CN103399847A (en) * 2012-05-31 2013-11-20 微软公司 Application language library for managing computing environment language
CN105893517A (en) * 2008-03-12 2016-08-24 埃克斯凯利博Ip有限责任公司 System, method, and/or apparatus for reordering search results
CN106663116A (en) * 2014-11-19 2017-05-10 谷歌公司 Method, systems, and media for presenting links to media content
CN103678482B (en) * 2006-12-19 2018-02-16 飞扬管理有限公司 For including the technology of collection items in search result
US10282529B2 (en) 2012-05-31 2019-05-07 Microsoft Technology Licensing, Llc Login interface selection for computing environment user login

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103678482B (en) * 2006-12-19 2018-02-16 飞扬管理有限公司 For including the technology of collection items in search result
CN105893517A (en) * 2008-03-12 2016-08-24 埃克斯凯利博Ip有限责任公司 System, method, and/or apparatus for reordering search results
WO2009124480A1 (en) * 2008-04-11 2009-10-15 华为技术有限公司 Search method, device and system
CN101556588B (en) * 2008-04-11 2013-01-23 华为技术有限公司 Method, equipment and system for searching
CN103399847A (en) * 2012-05-31 2013-11-20 微软公司 Application language library for managing computing environment language
CN103399847B (en) * 2012-05-31 2017-11-17 微软技术许可有限责任公司 Applicational language storehouse for managing computing environment language
US10282529B2 (en) 2012-05-31 2019-05-07 Microsoft Technology Licensing, Llc Login interface selection for computing environment user login
CN106663116A (en) * 2014-11-19 2017-05-10 谷歌公司 Method, systems, and media for presenting links to media content

Also Published As

Publication number Publication date
CN100543730C (en) 2009-09-23

Similar Documents

Publication Publication Date Title
US9183311B2 (en) Ordering of search results based on language and/or country of the search results
JP5065679B2 (en) System and method for providing preferred country bias for search results
JP4731479B2 (en) Search system and search method
US7668887B2 (en) Method, system and software product for locating documents of interest
US6463430B1 (en) Devices and methods for generating and managing a database
US8484177B2 (en) Apparatus for and method of searching and organizing intellectual property information utilizing a field-of-search
JP5114380B2 (en) Reranking and enhancing the relevance of search results
US7809710B2 (en) System and method for extracting content for submission to a search engine
CN1559044A (en) Content information analyzing method and apparatus
US20060122997A1 (en) System and method for text searching using weighted keywords
CN1509446A (en) System and method for content delivery over wireless communication medium to protable computing device
US7949674B2 (en) Integration of documents with OLAP using search
US20020138465A1 (en) Apparatus for and method of searching and organizing intellectual property information utilizing a classification system
US7310633B1 (en) Methods and systems for generating textual information
CN1622086A (en) Searching in a computer network
US20090299978A1 (en) Systems and methods for keyword and dynamic url search engine optimization
CN101044481A (en) A method, system, and computer program product for searching for, navigating among, and ranking of documents in a personal web
CN1685341A (en) Blinking annotation callouts highlighting cross language search results
CN1282928A (en) Method and system for providing national language inquiry service
EP1620815A2 (en) System and method for providing preferred language ordering of search results
CN1492367A (en) Inquire/response system and inquire/response method
CN101063975A (en) Method and system for electronic text-processing and searching
US20100131495A1 (en) Lightning search aggregate
CN1620648A (en) Document file read system using network
JP2010140200A (en) Search result classification device and method using click log

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
C56 Change in the name or address of the patentee
CP01 Change in the name or title of a patent holder

Address after: California, USA

Patentee after: GOOGLE Inc.

Address before: California, USA

Patentee before: Google Inc.

CP01 Change in the name or title of a patent holder

Address after: California, USA

Patentee after: Google Inc.

Address before: California, USA

Patentee before: Google Inc.

CP01 Change in the name or title of a patent holder
CX01 Expiry of patent term

Granted publication date: 20090923

CX01 Expiry of patent term