CN106293119A

CN106293119A - A kind of method and apparatus carrying out information recommendation in input method

Info

Publication number: CN106293119A
Application number: CN201610616201.3A
Authority: CN
Inventors: 张润琦
Original assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Current assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Priority date: 2016-07-29
Filing date: 2016-07-29
Publication date: 2017-01-04

Abstract

It is an object of the invention to provide a kind of method and apparatus for carrying out information recommendation in input method.First the present invention determines one or more interest informations of user；Then when described user's input information, list entries based on described user, in conjunction with described interest information, determine the one or more recommended links corresponding with described list entries and described interest information；Finally in input method, described recommended links is presented to described user.Compared with prior art, present invention achieves the information recommendation settled at one go, it is possible to provide the user the recommendation information meeting user's request easily, decrease user and obtain the operating procedure of information, improve user and obtain the efficiency of information, improve Consumer's Experience.

Description

A kind of method and apparatus carrying out information recommendation in input method

Technical field

The present invention relates to input method technique field, particularly relate to a kind of skill for carrying out information recommendation in input method Art.

Background technology

Input method is the crucial channel of man-machine interaction, and user carries out information input by input normal direction equipment, and input method It is only capable of the character simply user inputted to input, without producing the mutual of other with user.

In the prior art, when user wishes to obtain certain link information, it usually needs by instruments such as search engines Retrieve, or open specific website and inquire about.This method needs through " clicking on website input inquiry " Etc. multiple steps, and if institute's fruiting quantities of retrieving or inquiring about too much, in addition it is also necessary to user screens, therefore complex operation, and Occupy a large amount of system resource.

Although some input method can directly provide a user with phase based on the website list entries that user is inputted Corresponding web site url, such as, inputs " Taobao " in input method, then provides the website links of " Taobao " in input method " www.taobao.com ", but this information providing formula is only capable of, for limited specific website list entries, therefore cannot Meet user and obtain the demand of link information.

Summary of the invention

It is an object of the invention to provide a kind of method and apparatus for carrying out information recommendation in input method.

According to an aspect of the invention, it is provided a kind of method carrying out information recommendation in input method, wherein, the party Method comprises the following steps:

A determines one or more interest informations of user；

B when described user's input information, list entries based on described user, in conjunction with described interest information, determine with Described list entries and the corresponding one or more recommended links of described interest information；

Described recommended links is presented to described user in input method by c.

Alternatively, described step a includes:

-according to the application related information of the one or more application in equipment corresponding to described input method and/or user institute Corresponding user related information, determines one or more interest informations of described user.

Alternatively, described application related information includes following at least any one:

The application type information of-described application；

The application label information of-described application；

The application precedence information of-described application.

Alternatively, described step b includes:

B1 when described user's input information, list entries based on described user, determine that the input of described user is crucial Word；

B2, based on described input key word, in conjunction with described interest information, determines corresponding with described list entries Or multiple recommended links.

Alternatively, described step b1 includes following at least any one:

-when described user's input information, list entries based on described user, determine that the input of described user is crucial Word, wherein, described input key word is the input high frequency words of described user, described input high frequency words history based on described user List entries and determine；

-when described user's input information, list entries based on described user, determine that the input of described user is crucial Word, wherein, described input key word be described user browse high frequency words, described in browse high frequency words history based on described user Browse record and determine.

Alternatively, described step b2 includes:

B21, based on described input key word, in conjunction with described interest information, determines the pass corresponding with described input key word Connection interest information；

-according to described input key word and described association interest information, determine corresponding with described list entries Individual or multiple recommended links.

Alternatively, described step b21 includes:

-based on literary composition information before and after corresponding to described input key word and described input key word, in conjunction with described interest Information, determines the association interest information corresponding with described input key word.

Alternatively, described step c includes:

-based on predetermined ordering rule, the one or more recommended links is ranked up, preferably recommends chain to determine Connect；

-in input method, described preferred recommended links is presented to described user.

Alternatively, described predetermined ordering rule includes following at least any one:

-based on the page type information corresponding to described recommended links；

-ageing based on corresponding to described recommended links；

-distance based on the geographical position corresponding to described recommended links with the geographical position corresponding to described user；

-based on the reliability corresponding to described recommended links；

-based on the access related information corresponding to described recommended links.

According to a further aspect in the invention, additionally provide a kind of recommendation apparatus carrying out information recommendation in input method, its In, this device includes:

Interest determines device, for determining one or more interest informations of user；

Link determines device, and for when described user's input information, list entries based on described user, in conjunction with described Interest information, determines the one or more recommended links corresponding with described list entries and described interest information；

Present device, for described recommended links being presented to described user in input method.

Alternatively, described interest determine device for:

The application type information of-described application；

The application label information of-described application；

The application precedence information of-described application.

Alternatively, described link determines that device includes:

Key word determines unit, is used for when described user's input information, list entries based on described user, determines institute State the input key word of user；

Recommend to determine unit, for based on described input key word, in conjunction with described interest information, determine and described input sequence Arrange corresponding one or more recommended links.

Alternatively, described key word determines that unit is for following at least any one:

Alternatively, described recommendation determines that unit includes:

Association interest determines unit, for based on described input key word, in conjunction with described interest information, determines defeated with described Enter the association interest information that key word is corresponding；

Association link determines unit, for according to described input key word and described association interest information, determines and institute State one or more recommended links that list entries is corresponding.

Alternatively, described association interest determine unit for:

Alternatively, present described in device for:

-ageing based on corresponding to described recommended links；

-based on the reliability corresponding to described recommended links；

According to a further aspect in the invention, a kind of input equipment is additionally provided, including recommendation apparatus as described above.

Compared with prior art, the present invention is determined by the interest information of user, and when user's input information, based on The list entries at family and described interest information, determine corresponding recommended links for user, and push away described in input method Recommend link to present；Thus, present invention achieves the information recommendation settled at one go, it is possible to provide the user easily and meet use The recommendation information of family demand, decreases user and obtains the operating procedure of information, improves user and obtains the efficiency of information, improves Consumer's Experience.

And, the present invention can also be correlated with according to the application of the one or more application in equipment corresponding to described input method User related information corresponding to information and/or user, determines one or more interest informations of described user；Thus the present invention Can automatically determine the interest information of the user more meeting user's request, and can the change of real-time tracing user interest information, The recommendation information more meeting user's request finally can be provided, improve Consumer's Experience.

And, when the present invention can also work as described user's input information, list entries based on described user, determine described The input key word of user, is then based on described input key word, in conjunction with described interest information, determines and described list entries phase Corresponding one or more recommended links；It is highly preferred that the input high frequency words that described input key word is described user, described defeated Enter high frequency words history based on described user list entries and determine；Or, it is highly preferred that described input key word is described use Family browse high frequency words, described in browse high frequency words historical viewings based on described user record and determine.Thus the present invention can From described list entries, determine the input key word of core, be therefore not necessarily based on a large amount of insignificant list entries and determine Recommended links so that determined by recommended links have more specific aim, reduce resource occupation, improve treatment effeciency, finally can Provide the user the recommendation information meeting user's request easily, decrease user and obtain the operating procedure of information, improve use Family obtains the efficiency of information, improves Consumer's Experience.

And, the present invention is also based on described input key word, in conjunction with described interest information, determines and closes with described input The association interest information that keyword is corresponding, then according to described input key word and described association interest information, determines and institute State one or more recommended links that list entries is corresponding；It is highly preferred that the present invention is also based on described input key word And described input key word corresponding before and after literary composition information, in conjunction with described interest information, determine and described input key word phase Corresponding association interest information.Thus, the present invention can judge the point of interest of user accurately, determines and more meets user's request Recommended links, decrease user and obtain the operating procedure of information, improve user and obtain the efficiency of information, improve user's body Test.

And, the present invention is also based on predetermined ordering rule, is ranked up the one or more recommended links, with Determine preferred recommended links, then in input method, described preferred recommended links is presented to described user；So that institute in Existing preferred recommended links has uniqueness, it is achieved that the information recommendation settled at one go, it is possible to provides the user easily and meets The recommendation information of user's request, decreases user and obtains the operating procedure of information, improves user and obtains the efficiency of information, improves Consumer's Experience.

Accompanying drawing explanation

By the detailed description that non-limiting example is made made with reference to the following drawings of reading, other of the present invention Feature, purpose and advantage will become more apparent upon:

Fig. 1 illustrates that a kind of recommendation apparatus for carrying out information recommendation in input method according to one aspect of the invention shows It is intended to；

It is a kind of for carrying out pushing away of information recommendation in input method that Fig. 2 illustrates according to a preferred embodiment of the present invention Recommend device schematic diagram；

Fig. 3 illustrates a kind of method flow for carrying out information recommendation in input method according to a further aspect of the present invention Figure；

Fig. 4 illustrates a kind of side for carrying out information recommendation in input method according to a preferred embodiment of the present invention Method flow chart；

Fig. 5 to Fig. 8 be shown respectively according to a preferred embodiment of the invention for carrying out information recommendation in input method Present schematic diagram.

In accompanying drawing, same or analogous reference represents same or analogous parts.

Detailed description of the invention

It should be mentioned that, some exemplary embodiments are described as before being discussed in greater detail exemplary embodiment The process described as flow chart or method.Although flow chart operations is described as order process, but therein permitted Multioperation can be implemented concurrently, concomitantly or simultaneously.Additionally, the order of operations can be rearranged.When it When having operated, described process can be terminated, it is also possible to have the additional step being not included in accompanying drawing.Described process Can correspond to method, function, code, subroutine, subprogram etc..

Alleged by within a context, " recommendation apparatus " is " computer equipment ", and also referred to as " computer ", referring to can be by fortune Row preset program or instruction perform the intelligent electronic device of the predetermined process process such as numerical computations and/or logical calculated, and it can To include processor and memorizer, processor the survival instruction performing to prestore in memory performs predetermined process process, Or performed predetermined process process by the hardware such as ASIC, FPGA, DSP, or combined by said two devices and realize.

Described computer equipment includes subscriber equipment and/or the network equipment.Wherein, described subscriber equipment includes but not limited to Computer, smart mobile phone, PDA etc.；The described network equipment includes but not limited to that single network server, multiple webserver form Server group or the cloud being made up of a large amount of computers or the webserver based on cloud computing (Cloud Computing), its In, cloud computing is the one of Distributed Calculation, the super virtual machine being made up of a group loosely-coupled computer collection. Wherein, described computer equipment isolated operation can realize the present invention, it is possible to access network and by with in network other meter The interactive operation calculating machine equipment realizes the present invention.Wherein, the network residing for described computer equipment includes but not limited to interconnection Net, wide area network, Metropolitan Area Network (MAN), LAN, VPN etc..

Those skilled in the art will be understood that heretofore described " recommendation apparatus " can be only subscriber equipment, i.e. by Subscriber equipment performs to operate accordingly；Can also be formed, i.e. by subscriber equipment is mutually integrated with the network equipment or server Being matched with the network equipment by subscriber equipment, incompatible execution is corresponding to be operated.

It should be noted that described subscriber equipment, the network equipment and network etc. are only for example, other are existing or from now on may be used Computer equipment or the network that can occur such as are applicable to the present invention, within also should being included in scope, and to draw It is incorporated herein by mode.

It should be noted that preferably, of the present invention to " recommendation apparatus " input method can be included, be included in input method Device etc..Wherein, recommendation apparatus of the present invention can be pacified in advance by the production firm of computer equipment or sales service business It is filled to this computer equipment, it is possible to be loaded into computer equipment by computer equipment from server.Those skilled in the art should be able to Understand, any device that can be used for realizing the function in the present invention, regardless of whether be loaded in computer equipment, be all contained in In protection scope of the present invention.

Here, those skilled in the art will be understood that present invention can apply to mobile terminal and non-moving end, such as, when with When family uses mobile phone or PC, all may utilize method or apparatus of the present invention and recommend and present.

Concrete structure disclosed herein and function detail are the most representational, and are for describing showing of the present invention The purpose of example embodiment.But the present invention can be implemented by many alternative forms, and is not interpreted as It is limited only by the embodiments set forth herein.

Although it should be appreciated that here may have been used term " first ", " second " etc. to describe unit, But these unit should not be limited by these terms.These terms are used to be only used to a unit and another unit Make a distinction.For example, in the case of without departing substantially from the scope of exemplary embodiment, it is single that first module can be referred to as second Unit, and second unit can be referred to as first module similarly.Term "and/or" used herein above include one of them or Any and all combination of more listed associated item.

Term used herein above is only used to describe specific embodiment and be not intended to limit exemplary embodiment.Unless Context refers else clearly, and singulative the most used herein above " ", " one " also attempt to include plural number.Also should When being understood by, term used herein above " include " and/or " comprising " specify stated feature, integer, step, operation, Unit and/or the existence of assembly, and do not preclude the presence or addition of other features one or more, integer, step, operation, unit, Assembly and/or a combination thereof.

It should further be mentioned that in some replace implementation, the function/action being previously mentioned can be attached according to being different from The order indicated in figure occurs.For example, depending on involved function/action, the two width figures in succession illustrated actually may be used Substantially simultaneously to perform or sometimes can perform in a reverse order.

Below in conjunction with the accompanying drawings the present invention is described in further detail.

Fig. 1 illustrates that a kind of recommendation apparatus for carrying out information recommendation in input method according to one aspect of the invention shows It is intended to；Wherein, described recommendation apparatus includes that interest determines that device 1, link determine device 2, present device 3.

Specifically, described interest determines that device 1 determines one or more interest informations of user；When described user inputs letter During breath, described link determines device 2 list entries based on described user, in conjunction with described interest information, determines and described input Sequence and the corresponding one or more recommended links of described interest information；Described present device 3 in input method by described Recommended links presents to described user.

Described interest determines that device 1 determines one or more interest informations of user.

Specifically, described interest determine device 1 by with described user's direct interaction, or by institute can be provided with other The equipment of the interest information stating user interacts, to obtain one or more interest informations that user is provided.Here, it is described Interest information can be by above-mentioned mutual directly acquisition, it is also possible to obtain one or more interest phases by above-mentioned interactive mode Pass information, and described interest relevant information is analyzed acquisition after extraction.Such as, user is inputted by modes such as chat inputs The interest relevant information of oneself, such as " I likes singing ", then can extract its interest information for " singing ", if user is by searching for Input inputs the name of multiple songs continuously, then can determine that its interest information is " music ".

Or, described interest determine device 1 by one or more relevant informations are analyzed, to determine described user One or more interest informations.Wherein, during described relevant information includes but not limited to equipment corresponding to described input method User related information corresponding to the individual or application related information of multiple application and/or user.

Topic under wherein, described interest information includes but not limited to various ranks that user paid close attention to or various classification, Theme.Such as, described interest information is " ornaments, play, travel, cuisines ", " China News ", " star Zhao Wei " etc..

Preferably, described interest determines that device 1 is according to the one or more application in equipment corresponding to described input method User related information corresponding to application related information and/or user, determines one or more interest informations of described user.

Specifically, described interest determine device 1 by with described input method corresponding to equipment interact, described to obtain One or more application included in equipment；Then, based on being directly analyzed described application obtaining, or answer based on described Title and/or version number etc. by modes such as real-time search, obtain the application related information corresponding to described application.

Preferably, described application related information includes following at least any one:

The application type information of-described application: wherein, described application type information includes but not limited to that application institute itself is right The type in the field that the type answered or application are applied, as game, video, do shopping, office etc.；

The application label information of-described application: wherein, described application label information includes but not limited to apply default setting Application label or the application label etc. that provided of a large number of users, such as, described application label information is clothing, cuisines, decorations Product, film；

The application precedence information of-described application: wherein, described application precedence information can be based on such as the access frequency of user Grading etc. secondary, application vendor determines；One application application priority the highest, then its corresponding various information (as Application type information or application label information etc.) be confirmed as interest information probability the highest.

Or, described interest determines that device 1, by interacting with described user, or can obtain described user with other The equipment of user related information interact, to obtain the user related information of described user.

Wherein, described user related information includes but not limited to following at least any one:

The userspersonal information of-described user: such as age, sex, educational background etc.；

The customer position information of-described user: the customer position information as acquired in position based on GPS, defeated based on user Address of theenduser information etc. determined by entering；

The input historical information of-described user: input certain entry or certain class entry such as user frequently, or user institute is defeated The entry major part entered comes from certain theme dictionary, then using this entry or this dictionary as described interest information；

Browsing histories information corresponding to-described user: such as net page browse information, the browsing file information of described user Deng, described browsing histories information can include browsing content, browse duration, browse frequency etc..

Then, described interest determines device 1 answering according to the one or more application in equipment corresponding to described input method With the analysis of the user related information corresponding to relevant information and/or user, directly determine or weight and determine the one of described user Individual or multiple interest informations.

Such as, " trip is comprised in described equipment containing the application of multiple game class or having the application label of multiple application Play ", then it is believed that the interest information of this user is " game "；Or, such as, should if described equipment includes 2 game class With with the application of cuisines class, but the priority corresponding to the application of cuisines class is higher, and after being weighted calculating, cuisines class is applied Corresponding weights are higher, then it is believed that the interest information of this user is " cuisines "；Or, such as, if described individual subscriber letter Breath is young woman, then it is believed that the interest information of this user is " dress ornament "；Or, such as, if the customer location of described user Information is " Shanghai ", then it is believed that the interest information of this user is " Shanghai "；Or, such as, if the browsing histories of described user Information contains user certain theme browsed record frequently, then this theme is defined as the interest information of this user； Or, such as, the browsing histories information of described user contains user multiple themes are browsed record, then can be based on user Different themes is browsed frequency, browses the information such as duration, by one or several the highest for weights after wherein weighted calculation theme It is defined as the interest information etc. of this user.

When described user's input information, described link determines device 2 list entries based on described user, in conjunction with described Interest information, determines the one or more recommended links corresponding with described list entries and described interest information.

Specifically, when described user's input information, described link determines that device 2 is by described interest information and described user List entries be combined, to scan for inquiry, so that it is determined that relative with described list entries and described interest information The one or more recommended links answered.

Wherein, the mode of described combination includes but not limited to described " interest information " and " list entries " as search word Inquire about, or in the data base corresponding to described " interest information " or website, inquire about " the input sequence of described user Row ", or in the one or more candidate link corresponding to described " list entries ", inquiry is described " interest information "；Or base In described " interest information " and " list entries ", redefine one or more search word, to come really based on above-mentioned search word Determine recommended links.

Wherein, described recommended links is corresponding with described " list entries " and described " interest information ".Described recommendation chain The quantity connect can be one optimum, it is also possible to be multiple.

Such as, as it is shown in figure 5, when the interest information of described user is " ornaments ", as user's input " descendant of the sun ", Then search " descendant of the sun, ornaments ", and using its corresponding link as described recommended links, wherein, described recommended links Corresponding theme is " Song Huiqiao is with money ornaments "；Or, can using " Song Huiqiao is with money ornaments " as the search word redefined, And using its corresponding commodity page link as described recommended links etc..

Here, those skilled in the art will be understood that and above-mentioned determine that the method for recommended links is merely illustrative, not to this Bright restriction, other can " list entries of described interest information with described user is combined, with scan for inquiry, So that it is determined that the one or more recommended links corresponding with described list entries and described interest information " it is equally applicable to this Invention, and be included in protection scope of the present invention.

Described recommended links is presented to described user in input method by the described device 3 that presents.

Specifically, described in present device 3 in input method, by the skin of input method or on by ejecting the side of bubble Formula, presents to described user by described recommended links.

If here, those skilled in the art will be understood that described recommended links is single link, then this chain can be directly given Title, theme or other information connect, user clicks on this information, then can jump to the page corresponding to this link or application.If Described be linked as multiple link, then can show by the way of with single theme this group link corresponding to theme, when user's point When hitting this theme, launch the list corresponding to multiple recommended links, or by modes such as the icons near click theme to launch List etc. corresponding to multiple recommended links；When user clicks on above-mentioned any recommended links, then jump to corresponding to this link The page or application.

Preferably, present device 3 described in based on predetermined ordering rule, the one or more recommended links to be carried out Sequence, to determine preferred recommended links；In input method, described preferred recommended links is presented to described user.

Specifically, present described in device 3 can by default setting, by determining in real time by the way of utilizing machine learning, Or by other means etc., determine one or more predetermined ordering rule.

Preferably, described predetermined ordering rule includes following at least any one:

-based on the page type information corresponding to described recommended links: wherein, described page type information includes but does not limits In shopping page, service providing page, news pages, the retrieval page etc..The different weighteds corresponding to page type, example As, the weight of shopping page is higher, and the weight of the retrieval page is relatively low；Further, for the same page of different users Type, corresponding weight is likely to difference, and such as, if the point of interest of first user is shopping, then the weight of shopping page is wanted Higher than the retrieval page, if the point of interest of the second user is news, then retrieve the weight of the page higher than shopping page etc..

-ageing based on corresponding to described recommended links: wherein, described ageing includes but not limited to this recommended links The on-line time of the corresponding page itself, if the page ageing the highest of more line in evening, its weight is the highest；Time described Effect property also includes the ageing of the service of the page as corresponding to this recommended links, such as, purchase by group and purchase by group service corresponding to the page The most expired etc., such as, same two reached the standard grade for a day purchase by group the page, and the service that purchases by group of the page one cuts after one month Only, the service that purchases by group of the page two ends after one year, then the weight of the page two can be first-class higher than the page.

-distance based on the geographical position corresponding to described recommended links with the geographical position corresponding to described user: i.e. Both distances are the nearest, then weight is the highest.

-based on the reliability corresponding to described recommended links: wherein, described reliability includes but not limited to whether be safety Certification page (such as the https page), whether it is the authority page etc. in this field.The recommended links that reliability is the highest, its institute is right The weight answered is the highest.

-based on the access related information corresponding to described recommended links: wherein, described access related information includes but does not limits In total clicking rate, always place an order rate, recent clicking rate, place an order rate, access time etc. in the recent period.Above-mentioned every corresponding numerical value is more Height, then the weight corresponding to this recommended links is the highest.

The one or more link, based on one or more predetermined ordering rules above-mentioned, is carried out by the described device 3 that presents Sequence or weighting sequence, thus using recommended links the highest for numerical value after sequence as preferred recommended links or by super after sequence Cross the recommended links of the certain threshold value quantity as preferred recommended links or according to predetermined preferred recommended links and select one Individual or multiple preferred recommended links etc..

Then, present device 3 described in, in described input method, described preferred recommended links is presented to described user.Its In, described rendering method with described in Fig. 1 to present device 3 same or similar, therefore do not repeat them here, and by way of reference It is incorporated herein.

It is a kind of for carrying out pushing away of information recommendation in input method that Fig. 2 illustrates according to a preferred embodiment of the present invention Recommend device schematic diagram；Wherein, described recommendation apparatus includes that interest determines that device 1 ', link determine device 2 ', present device 3 ', institute State link and determine that device 2 ' includes that key word determines unit 21 ' and recommends to determine unit 22 '.

Specifically, described interest determines that device 1 ' determines one or more interest informations of user；When described user inputs During information, described key word determines unit 21 ' list entries based on described user, determines the input key word of described user； Described recommendation determines that unit 22 ', based on described input key word, in conjunction with described interest information, determines relative with described list entries The one or more recommended links answered；Described recommended links is presented to described user in input method by the described device 3 ' that presents.

Wherein, described interest determines with described, device 1 ' presents that device 3 ' is identical with corresponding intrument shown in Fig. 1 or basic phase With, therefore here is omitted, and be incorporated herein by way of reference.

When described user's input information, described key word determines unit 21 ' list entries based on described user, determines The input key word of described user.

Specifically, when described user's input information, described key word determine unit 21 ' based on predetermined grammatical rules or Based on data bases such as predetermined key word dictionaries, the list entries of described user is analyzed, to determine that described user's is defeated Enter key word.

Such as, if described user have input a SVO sentence, such as " we go to see the descendant of the sun ", the most described key Word determines that unit 21 ' can be by the object in this sentence, and i.e. " descendant of the sun " is as described input key word；Can also be based on lacking Save the key word dictionary arranging or obtaining in real time, owing to " descendant of the sun " in this sentence is in described key word dictionary, because of This is using the descendant of the sun " as described input key word.

Additionally, described key word determines that unit 21 ' is also based on the list entries that user is on the screen, determine described The input key word of user.Such as, identify the input content in input frame, and as input key word etc..

Preferably, when described user's input information, described key word determines unit 21 ' input based on described user sequence Row, determine the input key word of described user, and wherein, described input key word is the input high frequency words of described user, described defeated Enter high frequency words history based on described user list entries and determine.

Specifically, described key word determines that unit 21 ', by the way of real-time or non real-time, obtains or many of user Individual input record, and described input record is analyzed, to obtain of described user or many from described input record Individual history list entries.

Then, described key word determine unit 21 ' to described history input record be analyzed, to as described in history input List entries in record is added up, or adds up after being split by described list entries, if the result of statistics gained In include and one or more exceed predetermined input number of times or the word of predetermined incoming frequency, then using this word as described input high frequency Word.When user inputs described input high frequency words again, using described input high frequency words as described input key word.

Such as, if described input record comprises " I likes the descendant of the sun ", " going to see the descendant of the sun together ", " too The story of a play or opera of descendant of sun " etc. history list entries, then by the analysis to described history list entries, it is known that above-mentioned multiple sequences In all include " descendant of the sun ", therefore, by " descendant of the sun " as described input high frequency words.

Here, those skilled in the art will be understood that described input high frequency words includes that inputting number of times exceedes pre-determined number Word, if input number of times is more than 10 times；Incoming frequency can also be comprised and exceed the word of preset frequency, as inputted number of times in one month Inferior more than 3.Here, other determine that the mode of input high frequency words such as can be suitably used for the present invention, it is also contained in the guarantor of the present invention In the range of protecting.

Or, it is preferable that when described user's input information, described key word determines that unit 21 ' is based on described user List entries, determines the input key word of described user, wherein, described input key word be described user browse high frequency words, Described browse high frequency words historical viewings based on described user record and determine.

Specifically, described key word determines that unit 21 ', by the way of real-time or non real-time, obtains or many of user Individual historical viewings record, and described historical viewings record is analyzed, to obtain described use from described historical viewings record One or more browsing contents at family.

Such as, when user browses certain page, described browsing content includes but not limited to the mark of user institute browsing pages Topic, label, brief introduction, text etc., can include but not limited to the multimedia messages such as word, picture, video in browsing content.

Then, described key word determines that the browsing content in described historical viewings record is analyzed by unit 21 ', as logical Cross and analyze the title of the described page, content etc., therefrom extract one or more phrase, the result of statistics gained includes one Individual or multiple words exceeding predetermined number of visits or making a reservation for browse frequency, then browse high frequency words using this word as described.Work as user When again browsing high frequency words described in input, using the described high frequency words that browses as described input key word.

Such as, if described historical viewings record comprises " descendant _ Baidupedia of the sun ", " descendant's video of the sun ", Browsing contents such as " descendant's casts of the sun ", then by the analysis to described browsing content, it is known that above-mentioned multiple browsing contents In all include " descendant of the sun ", therefore, using " descendant of the sun " as described in browse high frequency words.

Here, those skilled in the art will be understood that described in browse high frequency words and include that number of visits exceedes pre-determined number Word, if input number of times is more than 10 times；Can also comprise and browse frequency and exceed the word of preset frequency, such as number of visits in month Inferior more than 3.Here, other determine that the mode browsing high frequency words such as can be suitably used for the present invention, it is also contained in the guarantor of the present invention In the range of protecting.

Described recommendation determines that unit 22 ', based on described input key word, in conjunction with described interest information, determines and described input One or more recommended links that sequence is corresponding.

Specifically, described recommendation determines that described input key word and described interest information are combined to carry out by unit 22 ' Search inquiry, so that it is determined that the one or more recommended links corresponding with described list entries.

Wherein, the mode of described combination includes but not limited to described " interest information " and " input key word " as search Word is inquired about, or inquires about " the input pass of described user in the data base corresponding to described " interest information " or website Keyword ", or in the one or more candidate link corresponding to described " input key word ", inquiry is described " interest information "；Or It is based on described " interest information " and " input key word ", redefines one or more search word, with based on above-mentioned search Word determines recommended links.

Here, with linking described in Fig. 1, the example of described combination determines that device 2 is same or similar, therefore do not repeat them here, And be incorporated herein by way of reference.

Preferably, described recommendation determines that unit 22 ' includes that associating interest determines that unit (not shown) and association link are true Cell (not shown)；Wherein, described association interest determine unit based on described input key word, in conjunction with described interest information, Determine the association interest information corresponding with described input key word；Described association link determines that unit is according to described input key Word and described association interest information, determine the one or more recommended links corresponding with described list entries.

Specifically, after described key word determines that unit 21 ' determines described input key word, described association interest determines Unit is based on described input key word and one or more interest informations of described user, by dividing based on to described user Analysis or the analysis to big data, screen, to determine the pass corresponding with described input key word from described interest information Connection interest information.

Those skilled in the art will be understood that described association interest information can be unique, it is also possible to is multiple.

Wherein, such as analysis to the common point of interest of described user, the analysis to big data to the analysis of described user Such as to other multiple users selected or analysis of point of interest of determining etc. when inputting described input key word.

Such as, if the interest information of described user includes " cuisines ", " ornaments ", " tourism ", when the input key word of user When being " descendant of the sun ", then based on the interest information of " descendant of the sun " institute default setting, association interest information can be determined For " ornaments ", as shown in Figure 5.

Or, such as, if user has frequently clicked on many times point of interest about " cuisines ", if then the input of user is crucial When word is " descendant of the sun ", it is possible to described association interest information is defined as " cuisines ".

Or, such as, if the input key word of user is " Hangzhou West Lake ", then can be based on the recommended links to multiple users Analysis, the recommended links as corresponding to several users click " Hangzhou West Lake+tourism ", it is determined that with " Hangzhou West Lake " institute right The association interest information answered is " tourism ", and then can linking the tourist service corresponding with " Hangzhou West Lake " as recommending chain Connect, as shown in Figure 6.

Or, such as, if the input key word of user is " from you of star ", then can be based on the recommendation to multiple users Link analysis, as several users click " from star you+cuisines " corresponding to recommended links, it is determined that with " from You of star " corresponding to association interest information be " cuisines ", and using the link of corresponding " medicated beer fried chicken " as recommending chain Connect, as shown in Figure 7.

It is highly preferred that described association interest determines that unit is right based on described input key word and described input key word institute Before and after Ying, literary composition information, in conjunction with described interest information, determines the association interest information corresponding with described input key word.

Specifically, described association interest determines that unit can also be by literary composition before and after obtaining corresponding to described input key word Information, determines the association interest information corresponding to described input key word；Wherein, described context information includes but not limited to Input on the screen or the information of not yet upper screen input.Usually, described context information includes straight with described input key word Connect adjacent information；Preferably, described context information also includes the input information apart from the described certain character of input key word.

Such as, if described input key word is " spending thousand bones ", if " spending thousand bones " is " seeing " above, then associated interest letter Breath is defined as " video "；If the information above " spending thousand bones " is " object for appreciation ", is then associated interest information and be defined as " game ".

Described " seeing " or " object for appreciation " may each be upper screen or the information of not upper screen.As a example by Fig. 8, if described user is the most defeated Enter " object for appreciation ", have input the most again " spending thousand bones ", then " game " has been associated interest information as it, and " the thousand same masterhands of bone will be spent Trip " corresponding to link as described recommended links.

Described association link determines that unit, according to described input key word and described association interest information, determines with described One or more recommended links that list entries is corresponding.

Wherein, described association link determines that unit determines that the method for described recommended links determines dress with linking described in Fig. 1 Put 2 same or similar, therefore do not repeat them here, and be incorporated herein by way of reference.

Fig. 3 illustrates a kind of method flow for carrying out information recommendation in input method according to a further aspect of the present invention Figure.

Specifically, in step sl, described recommendation apparatus determines one or more interest informations of user；In step S2 In, when described user's input information, described recommendation apparatus list entries based on described user, in conjunction with described interest information, Determine the one or more recommended links corresponding with described list entries and described interest information；In step s3, described Described recommended links is presented to described user in input method by recommendation apparatus.

In step sl, described recommendation apparatus determines one or more interest informations of user.

Specifically, in step sl, described recommendation apparatus by with described user's direct interaction, or by can with other The equipment providing the interest information of described user interacts, to obtain one or more interest informations that user is provided.? This, described interest information can be by above-mentioned mutual directly acquisition, it is also possible to obtain one or many by above-mentioned interactive mode Plant interest relevant information, and described interest relevant information is analyzed acquisition after extraction.Such as, user is by chat input etc. Mode have input the interest relevant information of oneself, and such as " I likes singing ", then can extract its interest information is " singing ", if user Inputted the name of multiple songs by search input continuously, then can determine that its interest information is " music ".

Or, in step sl, described recommendation apparatus is by being analyzed one or more relevant informations, to determine State one or more interest informations of user.Wherein, described relevant information includes but not limited to equipment corresponding to described input method In the application related information of one or more application and/or user related information corresponding to user.

Preferably, in step sl, one or more according in equipment corresponding to described input method of described recommendation apparatus User related information corresponding to the application related information of application and/or user, determines one or more interest of described user Information.

Specifically, in step sl, described recommendation apparatus by with described input method corresponding to equipment interact, to obtain Take the one or more application included in described equipment；Then, based on being directly analyzed described application obtaining, or based on The title of described application and/or version number etc., by modes such as real-time search, obtain the relevant letter of the application corresponding to described application Breath.

Then, in step sl, described recommendation apparatus according in equipment corresponding to described input method one or more should Application related information and/or user corresponding to the analysis of user related information, directly determine or weight and determine described use One or more interest informations at family.

In step s 2, when described user's input information, described recommendation apparatus list entries based on described user, knot Close described interest information, determine the one or more recommended links corresponding with described list entries and described interest information.

Specifically, in step s 2, when described user's input information, described recommendation apparatus is by described interest information and institute The list entries stating user is combined, to scan for inquiry, so that it is determined that believe with described list entries and described interest One or more recommended links that manner of breathing is corresponding.

In step s3, described recommended links is presented to described user in input method by described recommendation apparatus.

Specifically, in step s3, described recommendation apparatus in input method, by the skin of input method or on pass through bullet Go out the mode of bubble, described recommended links is presented to described user.

Preferably, in step s3, the one or more can be pushed away by described recommendation apparatus based on predetermined ordering rule Recommend link to be ranked up, to determine preferred recommended links；In input method, described preferred recommended links is presented to described user.

Specifically, in step s3, described recommendation apparatus can by default setting, by utilizing machine learning by the way of Determine in real time or by other means etc., determine one or more predetermined ordering rule.

In step s3, described recommendation apparatus is based on one or more predetermined ordering rules above-mentioned, to one or many Individual link is ranked up or weights sequence, thus will numerical value is the highest after sequence recommended links as preferred recommended links or The recommended links of certain threshold value will be exceeded after sequence as preferred recommended links or according to the number of predetermined preferred recommended links Amount selects one or more preferred recommended links etc..

Then, in step s3, described preferred recommended links is presented to institute in described input method by described recommendation apparatus State user.Wherein, described rendering method is same or similar with step S3 described in Fig. 3, therefore does not repeats them here, and by quoting Mode be incorporated herein.

Fig. 4 illustrates a kind of side for carrying out information recommendation in input method according to a preferred embodiment of the present invention Method flow chart.

Specifically, in step S1 ' in, described recommendation apparatus determines one or more interest informations of user；In step In S21 ', when described user's input information, described recommendation apparatus list entries based on described user, determine described user's Input key word；In step S22 ' in, described recommendation apparatus, based on described input key word, in conjunction with described interest information, determines The one or more recommended links corresponding with described list entries；In step S3 ' in, described recommendation apparatus will in input method Described recommended links presents to described user.

Step corresponding with shown in Fig. 3 is identical or essentially identical, so place is or not wherein, described step S1 ' with described step S3 ' Repeat again, and be incorporated herein by way of reference.

In step S21 ' in, when described user's input information, described recommendation apparatus list entries based on described user, Determine the input key word of described user.

Specifically, in step S21 ' in, when described user's input information, described recommendation apparatus is based on predetermined grammer rule Then or based on data bases such as predetermined key word dictionaries, the list entries of described user is analyzed, to determine described user Input key word.

Such as, if described user have input a SVO sentence, such as " we go to see the descendant of the sun ", then in step In S21 ', described recommendation apparatus can be by the object in this sentence, and i.e. " descendant of the sun " is as described input key word；Can also Based on default setting or the key word dictionary of acquisition in real time, owing to " descendant of the sun " in this sentence is at described key word dictionary In, therefore using the descendant of the sun " as described input key word.

Additionally, in step S21 ' in, described recommendation apparatus is also based on the list entries that user is on the screen, determines The input key word of described user.Such as, identify the input content in input frame, and as input key word etc..

Preferably, in step S21 ' in, when described user's input information, described recommendation apparatus is based on described user defeated Entering sequence, determine the input key word of described user, wherein, described input key word is the input high frequency words of described user, institute State input high frequency words history based on described user list entries and determine.

Specifically, in step S21 ' in, described recommendation apparatus, by the way of real-time or non real-time, obtains of user Or multiple input record, and described input record is analyzed, to obtain of described user from described input record Or multiple history list entries.

Then, in step S21 ' in, described recommendation apparatus to described history input record be analyzed, to as described in history List entries in input record is added up, or adds up after being split by described list entries, if statistics gained Result includes and one or more exceeding makes a reservation for input number of times or the word of predetermined incoming frequency, then using this word as described input High frequency words.When user inputs described input high frequency words again, using described input high frequency words as described input key word.

Or, it is preferable that in step S21 ' in, when described user's input information, described recommendation apparatus is based on described use The list entries at family, determines the input key word of described user, wherein, described input key word be described user browse high frequency Word, described in browse high frequency words historical viewings based on described user record and determine.

Specifically, in step S21 ' in, described recommendation apparatus, by the way of real-time or non real-time, obtains of user Or multiple historical viewings record, and described historical viewings record is analyzed, to obtain institute from described historical viewings record State one or more browsing contents of user.

Then, in step S21 ' in, the browsing content in described historical viewings record is analyzed by described recommendation apparatus, By the title of the page, content etc. as described in analysis, therefrom extract one or more phrase, the result of statistics gained comprises There is one or more word exceeding predetermined number of visits or making a reservation for browse frequency, then this word is browsed high frequency words as described.When User again input described in when browsing high frequency words, using the described high frequency words that browses as described input key word.

In step S22 ' in, described recommendation apparatus is based on described input key word, in conjunction with described interest information, determines and institute State one or more recommended links that list entries is corresponding.

Specifically, in step S22 ' in, described input key word and described interest information are combined by described recommendation apparatus To scan for inquiry, so that it is determined that the one or more recommended links corresponding with described list entries.

Here, the example of described combination is same or similar with step S2 described in Fig. 3, therefore do not repeat them here, and pass through The mode quoted is incorporated herein.

Preferably, described step S22 ' include step S221 ' (not shown) and step S222 ' (not shown)；Wherein, exist Step S221 ' in, described recommendation apparatus is based on described input key word, in conjunction with described interest information, determines and closes with described input The association interest information that keyword is corresponding；In step S222 ' in, described recommendation apparatus is according to described input key word and described Association interest information, determines the one or more recommended links corresponding with described list entries.

Specifically, when described step S21 in ' determine described input key word after, in step S221 ', described recommend dress Put the one or more interest informations based on described input key word and described user, by based on the analysis to described user Or the analysis to big data, screen from described interest information, to determine the association corresponding with described input key word Interest information.

It is highly preferred that in step S221 ' in, described recommendation apparatus is crucial based on described input key word and described input Literary composition information before and after word is corresponding, in conjunction with described interest information, determines the association interest letter corresponding with described input key word Breath.

Specifically, in step S221 ' in, described recommendation apparatus can also be by obtaining corresponding to described input key word Context information, determines the association interest information corresponding to described input key word；Wherein, described context information include but It is not limited to input on the screen or the information of not yet upper screen input.Usually, described context information includes closing with described input The information of keyword direct neighbor；Preferably, what described context information also included apart from the described certain character of input key word is defeated Enter information.

In step S222 ' in, described recommendation apparatus, according to described input key word and described association interest information, determines The one or more recommended links corresponding with described list entries.

Wherein, in step S222 ' in, described recommendation apparatus determines method and the step described in Fig. 3 of described recommended links S2 is same or similar, therefore does not repeats them here, and is incorporated herein by way of reference.

It should be noted that the present invention can be carried out at software and/or software in the assembly of hardware, such as, can adopt Realize with special IC (ASIC), general purpose computer or any other similar hardware device.An embodiment In, the software program of the present invention can perform to realize steps described above or function by processor.Similarly, the present invention Software program (including the data structure being correlated with) can be stored in computer readable recording medium storing program for performing, such as, and RAM memory, Magnetically or optically driver or floppy disc and similar devices.It addition, some steps of the present invention or function can employ hardware to realize, example As, perform the circuit of each step or function as coordinating with processor.

It addition, the part of the present invention can be applied to computer program, such as computer program instructions, when its quilt When computer performs, by the operation of this computer, can call or provide the method according to the invention and/or technical scheme. And call the programmed instruction of the method for the present invention, it is possibly stored in fixing or movably in record medium, and/or passes through Data stream in broadcast or other signal bearing medias and be transmitted, and/or be stored in and run according to described programmed instruction In the working storage of computer equipment.Here, include a device according to one embodiment of present invention, this device includes using In the memorizer of storage computer program instructions with for performing the processor of programmed instruction, wherein, when this computer program refers to When order is performed by this processor, trigger this plant running method based on aforementioned multiple embodiments according to the present invention and/or skill Art scheme.

It is obvious to a person skilled in the art that the invention is not restricted to the details of above-mentioned one exemplary embodiment, Er Qie In the case of the spirit or essential attributes of the present invention, it is possible to realize the present invention in other specific forms.Therefore, no matter From the point of view of which point, all should regard embodiment as exemplary, and be nonrestrictive, the scope of the present invention is by appended power Profit requires rather than described above limits, it is intended that all by fall in the implication of equivalency and scope of claim Change is included in the present invention.Should not be considered as limiting involved claim by any reference in claim.This Outward, it is clear that " including ", a word was not excluded for other unit or step, and odd number is not excluded for plural number.In device claim, statement is multiple Unit or device can also be realized by software or hardware by a unit or device.The first, the second word such as grade is used for table Show title, and be not offered as any specific order.

Claims

1. the method carrying out information recommendation in input method, wherein, the method comprises the following steps:

A determines one or more interest informations of user；

Method the most according to claim 1, wherein, described step a includes:

-according to corresponding to the application related information of the one or more application in equipment corresponding to described input method and/or user User related information, determine one or more interest informations of described user.

Method the most according to claim 2, wherein, described application related information includes following at least any one:

The application type information of-described application；

The application label information of-described application；

The application precedence information of-described application.

The most according to the method in any one of claims 1 to 3, wherein, described step b includes:

B1 when described user's input information, list entries based on described user, determine the input key word of described user；

B2, based on described input key word, in conjunction with described interest information, determines corresponding with described list entries or many Individual recommended links.

Method the most according to claim 4, wherein, described step b1 includes following at least any one:

-when described user's input information, list entries based on described user, determine the input key word of described user, its In, described input key word is the input high frequency words of described user, and described input high frequency words history based on described user inputs Sequence and determine；

-when described user's input information, list entries based on described user, determine the input key word of described user, its In, described input key word be described user browse high frequency words, described in browse high frequency words historical viewings based on described user Record and determine.

6. according to the method described in claim 4 or 5, wherein, described step b2 includes:

B21 is based on described input key word, in conjunction with described interest information, determines that the association corresponding with described input key word is emerging Interest information；

-according to described input key word and described association interest information, determine corresponding with described list entries or Multiple recommended links.

Method the most according to claim 6, wherein, described step b21 includes:

-based on literary composition information before and after corresponding to described input key word and described input key word, in conjunction with described interest information, Determine the association interest information corresponding with described input key word.

Method the most according to any one of claim 1 to 7, wherein, described step c includes:

-based on predetermined ordering rule, the one or more recommended links is ranked up, to determine preferred recommended links；

Method the most according to claim 8, wherein, described predetermined ordering rule includes following at least any one:

-ageing based on corresponding to described recommended links；

-based on the reliability corresponding to described recommended links；

10. carrying out a recommendation apparatus for information recommendation in input method, wherein, this device includes:

11. recommendation apparatus according to claim 10, wherein, described interest determine device for:

12. recommendation apparatus according to claim 11, wherein, described application related information includes following at least any one:

The application type information of-described application；

The application label information of-described application；

The application precedence information of-described application.

13. according to the recommendation apparatus according to any one of claim 10 to 12, and wherein, described link determines that device includes:

Key word determines unit, is used for when described user's input information, list entries based on described user, determines described use The input key word at family；

Recommend to determine unit, for based on described input key word, in conjunction with described interest information, determine and described list entries phase Corresponding one or more recommended links.

14. recommendation apparatus according to claim 13, wherein, described key word determines that unit is for following arbitrary :

15. according to the recommendation apparatus described in claim 13 or 14, and wherein, described recommendation determines that unit includes:

Association interest determines unit, for based on described input key word, in conjunction with described interest information, determines and closes with described input The association interest information that keyword is corresponding；

Association link determines unit, for according to described input key word and described association interest information, determines defeated with described Enter one or more recommended links that sequence is corresponding.

16. recommendation apparatus according to claim 15, wherein, described association interest determine unit for:

17. according to the recommendation apparatus according to any one of claim 10 to 16, wherein, described in present device for:

18. recommendation apparatus according to claim 17, wherein, described predetermined ordering rule includes following at least any one:

-ageing based on corresponding to described recommended links；

-based on the reliability corresponding to described recommended links；

19. 1 kinds of input equipments, including the recommendation apparatus as according to any one of claim 10 to 18.