Specific embodiment
The exemplary embodiment of the disclosure is more fully described below with reference to accompanying drawings.Although the disclosure is shown in accompanying drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure and should not be by embodiments set forth here
Limited.On the contrary, there is provided these embodiments are able to be best understood from the disclosure, and can be by the scope of the present disclosure
Complete conveys to those skilled in the art.
To solve the above problems, reach and provide a user with more accurately picture search result, also provide a user with regard to figure
The purpose of the extension information of piece content, embodiments provide it is a kind of by picture associate with structural data it is new
Inventive concept.Wherein structural data is the data that constituted of field through prescribed form, typically can be by being stored in
The corresponding data message of entry in encyclopaedia data base extracts processing and obtains, and makes full use of encyclopaedia database information amount big and information
The stronger characteristic of reliability, the structural data for obtaining is stored in encyclopaedia data base.Certainly discrete structure data also may be used
To be obtained by way of human-edited, but rely solely on human-edited cannot meet magnanimity structural data generate and its
The needs that field updates.
Conceived based on foregoing invention, embodiments provide the correlating method of a kind of structural data and picture.Fig. 1
Show the process chart of structural data according to an embodiment of the invention and the correlating method of picture.Referring to Fig. 1, should
Structural data at least includes step S102 to step S108 with the correlating method of picture.
Step S102, the character description information for obtaining picture, carry out meaning of one's words extension to character description information, after being expanded
Extended description information.
Step S104, extended description information is matched with the encyclopaedia data base of the structural data that is stored with, it is determined that with
The theme of extended description information matches.
Step S106, in a plurality of structural data that includes of theme of matching, select and extended description information correlation
More than at least one structural data of specified degree of association.
Step S108, by select at least one structural data and picture be associated.
In the embodiment of the present invention, to make picture be associated with structural data, the character description information of picture is obtained first,
And meaning of one's words extension is carried out to the character description information of picture be expanded description information.Extended description information covers word description letter
The content of breath simultaneously can will describe expanded range by meaning of one's words extension.In encyclopaedia data base, a large amount of structural datas are according to master
Topic carries out classification storage, includes a plurality of structural data in encyclopaedia data base under each theme.Using extended description information and hundred
Section data base matches, it is determined that the theme with extended description information match, and then it is determined that theme under select to retouch with extension
State some structural datas of information matches.As extended description information is obtained on the basis of character description information,
Therefore by determining the theme with extended description information match, it is equal to belonging to determination structural data corresponding with picture
Theme, it is this to determine that the matching way of theme ensure that the last accuracy for obtaining structural data first, when different pictures
Character description information occur to make a distinction by theme during identical situation, it is to avoid there is the pass of picture and structural data
Connection mistake.Degree of association is specified by setting, is selected to exceed with the degree of association of extended description information under the theme for matching and is specified phase
At least one structural data and the picture of Guan Du is associated, and ensure that the matching degree of structural data and picture, and
Structural data as much as possible is associated with for picture.As can be seen here, cannot provide the user relative to prior art more accurate
Search Results, cannot also obtain the problem of the extension information with regard to image content, using the affiliated party in the embodiment of the present invention
Method can reach the purpose that structural data is associated with picture, can carry out standard to picture based on associated structural data
True identification such that it is able to provide the user more accurately Search Results, moreover it is possible to provide the user the extension with regard to image content
Information.
To make the embodiment of the present invention easy to understand, now illustrate and which is specifically described.It is little as Liu star with a personage
Bright picture citing, image content be Liu little Ming and magnify it is bright together attend Asia film festival awards ceremony, to make user in inspection
During the picture of Suo star Liu little Ming, this pictures can be accurately retrieved, moreover it is possible to obtain the information with regard to star Liu little Ming, should
It is that structural data by the picture with star Liu little Ming in encyclopaedia data base is associated with the purpose of the embodiment of the present invention.Close
The detailed process of connection is as follows.First obtain picture character description information, character description information be " Liu little Ming and magnify it is bright together
Give out an award for Asia film festival ", there are many different famous persons for being Liu little Ming in encyclopaedia data base, have prominent professors Liu little
Bright, star Liu little Ming etc., has many structural datas with regard to each Liu little Ming, such as with regard to the structuring number of star Liu little Ming
According to including:The data such as age of star Liu little Ming, birthplace, his nearest news, activity and films and television programs, musical works
One or the combination of several in information;Structural data with regard to teaching Liu little Ming includes:The age of professor Liu little Ming, work
The combination of or several in school, teaching experience and obtained honor.It is possible thereby to find, due to existing in encyclopaedia data base
Several different identities but name identical Liu Xiao Ming, have many structural datas for each Liu little Ming, therefore only lead to
The character description information for crossing picture cannot determine that picture should be associated with the structural data of which Liu little Ming.
It is the structural data that star Liu little Ming is filtered out in the structural data of the different Liu little Ming of several identity,
And which is associated with picture, need to be analyzed the structural data in encyclopaedia data base.Structural data be through
The data constituted by the field of prescribed form, typically can be by the corresponding data letter of entry to being stored in encyclopaedia data base
Breath extracts processing and obtains.
The extraction course of processing of data message can adopt various ways, such as extract whole entries in encyclopaedia data base first
Corresponding all of data message, the corresponding data message of each entry are typically a title, then correspond to a segment description
Information.Then each word in all data messages for obtaining to extraction is carried out respectively based on TF-IDF (term
Frequency-inverse document frequency, word frequency-anti-document frequency) algorithm weight computing.Based on TF-
The number of times that the weight computing of certain word of IDF algorithms can occur divided by the word with the number of total vocabulary in all data messages come
Realize.Therefore when one timing of number of total vocabulary in all data messages, the word just relatively more for natural occurrence number, such as
The words without essential meaning such as ", ", its weights are relatively small.By weight computing can exclude natural occurrence number compared with
The relatively small word without essential meaning of many, weights, so as to filter out the word for having essential meaning in data message.Finally to weights
Data message after calculating carries out a series of process of preset rules, can be format analysis processing to each field etc..So far
The extraction processing of data message is completed, structural data has been obtained.
Encyclopaedia data base has carried out theme division to structural data, comprising a large amount of structurings with theme under each theme
Data.In encyclopaedia data base to the determination of theme based on Term co-occurrence calculating.Term co-occurrence refers to several words and goes out jointly
It is existing.If several words Jing often occurs jointly, in can occurring in jointly in short, or in a paragragh, then recognize
To be to be mutually related between the meaning of one's words of these words.Such as " 360 ", " security guard ", Jing is normal in " computer health check-up " these three words
In occurring in jointly in short, it is taken as that the meaning of one's words between these three words is association.Term co-occurrence rate refers to several words
The common probability for occurring.The co-occurrence rate of several words is higher, then between the meaning of one's words association it is closer.Therefore can be by word altogether
Now several higher words of rate are divided into a theme.PLSA (Probability Latent Semantic can be utilized
Analysis, probability latent semantic analysis) topic model, calculated by Term co-occurrence and matrix equation is to the institute in structural data
There is word to carry out calculating analysis, several co-occurrence rate higher words are divided into into a theme, while obtaining each word in theme
The probability of appearance, a word can repeat in each theme.PLSA topic models are such as utilized, by " 360 ", " is defended safely
Scholar ", " computer health check-up " three words are divided in a theme, and the probability that " 360 " occur in the theme is 0.6777, meanwhile,
PLSA topic models are also divided in " 360 " and " Internet enterprises " two words in one theme, and in the theme, " 360 " occur
Probability is 0.553.Word under same theme can be ranked up according to certain rule, can according to word probability of occurrence by
It is high to Low to be ranked up, it is also possible to be ranked up according to Else Rule.
Acquisition and storage method based on above structural data, encyclopaedia data base is by the structuring with regard to star Liu little Ming
Under entertainment theme, the structural data with regard to teaching Liu little Ming is stored under educational topics data storage.Based on this point of master
The storage mode of topic, in order to filter out the structural data of star Liu little Ming, can determine the corresponding structuring number of picture first
According to affiliated theme, the structural data of needs is then selected under affiliated theme, so as to avoid due to originals such as personage's duplications of name
Structural data because making picture associated errors.In order to determine the theme belonging to structural data corresponding with picture, can be right
The character description information of picture carries out meaning of one's words extension, obtains covering the content of character description information and than character description information model
Enclose wider array of extended description information, matched with encyclopaedia data base using extended description information, it is determined that with extended description information phase
The theme of matching, so that it is determined that the theme belonging to structural data corresponding with picture.
The character description information " Liu little Ming is all the prize-giving of Asia film festival with magnifying bright one " of picture in this example is carried out into language
Meaning extension, is famous video display director due to magnifying bright, therefore by being extended to name " magnifying bright ", can obtain " director,
A series of relevant words such as video display, amusement "." Asia film festival " is extended, " film, activity, amusement, shadow can be obtained
Depending on " etc. a series of words.The word obtained after being extended by " magnifying bright ", " Asia film festival " collectively constitutes extended description information.
Above-mentioned extended description information is carried out matching with the classification of theme in encyclopaedia data base and be can determine that belonging to above-mentioned extended description information
Theme be entertainment theme, therefore structural data corresponding with " Liu little Ming " in the picture is stored under entertainment theme.
In this example, under entertainment theme, there is the structural data of many amusement classes, being can be in substantial amounts of structural data
In get the data higher with extended description information correlation, so as to provide accurate data message for picture, setting is specified
Standard of the degree of association as mating structure data.Extended description information is calculated in this example related to a certain bar structural data
When spending, the common factor and union of extended description information and the structural data are taken first, and common factor is that the neutralization of extended description information should
The set of the word all included in bar structural data, union are the institutes occurred during extended description information neutralizes the structural data
The set of some words.In common factor the number of word with and concentrate word number ratio, be exactly extended description information and the structuring
The degree of association of data.In common factor the number of word with and concentrate word number ratio it is bigger, illustrate extended description information and this tie
Structure Data mutuality degree is higher.
Degree of association specific formula for calculation isWherein SijRepresent the degree of association of text i and text j, DiRepresent text
Contained word, D in this ijContained word, D in expression text ji∩DjRepresent text i and text j common factor, i.e. text i and
The set of the word all included in text j, Di∪DjRepresent the union of text i and text j, the i.e. institute occurred in text i and text j
The set of some words.
In embodiments of the present invention, in order to the structural data with regard to Liu little Ming is got from amusement class theme, set
Specified degree of association is 70%.For the structural data unrelated with Liu little Ming under entertainment theme with regard to other stars, can be with fast
Speed is judged to be less than 70% with extended description information correlation.In structural data of the degree of association more than 70%, select at least
The structural data of one Liu little Ming is associated with picture, so as to provide the user the information with regard to Liu little Ming.Based on this example
In extended description information, the structural data of the Liu little Ming selected in this example can be that " Liu little Ming serves as Asia in continuous 5 years
The invited guests for granting awards of film festival, once self-destruction want to change one's profession to work as director.”
In this example, using the correlating method of the structural data in the embodiment of the present invention and picture, due to first with
Extended description information determines the theme of matching, i.e., first determine entertainment theme corresponding with Liu little Ming, then carry out from entertainment theme
The selection of structural data, ensure that the accuracy of the structural data of last acquisition, it is to avoid picture is with professor Liu little Ming's
Structural data is interrelated.Setting is specified degree of association and selects to be exceeded with extended description information correlation and specifies degree of association extremely
A few structural data is associated with picture, ensure that the extension provided the user accurately with regard to image content is believed
Breath.
For picture relational structure data when, it sometimes appear that the situation of incidence relation cannot be set up.Such as picture
The picture of theme entitled " Liu little Ming and Wang great Chuan together attends a banquet ", this text description information of correspondence subject occur in that " Liu
The word such as Xiao Ming ", " Wang great Chuan ", " attending a banquet ", by extensions such as " Wang great Chuan " extensions available " famous enterpriser, economist "
Information, so as to judge to belong to economic field, and is extended to " attending a banquet " and cannot determine subject fields, and several in data base
There is no the situation with economic field or type matching in " Liu little Ming ", even if therefore now cannot determine text by extension yet
Theme belonging to this description information, so as to cannot also set up this picture with the structural data of some Liu little Ming in data base
Incidence relation.
When selecting structural data under theme, it is understood that there may be have some degree of association to exceed the structuring for specifying degree of association
Data, and these structural datas have a case that identical degree of association.In another embodiment, it is same to provide
One picture with regard to star Liu little Ming, character description information is " Liu little Ming makes a film in Heng Dian ", by believing to word description
Breath is extended and obtains extended description information such as " film, amusement, star, shootings ", and these extended description information can be with encyclopaedia
Entertainment theme in data base is matched, and retrieves many structural datas with regard to Liu little Ming, such as under entertainment theme
" Liu little Ming, famous film performer are good at shooting acrobatic fighting play " was with " Liu little Ming, once shot 50 multi-section films, newly played just in horizontal stroke
Shop start shooting " etc., the structural data that these structural datas are all with extended description information correlation is 80%, if at this moment this
Type of structured data quantity is few, meets the quantitative requirement being associated with picture simultaneously, can choose simultaneously and picture
Matched.If all very high structural data quantity of this degree of association is big, such as up to thousands of, it is impossible to while and picture
It is associated, can therefrom selects content newest, information most comprehensive some is matched with picture.When picture with it is a plurality of
When structural data is matched, structural data can be ranked up according to degree of association grade.
In some cases, when scanning under theme, lifting the degree of association for setting may be too high, causes retrieval
Less than suitable structural data.Such as under certain theme, specified degree of association is set to into 90%, by retrieval it may be found that looking for not
To the structural data with extended description information matches, or the energy mating structure data for retrieving are seldom, and retrieving
Middle discovery is many with the structural data that the degree of association of extended description information is 80%, and quantity of information, at this moment can be automatic than larger
It is 80% to reduce degree of association, so as to provide the user the extension information with regard to image content as much as possible.
The bar number of the numerical value of the degree of association in above-mentioned instantiation and the structural data with picture match selected
It is only to enumerate, the present invention is not caused to limit.In the specific implementation, the numerical value of degree of association and select and picture
Depending on the bar number of the structural data of matching is by concrete condition.
Another preferred embodiment of the present invention additionally provides a picture for showing white clouds and the earth, the word description of picture
Information is " the earth, white clouds ", is that the structural data of the earth is carried out with picture using the purpose of the correlating method shown in Fig. 1
Association.Structural data of the several bar with regard to the earth is there may be in encyclopaedia data base, it may be possible to the earth in natural landscape
Structural data, it is also possible to the structural data of the song of certain entitled the earth, if directly with character description information with
Encyclopaedia data base match, it is impossible to it is determined that any bar structural data is associated, while association occurs mistake certainly.At this moment to picture
Character description information " the earth, white clouds " be extended, can determine that the theme for matching is nature by extended description information
Landscape, so as to the structural data for selecting to match under natural landscape theme, finally can by natural landscape with regard to big
Ground structure data are associated with picture, it is ensured that picture is associated with accurate structural data.
From foregoing invention embodiment, using the correlating method in the embodiment of the present invention can by picture with it is corresponding
The means that structural data is mutually matched, reach the purpose of structural data and picture accurate correlation, based on associated structure
Change data can accurately recognize picture such that it is able to provide the user more accurately Search Results, moreover it is possible to for user
Extension information with regard to image content is provided.
In a preferred embodiment of the invention, mark of the character description information of picture at least including the picture place page
One or more in text, the anchor texts (link Anchor Text) and the title of picture of picture around topic, picture.
After picture is got, can be by the text around the title of the picture place page, picture, the anchor of picture
The title of one or more acquisition picture in text (link Anchor Text) and the title of picture.Obtain many pictures as far as possible to retouch
State information to be easier after expansion to be matched with the theme in encyclopaedia data base.
When being extended to character description information, it is found that the frequency that some words occur is very high, but but does not have
Essential meaning, this kind of word are referred to as stop-word.The presence of stop-word does not have help for meaning of one's words extension, therefore word description is being believed
When breath is extended, first character description information is analyzed, these stop-words without essential meaning is deleted, it is ensured that leave one
A bit for the significant word of matching operation.It is that common stop-word has, etc. word.That stop-word is included but are not limited to,
, etc. word, also including a large amount of insignificant semantic words.The more words of occurrence number generally in encyclopaedia data base, in extension
When will be considered which closer to meaningless, therefore will be considered that and be off word and deleted.Meaning of one's words expansion is carried out to the part after deletion
Exhibition, increased the accuracy of meaning of one's words extension.By the weight computing based on TF-IDF algorithms is carried out to character description information, can
Exclude the stop-word nonsensical for extension.Stop-word is also referred to as stop words.
In another preferred embodiment of the invention, " Liu little Ming and magnify bright one for what is mentioned in the above example
Be all Asia film festival prize-giving " this character description information, " and ", " being " be exactly stop-word, when being extended, first to which
Deleted.Again parsing sentence " Liu little Ming magnify bright together Asia film festival prize-giving " it is found that " together " and " prize-giving " though
Be not so meaningless word, be not belonging to stop word category, but help do not played to matching theme yet, therefore stop deleting
After word, also tackle character description information and be analyzed, extract and then can be expanded again as the noun of search key
Exhibition.It is found that effective noun is " magnifying bright ", " Asia film festival " in this example, therefore it is extended against two words,
It is easy for be matched with the entertainment theme in encyclopaedia data base.
Using stop-word is deleted from character description information in the embodiment of the present invention, keyword is extracted, then carries out meaning of one's words expansion
The method of exhibition is extended to character description information, it is ensured that the effectiveness of extension, can in encyclopaedia data base accurate match
To suitable theme.
Based on same inventive concept, the embodiment of the present invention additionally provides a kind of side of generating structure data search result items
Method.Fig. 2 shows the process chart of the method for generating structure data search result items according to an embodiment of the invention.
As shown in Fig. 2 the method for generating structure data search result items at least includes step S202 to S204.
S202, the picture for obtaining search query word Corresponding matching.
S204, the structural data being endowed according to picture, generate search result items.
In the embodiment of the present invention, after query word is got, the picture matched with query word can be obtained automatically, and is schemed
Piece is associated with the correlating method of picture with corresponding structural data by the structural data shown in Fig. 1, then
Search result items are generated with the structural data for associating according to picture.The search result items of generation can be one with query word phase
The picture of matching, and picture and corresponding structural data linked, and can be jumped to accordingly by clicking on picture
The structural data page, or one include the webpage of picture and corresponding structural data, search result items it is concrete
Embodiment is determined on a case-by-case basis.
By the method for the generating structure data search result items in the embodiment of the present invention, generated according to query word and searched for
Result items, can provide the user more accurately Search Results, moreover it is possible to provide the user the extension information with regard to image content.
As a example by searching for the picture of Liu little Ming, using the generating structure data search result items in the embodiment of the present invention
Method, after keyword " Liu little Ming " is got, can obtain the picture matched with " Liu little Ming " automatically, and picture has passed through figure
Structural data shown in 1 is associated with the correlating method of picture with corresponding structural data, according to picture with associate
Structural data generate search result items.Search result items can be a webpage comprising Liu little Ming pictures, while webpage
Automatic jump to by structural data of the AutoLink with regard to Liu little Ming, the picture for clicking on the Liu little Ming for searching little comprising Liu
The page of bright structural data, search result items can also be a webpages comprising Liu little Ming pictures, also provide in webpage
Structural data corresponding with Liu little Ming.Simply schematically illustrate herein, search result items can also show otherwise
To user.
Correspondingly, user can not only get the picture of Liu little Ming, moreover it is possible to get Liu little Ming when " Liu little Ming " is searched for
Structural data.It is suitability as clicked on picture that the user be given in this example obtains the method for the structural data of Liu little Ming
Citing, to the present invention do not cause limit.In the specific implementation, concrete grammar is determined on a case-by-case basis.
The structural data illustrated in correspondence Fig. 1 and the correlating method of picture, Fig. 3 provide structural data with picture
The structural representation of associated apparatus.As shown in figure 3, the embodiment of the present invention a kind of structural data and picture be provided associate dress
Put, including:
Expansion module 310, is configured to obtain the character description information of picture, carries out meaning of one's words extension to character description information,
Extended description information after being expanded.
Matching module 320, is configured to carry out the encyclopaedia data base of extended description information and the structural data that is stored with
Match somebody with somebody, it is determined that the theme with extended description information matches.
Selecting module 330, is configured in a plurality of structural data that the theme of matching includes, selects to believe with extended description
Breath degree of association exceedes at least one structural data for specifying degree of association.
Relating module 340, is configured at least one structural data for selecting is associated with picture.
In the embodiment of the present invention, to make picture match with structural data, the text of picture is obtained using expansion module 310
WD information, then and to the character description information of picture carries out meaning of one's words extension and is expanded description information, extended description letter
Breath is covered the content of character description information and is expanded the scope of description by conjunctive word.In encyclopaedia data base, a large amount of structurings
Data carry out classification storage according to theme, include a plurality of structural data in encyclopaedia data base under each theme.By extended description
Information is matched with encyclopaedia data base, it is determined that the theme being associated with extended description information, and then it is determined that theme under select
With some structural datas of extended description information matches.This utilization matching module 320 determines the match party of theme first
Formula ensure that the last accuracy for obtaining structural data, can be with when identical situation occurs in the character description information of different pictures
Made a distinction by theme, it is to avoid the associated errors of picture and structural data occur.Set in selecting module 330 and specify phase
Guan Du, selects to exceed at least one structural data for specifying degree of association and figure under the theme for matching using selecting module 330
Piece is associated, and ensure that the matching degree of structural data and extended description information, and is associated with for picture as many as possible
Structural data.As can be seen here, more accurately Search Results cannot be provided the user relative to prior art, cannot be also obtained
With regard to the problem of the extension information of image content, using the associated apparatus in the embodiment of the present invention can by picture with it is corresponding
The means that structural data is mutually matched, reach the purpose that structural data is associated with picture, based on associated structuring
Data can be accurately recognized to picture such that it is able to provide the user more accurately Search Results, moreover it is possible to carry for user
For the extension information with regard to image content.
In another embodiment of the present invention, expansion module 310 also needs to be analyzed character description information, deletes
Again insignificant stop-word wherein more than occurrence number, stop-word including but not limited to,, also including a large amount of insignificant
Semantic word.
In a preferred embodiment of the invention, mark of the character description information of picture at least including the picture place page
One or more in text, the anchor texts (link Anchor Text) and the title of picture of picture around topic, picture.Its
In, the title of picture at least one from above-mentioned text obtain.
The method of the generating structure data search result items illustrated in correspondence Fig. 2, Fig. 4 are shown according to of the invention one
The structural representation of the device of the generating structure data search result items of embodiment.As shown in figure 4, the embodiment of the present invention is provided
A kind of device of generating structure data search result items, including:
Acquisition module 410, is configured to obtain the corresponding picture of search query word.
Generation module 420, is configured to the structural data associated according to picture and picture, generates search results pages.
In the embodiment of the present invention, after the query word of search pictures, can be obtained automatically and query word by acquisition module 410
The picture for matching, picture are associated with corresponding structural data by the method shown in Fig. 1.Using generation mould
Block 420 generates result of page searching with the structural data for associating according to picture.Result of page searching can be one comprising searching
The webpage of the picture that rope is arrived, while one webpage containing corresponding structural data of webpage AutoLink, by webpage clicking
Mode can redirect, or a webpage comprising picture and corresponding structural data, search result items it is concrete
Embodiment is determined on a case-by-case basis.
By the device of the generating structure data search result items in the embodiment of the present invention, generated according to query word and searched for
Result items, can provide the user accurate Search Results, moreover it is possible to provide the user the extension information with regard to image content.
In a further advantageous embodiment, user is using the generating structure data search result items in the embodiment of the present invention
Device search for " Liu little Ming " when, can not only obtain the picture of Liu little Ming, by click on picture can also obtain Liu little Ming's
Structural data.It is suitability as clicked on picture that the user be given in this example obtains the method for the structural data of Liu little Ming
Citing, does not cause to limit to the present invention.In the specific implementation, concrete grammar is determined on a case-by-case basis.
Further, summary content, Fig. 5 show that generating structure data according to an embodiment of the invention are searched
The structural representation of the system of rope result items.The system of the generating structure data search result items includes:
Encyclopaedia data base 510, is configured to include multiple themes, each theme includes a plurality of structural data.
Picture database 520, is configured to the plurality of pictures that is stored with;Meaning of one's words expansion is carried out to the character description information of every pictures
Exhibition, the extended description information after being expanded;And, according to the description information after extension and encyclopaedia database matching, association should
Picture with match at least one structural data that obtains.
User terminal 530, is configured to be input into the search query word of picture.
Search engine 540, is configured to the corresponding picture of the search query word is searched for and obtained in picture database, with
And, the structural data associated with picture is searched for and obtained in encyclopaedia data base, with reference to obtain the picture and with figure
The related information of piece association, generates search results pages.
Using the system of the generating structure data search result items provided in the embodiment of the present invention, can be to image data
The character description information of the every pictures in storehouse 520 carries out meaning of one's words extension, and be expanded description information, believes further according to extended description
Breath is matched with encyclopaedia data base 510, determines the theme with extended description information match in encyclopaedia data base 510, then from correspondingly
Theme in select with higher at least one structural data more than specified threshold of extended description information correlation, be associated with
On corresponding picture.Wherein encyclopaedia data base 510 includes multiple themes, and each theme includes a plurality of structural data.So far
The process that picture is associated with structural data is completed just.
System in the embodiment of the present invention has user terminal 530, and user can utilize user terminal 530 to be input into picture
Search query word.System in the embodiment of the present invention also has search engine 540, and search engine 540 can be according to picture query
Word is searched in picture database 520 and obtains corresponding picture, moreover it is possible to got in encyclopaedia data base 510 relative with picture
The structural data answered, according to picture and the structural data for associating, can generate search results pages.
Using the system of the generating structure data search result items provided in the embodiment of the present invention, picture can with it is corresponding
Structural data be associated, when user searches sketch map piece, provide the user accurate Search Results, additionally it is possible to provide with regard to
The extension information of image content.
From the foregoing, it will be observed that in one embodiment of the invention, there is provided the correlating method of a kind of structural data and picture.To make
Picture is associated with structural data, obtains the character description information of picture first, and the character description information to picture is carried out
The meaning of one's words extends the description information that is expanded.Extended description information is covered the content of character description information and can be extended by the meaning of one's words
Expanded range will be described.In encyclopaedia data base, a large amount of structural datas carry out classification storage according to theme, in encyclopaedia data base
A plurality of structural data is included under each theme.Matched with encyclopaedia data base using extended description information, it is determined that retouching with extension
State the theme of information match, so it is determined that theme under select with extended description information matches some structuring numbers
According to.As extended description information is obtained on the basis of character description information, therefore by determining and extended description information
The theme for matching, is equal to the theme determined belonging to corresponding with picture structural data, this theme of determination first
The last accuracy for obtaining structural data be ensure that with mode, when identical situation occurs in the character description information of different pictures
When can be made a distinction by theme, it is to avoid there are the associated errors of picture and structural data.Degree of association is specified by setting,
Exceed with the degree of association of extended description information under the theme that matches of selection specify at least one structural data of degree of association with
Picture is associated, and ensure that the matching degree of structural data and picture, and is associated with structure as much as possible for picture
Change data.As can be seen here, more accurately Search Results cannot be provided the user relative to prior art, cannot be also obtained with regard to figure
The problem of the extension information of piece content, can reach structural data with picture phase using the correlating method in the embodiment of the present invention
The purpose of association, can accurately be recognized to picture based on associated structural data such that it is able to provide the user
More accurately Search Results, moreover it is possible to provide the user the extension information with regard to image content.
Correspondence said structure data and the correlating method of picture, additionally provide a kind of knot in another embodiment of the present invention
Structure data and the associated apparatus of picture, cannot provide the user more accurately Search Results relative to prior art, also cannot
Obtain the problem of the extension information with regard to image content, using the associated apparatus in the embodiment of the present invention can by picture with it is right
The means that the structural data answered is mutually matched, reach the purpose that structural data is associated with picture, based on associated knot
Structure data can be accurately recognized to picture such that it is able to provide the user more accurately Search Results, moreover it is possible to be use
Family provides the extension information with regard to image content.
Another embodiment of the present invention additionally provides a kind of method of generating structure data search result items, being capable of basis
Query word generates search result items such that it is able to provide the user more accurately Search Results, moreover it is possible to provide the user with regard to figure
The extension information of piece content.
The method of the above-mentioned generating structure data search result items of correspondence, additionally provides a kind of generation in another embodiment
The device of structured data searching result items, can generate search result items according to query word such that it is able to provide the user more
Accurate Search Results, moreover it is possible to provide the user the extension information with regard to image content.
Using the system of the generating structure data search result items provided in another embodiment of the present invention, picture can
It is associated with corresponding structural data, when user searches sketch map piece, provides the user accurate Search Results, additionally it is possible to carry
For the extension information with regard to image content.
In description mentioned herein, a large amount of details are illustrated.It is to be appreciated, however, that the enforcement of the present invention
Example can be put into practice in the case where not having these details.In some instances, known method, structure is not been shown in detail
And technology, so as not to obscure the understanding of this description.
Similarly, it will be appreciated that in order to simplify the disclosure and help understand one or more in each inventive aspect, exist
Above to, in the description of the exemplary embodiment of the present invention, each feature of the present invention is grouped together into single enforcement sometimes
In example, figure or descriptions thereof.However, should the method for the disclosure be construed to reflect following intention:I.e. required guarantor
The more features of feature is expressly recited in each claim by the application claims ratio of shield.More precisely, such as following
Claims it is reflected as, inventive aspect is less than all features of single embodiment disclosed above.Therefore,
Thus the claims for following specific embodiment are expressly incorporated in the specific embodiment, wherein each claim itself
All as the separate embodiments of the present invention.
Those skilled in the art are appreciated that can be carried out adaptively to the module in the equipment in embodiment
Change and they are arranged in one or more different from embodiment equipment.Can be the module or list in embodiment
Unit or component are combined into a module or unit or component, and can be divided in addition multiple submodule or subelement or
Sub-component.In addition at least some in such feature and/or process or unit is excluded each other, can adopt any
Combine to all features disclosed in this specification (including adjoint claim, summary and accompanying drawing) and so disclosed
Where all processes or unit of method or equipment are combined.Unless expressly stated otherwise, this specification (includes adjoint power
Profit is required, summary and accompanying drawing) disclosed in each feature can it is identical by offers, be equal to or the alternative features of similar purpose carry out generation
Replace.
Although additionally, it will be appreciated by those of skill in the art that some embodiments described herein include other embodiments
In some included features rather than further feature, but the combination of the feature of different embodiments means in of the invention
Within the scope of and form different embodiments.For example, in detail in the claims, embodiment required for protection one of arbitrarily
Can in any combination mode using.
The present invention all parts embodiment can be realized with hardware, or with one or more processor operation
Software module realize, or with combinations thereof realize.It will be understood by those of skill in the art that can use in practice
Microprocessor or digital signal processor (DSP) are realizing the pass of middle structural data according to embodiments of the present invention and picture
The some or all functions of some or all parts of coupling device.The present invention is also implemented as being retouched for performing here
Some or all equipment of the method stated or program of device (for example, computer program and computer program).
Such program for realizing the present invention can be stored on a computer-readable medium, or can have one or more signal
Form.Such signal can be downloaded from internet website and be obtained, or on carrier signal provide, or with it is any its
He provides form.
It should be noted that above-described embodiment the present invention will be described rather than limits the invention, and ability
Field technique personnel can design alternative embodiment without departing from the scope of the appended claims.In the claims,
Any reference markss between bracket should not be configured to limitations on claims.Word "comprising" is not excluded the presence of not
Element listed in the claims or step.Word "a" or "an" before element does not exclude the presence of multiple such
Element.The present invention can come real by means of the hardware for including some different elements and by means of properly programmed computer
It is existing.If in the unit claim for listing equipment for drying, several in these devices can be by same hardware branch
To embody.The use of word first, second, and third does not indicate that any order.These words can be explained and be run after fame
Claim.
So far, although those skilled in the art will appreciate that detailed herein illustrate and describe multiple showing for the present invention
Example property embodiment, but, without departing from the spirit and scope of the present invention, still can be direct according to present disclosure
It is determined that or deriving many other variations or modifications for meeting the principle of the invention.Therefore, the scope of the present invention is understood that and recognizes
It is set to and covers all these other variations or modifications.
The embodiment of the invention also discloses the correlating method of A1, a kind of structural data and picture, including:
The character description information of picture is obtained, meaning of one's words extension is carried out to the character description information, the expansion after being expanded
Exhibition description information;
The extended description information is matched with the encyclopaedia data base of the structural data that is stored with, it is determined that with the expansion
The theme of exhibition description information matching;
In a plurality of structural data that the theme of matching includes, select to exceed finger with the extended description information correlation
Determine at least one structural data of degree of association;
At least one structural data for selecting is associated with the picture.
A2, the method according to A1, wherein, meaning of one's words extension is carried out to the character description information, including:
Respectively the character description information is analyzed, stop-word therein is deleted, wherein, the stop-word is to point out
Now repeatedly but the word without essential meaning;
Meaning of one's words extension is carried out to the part after deletion.
A3, the method according to any one of A1 to A2, wherein, the stop-word includes at least one following:,
Ground.
A4, the method according to any one of A1 to A3, wherein, the character description information of the picture, including it is following extremely
It is one of few:
The title of the picture place page;
Text around picture;
The anchor texts of picture;
The title of picture, wherein, the title of the picture is obtained by least one above-mentioned text.
B5, a kind of method of generating structure data search result items, including:
Obtain the picture of search query word Corresponding matching;
According to the structural data that the picture is endowed, search result items are generated.
The associated apparatus of C6, a kind of structural data and picture, including:
Expansion module, is configured to obtain the character description information of picture, carries out meaning of one's words extension to the character description information,
Extended description information after being expanded;
Matching module, is configured to carry out the encyclopaedia data base of the extended description information and the structural data that is stored with
Match somebody with somebody, it is determined that the theme with the extended description information matches;
Selecting module, is configured in a plurality of structural data that the theme of matching includes, selects and the extended description
Information correlation exceedes at least one structural data for specifying degree of association;
Relating module, is configured at least one structural data for selecting is associated with the picture.
C7, the device according to C6, wherein, the expansion module is additionally configured to:
The character description information is analyzed, stop-word therein is deleted, wherein, the stop-word refers to that appearance is more
The secondary but word without essential meaning;
Meaning of one's words extension is carried out to the part after deletion.
C8, the device according to C7, wherein, the stop-word includes at least one following:,.
C9, the device according to any one of C6 to C8, wherein, the character description information of the picture, including it is following extremely
It is one of few:
The title of the picture place page;
Text around picture;
The anchor texts of picture;
The title of picture, wherein, the title of the picture is obtained by least one above-mentioned text.
D10, a kind of device of generating structure data search result items, including:
Acquisition module, is configured to obtain the corresponding picture of search query word;
Generation module, is configured to the structural data associated according to the picture and the picture, generates Search Results
Page.
E11, a kind of system of generating structure data search result items, including:
Encyclopaedia data base, is configured to include multiple themes, each theme includes a plurality of structural data;
Picture database, is configured to the plurality of pictures that is stored with;Meaning of one's words extension is carried out to the character description information of every pictures,
Extended description information after being expanded;And, according to the description information after the extension and encyclopaedia database matching, association should
Picture with match at least one structural data that obtains;
User terminal, is configured to be input into the search query word of picture;
Search engine, is configured to the corresponding picture of the search query word is searched for and obtained in the picture database,
And, the structural data associated with the picture is searched for and is obtained in the encyclopaedia data base, with reference to the figure for obtaining
Piece and the related information associated with the picture, generate search results pages.