CN106557554A - Display packing and device based on the Search Results of artificial intelligence - Google Patents
Display packing and device based on the Search Results of artificial intelligence Download PDFInfo
- Publication number
- CN106557554A CN106557554A CN201610974823.3A CN201610974823A CN106557554A CN 106557554 A CN106557554 A CN 106557554A CN 201610974823 A CN201610974823 A CN 201610974823A CN 106557554 A CN106557554 A CN 106557554A
- Authority
- CN
- China
- Prior art keywords
- participle
- search results
- degree
- search
- word
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/9038—Presentation of query results
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention proposes a kind of display packing and device of Search Results based on artificial intelligence, wherein, the display packing of the Search Results includes：Search Results are obtained according to the search word of input；Using the attention model of training in advance, the degree of correlation of each participle and search word in Search Results is obtained；Degree is highlighted according to what the degree of correlation of each participle and search word determined each participle in Search Results；Each participle in Search Results is highlighted according to the degree that highlights of each participle.The display packing of the Search Results based on artificial intelligence of the present invention, program is highlighted with difference in search results pages and is highlighted each participle in Search Results, the search results pages with stereovision are provided the user, has facilitated user that the degree of correlation that degree understands the participle in Search Results and search word is highlighted according to difference.
Description
Technical field
The present invention relates to natural language processing technique field, more particularly to a kind of Search Results based on artificial intelligence it is aobvious
Show method and apparatus.
Background technology
Artificial intelligence (Artificial Intelligence), english abbreviation is AI.It is study, be developed for simulation,
Extend and extend a new science of technology of theory, method, technology and the application system of the intelligence of people.Artificial intelligence is to calculate
One branch of machine science, it attempts the essence for understanding intelligence, and produce it is a kind of it is new can be in the way of human intelligence be similar
The intelligent machine made a response, the research in the field include robot, speech recognition, image recognition, natural language processing and specially
Family's system etc..
At present, with the fast development of Internet technology, search engine has become the important tool that people obtain information.
When user is input into search word (query) in search engine, search engine can be rapidly returned to the corresponding Search Results of the search word
Page.In addition, user navigates to rapidly wanted result in search results pages for convenience, can be to using in search results pages
The key word or near synonym of family input is highlighted, and the display effect of the word in other search results pages is the same.However,
Generally some of search results pages word can also reflect that user wants the content searched for, directly aobvious with highlighted or not highlighted form
Show, the single no stereovision of search results pages, it is impossible to embody the degree of correlation of search results pages and search word.
The content of the invention
It is contemplated that at least solving above-mentioned technical problem to a certain extent.
For this purpose, first purpose of the present invention is to propose a kind of display packing of the Search Results based on artificial intelligence,
Program is highlighted with difference in search results pages and is highlighted each participle in Search Results, provided the user and had
The search results pages of stereovision, facilitate user to highlight degree according to difference and understand participle and search word in Search Results
Degree of correlation.
Second object of the present invention is to propose a kind of display device of the Search Results based on artificial intelligence.
It is that up to above-mentioned purpose, embodiment proposes a kind of Search Results based on artificial intelligence according to a first aspect of the present invention
Display packing, comprise the following steps：Search Results are obtained according to the search word of input；Using the attention mould of training in advance
Type, obtains the degree of correlation of each participle and the search word in the Search Results；According to the phase of each participle and search word
What pass degree determined each participle in the Search Results highlights degree；Projected according to the degree that highlights of each participle
Show each participle in the Search Results.
The display packing of the Search Results based on artificial intelligence of the embodiment of the present invention, obtains in the search word according to input
After Search Results, by the attention model of training in advance, each participle and the search word in the Search Results is obtained
Degree of correlation, and highlighting for each participle in the Search Results is determined according to each participle and the degree of correlation of search word
Degree, and each participle in the Search Results is highlighted according to the degree that highlights of each participle.Thus, searching
Program is highlighted with difference in rope result page and is highlighted each participle in Search Results, provided the user and there is level
The search results pages of sense, facilitate the participle that user is highlighted according to difference in degree understanding Search Results related to search word
Degree.
Second aspect present invention embodiment proposes a kind of display device of the Search Results based on artificial intelligence, including：
First acquisition module, obtains Search Results for the search word according to input；Second acquisition module, for using training in advance
Attention model, obtains the degree of correlation of each participle and the search word in the Search Results；Determining module, for basis
What the degree of correlation of each participle and search word determined each participle in the Search Results highlights degree；Display module,
Each participle in the Search Results is highlighted for the degree that highlights according to each participle.
The display device of the Search Results based on artificial intelligence of the embodiment of the present invention, obtains in the search word according to input
After Search Results, by the attention model of training in advance, each participle and the search word in the Search Results is obtained
Degree of correlation, and highlighting for each participle in the Search Results is determined according to each participle and the degree of correlation of search word
Degree, and each participle in the Search Results is highlighted according to the degree that highlights of each participle.Thus, searching
Program is highlighted with difference in rope result page and is highlighted each participle in Search Results, provided the user and there is level
The search results pages of sense, facilitate the participle that user is highlighted according to difference in degree understanding Search Results related to search word
Degree.
The additional aspect and advantage of the present invention will be set forth in part in the description, and partly will become from the following description
Obtain substantially, or recognized by the practice of the present invention.
Description of the drawings
The above-mentioned and/or additional aspect and advantage of the present invention will become from the description with reference to accompanying drawings below to embodiment
It is substantially and easy to understand, wherein：
Fig. 1 is the flow chart of the display packing of the Search Results based on artificial intelligence according to one embodiment of the invention；
Fig. 2 is the refined flow chart of the S12 according to the step of one embodiment of the invention；
Fig. 3 is the flow process of the display packing of the Search Results based on artificial intelligence according to another embodiment of the present invention
Figure；
Fig. 4 is the structural representation of the display device of the Search Results based on artificial intelligence according to one embodiment of the invention
Figure；
Fig. 5 is that the structure of the display device of the Search Results based on artificial intelligence according to another embodiment of the present invention is shown
It is intended to；
Fig. 6 is that the structure of the display device of the Search Results based on artificial intelligence according to another embodiment of the invention is shown
It is intended to.
Specific embodiment
Embodiments of the invention are described below in detail, the example of the embodiment is shown in the drawings, wherein from start to finish
Same or similar label represents same or similar element or the element with same or like function.Below with reference to attached
The embodiment of figure description is exemplary, is only used for explaining the present invention, and is not considered as limiting the invention.
In describing the invention, it is to be understood that term " multiple " refers to two or more；Term " first ",
" second " is only used for describing purpose, and it is not intended that indicating or implying relative importance.
Below with reference to the accompanying drawings describe the Search Results based on artificial intelligence according to embodiments of the present invention display packing and
Device.
Fig. 1 is the flow chart of the display packing of the Search Results based on artificial intelligence according to one embodiment of the invention.
As shown in figure 1, the display packing of the Search Results based on artificial intelligence according to embodiments of the present invention, including it is following
Step.
S11, obtains Search Results according to the search word of input.
Specifically, when user is scanned for by search engine, search engine can provide search entrance, and by search
The searching request including search word of entrance receiving user's input.
Wherein, search word (i.e. query) can be the character (such as word, phonetic, symbol and/or numeral etc.) of various language
In one kind or combinations thereof.
Search engine obtains the multiple Search Results relevant with search word after the search word for receiving user input.
Wherein, Search Results include title and summary.
S12, using the attention model of training in advance, obtains the related journey of each participle and search word in Search Results
Degree.
In one embodiment of the invention, the thinning process of step S12, as shown in Fig. 2 may comprise steps of：
S121, obtains the first expression vector of each participle in Search Results using the attention model of training in advance.
Specifically, obtaining with after the Search Results of search word, for each Search Results, Search Results can carried out point
Word process, to generate the word sequence of Search Results, and ties according to the word order column-generation of default neutral net and Search Results search
First expression vector of each participle in fruit.
, wherein it is desired to be understood by, each word in the word sequence of Search Results is in predefined vocabulary
Individual word.
Wherein, it is, for example, RNN (Recurrent neural Network, Recognition with Recurrent Neural Network) net to preset neutral net
Network, LSTM networks (Long Short-Term Memory, long neutral net in short-term) etc..
As a kind of exemplary embodiment, it is assumed that default neutral net is RNN networks, can be by the word order of Search Results
The input layer that RNN networks are input into according to time serieses is arranged, then, the term vector layer of RNN networks is represented corresponding to each participle
Term vector.Wherein, term vector is the column vector that a length is EMBEDDING_SIZE.The hidden layer of RNN networks represents RNN nets
State of the network in each time point i, be a length be HIDDEN_SIZE column vector hi。
In one embodiment of the invention, in order to the memory of neutral net can be increased, can be according to default neutral net
It is vectorial with the first subrepresentation of each participle in the word order column-generation Search Results of Search Results, then, by the word of Search Results
The order of the word in sequence is overturned, to generate the first word sequence of Search Results, and according to the first word sequence and default god
Jing networks generate the second subrepresentation vector of each participle in Search Results, finally, sub according to the first subrepresentation vector sum second
Represent that vector generates the first expression vector of each participle of Search Results.
For example, it is assumed that Search Results are made up of T participle, are designated as：Text=(w1,…,wT), it is assumed that it is refreshing by RNN
In the word sequence of the forward direction of the hidden layer acquisition Search Results of Jing networks, the first subrepresentation vector of i-th participle is hi, pass through
In the reverse word sequence of the hidden layer acquisition Search Results of RNN neutral nets, the second subrepresentation vector of i-th participle is
rhi, the first expression vector H of i-th participlei=Concat (hi, rhi).Wherein, the expression sequence vector of Search Results is H1,
H2,…HT。
S122, obtains the second expression vector of search word using the attention model of training in advance.
, wherein it is desired to explanation, the execution of step S121 and S122 is in no particular order.
After search word is obtained, word segmentation processing can be carried out to search word, to generate the word sequence of search word, and according to default
Second expression vector of each participle in the word order column-generation search word of neutral net and search word.
Wherein, search word can be made up of at least one participle, when search word is made up of multiple participles, by default god
When Jing networks determine the second expression vector of search word, can be using the expression vector of last participle in search word as search word
The second expression vector, to calculate the overall degree of correlation of each participle and search word in Search Results.
In addition, in one embodiment of the invention, by default neutral net determine the second of search word represent to
During amount, it is also possible to which the expression vector of all participles of search word is added obtained expression vector as search word by element
Second represents vector.
According to the second expression vector sum first, S123, represents that vector obtains the phase of each participle and search word in Search Results
Pass degree.
Specifically, in search word is obtained after the first expression vector of the second expression vector sum search word of each participle,
The vectorial cosine value of the first expression vector and the second expression vector of each participle can be calculated.
In one embodiment of the invention, vector carries out linear transformation to be represented to the first expression vector sum second, to obtain
The first process vector sum second processing vector is taken, then, cosine calculating is carried out to the first process vector sum second processing vector, is obtained
Take the degree of correlation of each participle and search word in Search Results.Thus, by reducing to representing that vector carries out linear transformation
Redundancy in expression vector reduces complexity of the calculation, improves follow-up relatedness computation to the follow-up impact for calculating
Efficiency and accuracy.
S13, highlights journey according to what the degree of correlation of each participle and search word determined each participle in Search Results
Degree.
, wherein it is desired to be understood by, different degrees of correlation are corresponding to highlight degree difference.
S14, highlights each participle in Search Results according to the degree that highlights of each participle.
That is, when Search Results are provided the user, in search results pages, degree is highlighted showing each with different
Participle.
For example, Search Results are highlighted with yellow, it is assumed that default setting degree of correlation is higher, and highlighted color is deeper,
The depth of highlight color can be determined according to the degree of correlation of each participle in Search Results and search word, and in search results pages
In with the yellow of the different depths come the participle being highlighted in Search Results.
In summary it can be seen, the display packing of the Search Results based on artificial intelligence of the embodiment of the present invention, according to defeated
After the search word that enters obtains Search Results, by the attention model of training in advance, obtain in Search Results each participle and search
The degree of correlation of rope word, and the degree of correlation according to each participle with search word determines that the protrusion of each participle in Search Results shows
Show degree, and each participle in Search Results is highlighted according to the degree that highlights of each participle.Thus, in search
Program is highlighted with difference in result page and is highlighted each participle in Search Results, provided the user and there is stereovision
Search results pages, facilitate user according to difference highlight degree understand Search Results in participle and search word related journey
Degree.
In one embodiment of the invention, before the search word according to input obtains Search Results, training can be obtained
Data, and according to the title of sample searches word and its sample searches result for being noted as positive example or negative example accordingly to attention
Model is trained.
Wherein, training data includes sample searches word and its is noted as the sample searches result of positive example or negative example accordingly
Title.
Wherein, training data is obtained from the inquiry log of user.
Wherein, the sample searches result for being noted as positive example is the Search Results of clicked mistake.
Wherein, the sample searches result for being noted as negative example is the Search Results not clicked on.
Specifically, the fixed fixed expression represented with the title of sample searches result of sample searches word can be obtained, then
The parameter of attention model is trained using the mode of paired Pairwise, the algorithm based on gradient is particularly used, in meter
Calculate calculating is carried out come the power model that gains attention on cluster.
Fig. 3 is the flow chart of the display packing of the Search Results based on artificial intelligence according to one embodiment of the invention.
As shown in figure 3, the display packing of the Search Results based on artificial intelligence according to embodiments of the present invention, including it is following
Step.
S31, obtains Search Results according to the search word of input.
S32, using the attention model of training in advance, obtains the related journey of each participle and search word in Search Results
Degree.
In one embodiment of the invention, the attention model of training in advance can be adopted to obtain each point in Search Results
First expression vector of word, obtains the second expression vector of search word using the attention model of training in advance, according to the second table
Show that vector sum first represents that vector obtains the degree of correlation of each participle and search word in Search Results.
In one embodiment of the invention, vector carries out linear transformation to be represented to the first expression vector sum second, is obtained
First processes vector sum second processing vector, then, carries out cosine calculating to the first process vector sum second processing vector, obtains
The degree of correlation of each participle and search word in Search Results.
S33, determines the highlighted degree of each participle in Search Results according to the degree of correlation of each participle and search word.
, wherein it is desired to be understood by, different degrees of correlation are corresponding to highlight degree difference.
S34, is highlighted to each participle in Search Results according to the highlighted degree of each analysis.
For example, Search Results are highlighted with yellow, it is assumed that default setting degree of correlation is higher, and highlighted color is deeper,
The depth of highlight color can be determined according to the degree of correlation of each participle in Search Results and search word, and in search results pages
In with the yellow of the different depths come the participle being highlighted in Search Results.
In summary it can be seen, the display packing of the Search Results based on artificial intelligence of the embodiment of the present invention, according to defeated
After the search word that enters obtains Search Results, by the attention model of training in advance, obtain in Search Results each participle and search
The degree of correlation of rope word, and determine that the highlighted of each participle shows in Search Results according to each participle and the degree of correlation of search word
Show degree, and each participle that degree is highlighted in Search Results is highlighted according to each participle.Thus, in search
Program is highlighted with difference in result page and has been highlighted each participle in Search Results, provided the user and there is stereovision
Search results pages, facilitate user according to difference be highlighted degree understand Search Results in participle and search word related journey
Degree.
In order to realize above-described embodiment, the present invention also proposes a kind of display device of the Search Results based on artificial intelligence.
Fig. 4 is the structural representation of the display device of the Search Results based on artificial intelligence according to one embodiment of the invention
Figure.
As shown in figure 4, the display device of the Search Results based on artificial intelligence according to embodiments of the present invention, including first
Acquisition module 110, the second acquisition module 120, determining module 130 and display module 140, wherein：
First acquisition module 110 obtains Search Results for the search word according to input.
Second acquisition module 120 is used for the attention model using training in advance, obtain in Search Results each participle with
The degree of correlation of search word.
Determining module 130 determines each participle in Search Results for the degree of correlation according to each participle with search word
Highlight degree.
Display module 140 highlights each point in Search Results for the degree that highlights according to each participle
Word.
In one embodiment of the invention, on the basis of shown in Fig. 4, as shown in figure 5, the second acquisition module 120 is wrapped
First acquisition unit 121, second acquisition unit 122 and the 3rd acquiring unit 123 are included, wherein：
First acquisition unit 121 is used to obtain the of each participle in Search Results using the attention model of training in advance
One represents vector.
Second acquisition unit 122 is used for the second expression vector that search word is obtained using the attention model of training in advance.
3rd acquiring unit 123 obtains each point in Search Results for representing vectorial according to the second expression vector sum first
The degree of correlation of word and search word.
In one embodiment of the invention, using the expression vector of last participle in search word as the of search word
Two represent vector, or, the expression vector of all participles of search word is added into obtained expression vector as searching by element
Second expression vector of rope word.
In one embodiment of the invention the 3rd acquiring unit 123 specifically for：
Vector carries out linear transformation to be represented to the first expression vector sum second, obtain first process vector sum second processing to
Amount.
Cosine calculating is carried out to the first process vector sum second processing vector, each participle and search in Search Results is obtained
The degree of correlation of word.
In one embodiment of the invention, determining module 130 specifically for：It is related to search word according to each participle
Degree determines the highlighted degree of each participle in Search Results.
In one embodiment of the invention, display module 140 specifically for：According to the highlighted degree of each analysis to searching
Each participle in hitch fruit is highlighted.
In one embodiment of the invention, on the basis of shown in Fig. 4, as shown in fig. 6, the device can also include the
Three acquisition modules 150 and training module 160, wherein：
3rd acquisition module 150 is used to obtain training data, wherein, training data includes sample searches word and its corresponding
It is noted as the title of the sample searches result of positive example or negative example.
Training module 160 is for being noted as according to sample searches word and its accordingly the sample searches knot of positive example or negative example
The title of fruit is trained to attention model.
, wherein it is desired to explanation, the 3rd acquisition module 150 and training module 160 in the device embodiment shown in Fig. 6
Structure can also be included in the device embodiment shown in Fig. 5, the present invention this is not construed as limiting.
, wherein it is desired to explanation, the explanation of the display packing of the aforementioned Search Results to based on artificial intelligence
Suitable for the display device of the Search Results based on artificial intelligence of the embodiment, here is omitted.
The display device of the Search Results based on artificial intelligence of the embodiment of the present invention, obtains in the search word according to input
After Search Results, by the attention model of training in advance, the degree of correlation of each participle and search word in Search Results is obtained,
And determine the degree that highlights of each participle in Search Results according to the degree of correlation of each participle and search word, and according to
The degree that highlights of each participle highlights each participle in Search Results.Thus, with difference in search results pages
The program of highlighting is highlighted each participle in Search Results, has provided the user the search results pages with stereovision,
Facilitate user that the degree of correlation that degree understands the participle in Search Results and search word is highlighted according to difference.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
Example ", or the description of " some examples " etc. mean specific features with reference to the embodiment or example description, structure, material or spy
Point is contained at least one embodiment or example of the present invention.In this manual, to the schematic representation of above-mentioned term not
Identical embodiment or example must be directed to.And, the specific features of description, structure, material or feature can be with office
Combined in one or more embodiments or example in an appropriate manner.Additionally, in the case of not conflicting, the skill of this area
The feature of the different embodiments or example described in this specification and different embodiments or example can be tied by art personnel
Close and combine.
Additionally, term " first ", " second " are only used for describing purpose, and it is not intended that indicating or implying relative importance
Or the implicit quantity for indicating indicated technical characteristic.Thus, define " first ", the feature of " second " can express or
Implicitly include at least one this feature.In describing the invention, " multiple " are meant that two or more, unless separately
There is clearly specific restriction.
In flow chart or here any process described otherwise above or method description are construed as, expression includes
It is one or more for realizing specific logical function or process the step of the module of code of executable instruction, fragment or portion
Point, and the scope of the preferred embodiment of the present invention includes other realization, wherein the suitable of shown or discussion can not be pressed
Sequence, including according to involved function by it is basic simultaneously in the way of or in the opposite order, carry out perform function, this should be of the invention
Embodiment person of ordinary skill in the field understood.
Expression or here logic described otherwise above and/or step, for example, are considered use in flow charts
In the order list of the executable instruction for realizing logic function, may be embodied in any computer-readable medium, for
Instruction execution system, device or equipment (as computer based system, the system including processor or other can hold from instruction
The system of row system, device or equipment instruction fetch execute instruction) use, or with reference to these instruction execution systems, device or set
It is standby and use.For the purpose of this specification, " computer-readable medium " can any can be included, store, communicate, propagate or pass
The dress that defeated program is used for instruction execution system, device or equipment or with reference to these instruction execution systems, device or equipment
Put.The more specifically example (non-exhaustive list) of computer-readable medium is including following：With the electricity that one or more connect up
Connecting portion (electronic installation), portable computer diskette box (magnetic device), random access memory (RAM), read only memory
(ROM), erasable edit read-only storage (EPROM or flash memory), fiber device, and portable optic disk is read-only deposits
Reservoir (CDROM).In addition, computer-readable medium can even is that the paper that can print described program thereon or other are suitable
Medium, because for example by carrying out optical scanning to paper or other media edlin, interpretation can then be entered or if necessary with which
His suitable method is processed to electronically obtain described program, is then stored in computer storage.
It should be appreciated that each several part of the present invention can be realized with hardware, software, firmware or combinations thereof.Above-mentioned
In embodiment, the software that multiple steps or method can be performed in memory and by suitable instruction execution system with storage
Or firmware is realizing.For example, if realized with hardware, and in another embodiment, can be with well known in the art
Any one of row technology or their combination are realizing：With for the logic gates of logic function is realized to data signal
Discrete logic, the special IC with suitable combinational logic gate circuit, programmable gate array (PGA), scene
Programmable gate array (FPGA) etc..
Those skilled in the art are appreciated that to realize all or part of step that above-described embodiment method is carried
Suddenly the hardware that can be by program to instruct correlation is completed, and described program can be stored in a kind of computer-readable storage medium
In matter, the program upon execution, including one or a combination set of the step of embodiment of the method.
Additionally, each functional unit in each embodiment of the invention can be integrated in a processing module, it is also possible to
It is that unit is individually physically present, it is also possible to which two or more units are integrated in a module.Above-mentioned integrated mould
Block both can be realized in the form of hardware, it would however also be possible to employ the form of software function module is realized.The integrated module is such as
Fruit using in the form of software function module realize and as independent production marketing or use when, it is also possible to be stored in a computer
In read/write memory medium.
Storage medium mentioned above can be read only memory, disk or CD etc..Although having shown that above and retouching
Embodiments of the invention are stated, it is to be understood that above-described embodiment is exemplary, it is impossible to be interpreted as the limit to the present invention
System, one of ordinary skill in the art can be changed to above-described embodiment, change, replace and become within the scope of the invention
Type.
Claims (12)
1. a kind of display packing of the Search Results based on artificial intelligence, it is characterised in that comprise the following steps：
Search Results are obtained according to the search word of input；
Using the attention model of training in advance, the related journey of each participle and the search word in the Search Results is obtained
Degree；
Degree is highlighted according to what the degree of correlation of each participle and search word determined each participle in the Search Results；
Each participle in the Search Results is highlighted according to the degree that highlights of each participle.
2. the method for claim 1, it is characterised in that the attention model of the employing training in advance, obtains described
The degree of correlation of each participle and the search word in Search Results, including：
First expression vector of each participle in the Search Results is obtained using the attention model of training in advance；
Second expression vector of the search word is obtained using the attention model of training in advance；
According to described second represents vector sum, first represents that vector obtains each participle and the search word in Search Results
Degree of correlation.
3. the method for claim 1, it is characterised in that the degree of correlation according to each participle with search word determines
Each participle highlights degree in the Search Results, including：
The highlighted degree of each participle in the Search Results is determined according to the degree of correlation of each participle and search word；
The degree that highlights according to each participle highlights each participle in the Search Results, including：
Each participle in the Search Results is highlighted according to the highlighted degree of each analysis.
4. the method as any one of claim 1-3, it is characterised in that obtain in the search word according to input and search
Before hitch fruit, also include：
Training data is obtained, wherein, the training data includes sample searches word and its is noted as positive example or negative example accordingly
Sample searches result title；
According to the title of the sample searches word and its sample searches result for being noted as positive example or negative example accordingly to described
Attention model is trained.
5. the method as described in claim 2, it is characterised in that first table according to described second represents vector sum
Show that vector obtains the degree of correlation of each participle and the search word in Search Results, including：
Linear transformation is carried out to the second expression vector described in the described first expression vector sum, to obtain the first process vector sum second
Process vector；
Cosine calculating is carried out to second processing vector described in the described first process vector sum, each point in the Search Results is obtained
The degree of correlation of word and the search word.
6. the method as described in claim 2, it is characterised in that by the expression vector of last participle in the search word
As the second expression vector of the search word, or, the expression vector of all participles of the search word is added by element
Second expression vector of the expression vector for being obtained as the search word.
7. a kind of display device of the Search Results based on artificial intelligence, it is characterised in that include：
First acquisition module, obtains Search Results for the search word according to input；
Second acquisition module, for the attention model using training in advance, obtains each participle and institute in the Search Results
State the degree of correlation of search word；
Determining module, for determining the prominent of each participle in the Search Results according to the degree of correlation of each participle and search word
Go out display degree；
Display module, highlights each participle in the Search Results for the degree that highlights according to each participle.
8. device as claimed in claim 7, it is characterised in that second acquisition module, including：
First acquisition unit, obtains first of each participle in the Search Results for the attention model using training in advance
Represent vector；
Second acquisition unit, obtains the second expression vector of the search word for the attention model using training in advance；
For according to described second represents vector sum first, 3rd acquiring unit, represents that vector obtains each in Search Results
The degree of correlation of participle and the search word.
9. device as claimed in claim 8, it is characterised in that the determining module, specifically for：
The highlighted degree of each participle in the Search Results is determined according to the degree of correlation of each participle and search word；
The display module, specifically for：
Each participle in the Search Results is highlighted according to the highlighted degree of each analysis.
10. device as claimed in any one of claims 7-9, it is characterised in that also include：
3rd acquisition module, for obtaining training data, wherein, the training data includes sample searches word and its quilt accordingly
It is labeled as the title of the sample searches result of positive example or negative example；
Training module, for being noted as according to the sample searches word and its accordingly the sample searches result of positive example or negative example
Title the attention model is trained.
11. devices as described in claim 8, it is characterised in that the 3rd acquiring unit, specifically for：
Linear transformation is carried out to the second expression vector described in the described first expression vector sum, to obtain the first process vector sum second
Process vector；
Cosine calculating is carried out to second processing vector described in the described first process vector sum, each point in the Search Results is obtained
The degree of correlation of word and the search word.
12. devices as described in claim 8, it is characterised in that by the expression of last participle in the search word to
The second expression vector as the search word is measured, or, by the expression vector of all participles of the search word by element phase
Plus the expression vector for being obtained is used as the second expression vector of the search word.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610974823.3A CN106557554B (en) | 2016-11-04 | 2016-11-04 | The display methods and device of search result based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610974823.3A CN106557554B (en) | 2016-11-04 | 2016-11-04 | The display methods and device of search result based on artificial intelligence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106557554A true CN106557554A (en) | 2017-04-05 |
| CN106557554B CN106557554B (en) | 2019-11-08 |
Family
ID=58443948
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610974823.3A Active CN106557554B (en) | 2016-11-04 | 2016-11-04 | The display methods and device of search result based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106557554B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107301170A (en) * | 2017-06-19 | 2017-10-27 | 北京百度网讯科技有限公司 | The method and apparatus of cutting sentence based on artificial intelligence |
| CN107977363A (en) * | 2017-12-20 | 2018-05-01 | 北京百度网讯科技有限公司 | Title generation method, device and electronic equipment |
| CN109165361A (en) * | 2018-07-31 | 2019-01-08 | 优视科技新加坡有限公司 | The method, apparatus and equipment/terminal/server of page presentation in a kind of information flow |
| CN110909150A (en) * | 2018-09-13 | 2020-03-24 | 深圳市蓝灯鱼智能科技有限公司 | Search result display method and device, storage medium and electronic device |
| CN111831938A (en) * | 2020-06-17 | 2020-10-27 | 维沃移动通信有限公司 | Information display method, information display device, electronic equipment and medium |
| CN112732870A (en) * | 2020-12-31 | 2021-04-30 | 平安科技（深圳）有限公司 | Searching method, device and equipment based on word vector and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104572825A (en) * | 2014-12-04 | 2015-04-29 | 百度在线网络技术（北京）有限公司 | Method and device for recommending information |
| CN104598611A (en) * | 2015-01-29 | 2015-05-06 | 百度在线网络技术（北京）有限公司 | Method and system for sequencing search entries |
| CN104615767A (en) * | 2015-02-15 | 2015-05-13 | 百度在线网络技术（北京）有限公司 | Searching-ranking model training method and device and search processing method |
| CN104636465A (en) * | 2015-02-10 | 2015-05-20 | 百度在线网络技术（北京）有限公司 | Webpage abstract generating methods and displaying methods and corresponding devices |
| CN105808566A (en) * | 2014-12-30 | 2016-07-27 | 北京奇虎科技有限公司 | Method and device for extracting abstracts from webpages on basis of search words |
-
2016
- 2016-11-04 CN CN201610974823.3A patent/CN106557554B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104572825A (en) * | 2014-12-04 | 2015-04-29 | 百度在线网络技术（北京）有限公司 | Method and device for recommending information |
| CN105808566A (en) * | 2014-12-30 | 2016-07-27 | 北京奇虎科技有限公司 | Method and device for extracting abstracts from webpages on basis of search words |
| CN104598611A (en) * | 2015-01-29 | 2015-05-06 | 百度在线网络技术（北京）有限公司 | Method and system for sequencing search entries |
| CN104636465A (en) * | 2015-02-10 | 2015-05-20 | 百度在线网络技术（北京）有限公司 | Webpage abstract generating methods and displaying methods and corresponding devices |
| CN104615767A (en) * | 2015-02-15 | 2015-05-13 | 百度在线网络技术（北京）有限公司 | Searching-ranking model training method and device and search processing method |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107301170A (en) * | 2017-06-19 | 2017-10-27 | 北京百度网讯科技有限公司 | The method and apparatus of cutting sentence based on artificial intelligence |
| CN107301170B (en) * | 2017-06-19 | 2020-12-22 | 北京百度网讯科技有限公司 | Method and device for segmenting sentences based on artificial intelligence |
| CN107977363A (en) * | 2017-12-20 | 2018-05-01 | 北京百度网讯科技有限公司 | Title generation method, device and electronic equipment |
| CN107977363B (en) * | 2017-12-20 | 2021-12-17 | 北京百度网讯科技有限公司 | Title generation method and device and electronic equipment |
| CN109165361A (en) * | 2018-07-31 | 2019-01-08 | 优视科技新加坡有限公司 | The method, apparatus and equipment/terminal/server of page presentation in a kind of information flow |
| WO2020026015A1 (en) * | 2018-07-31 | 2020-02-06 | 优视科技新加坡有限公司 | Method and apparatus for page display in information stream, and device/terminal/server |
| CN110909150A (en) * | 2018-09-13 | 2020-03-24 | 深圳市蓝灯鱼智能科技有限公司 | Search result display method and device, storage medium and electronic device |
| CN111831938A (en) * | 2020-06-17 | 2020-10-27 | 维沃移动通信有限公司 | Information display method, information display device, electronic equipment and medium |
| CN112732870A (en) * | 2020-12-31 | 2021-04-30 | 平安科技（深圳）有限公司 | Searching method, device and equipment based on word vector and storage medium |
| CN112732870B (en) * | 2020-12-31 | 2024-03-05 | 平安科技（深圳）有限公司 | Word vector based search method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106557554B (en) | 2019-11-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106557554A (en) | Display packing and device based on the Search Results of artificial intelligence | |
| US11003856B2 (en) | Processing text using neural networks | |
| CN107798136B (en) | Entity relation extraction method and device based on deep learning and server | |
| Nagaraja et al. | Modeling context between objects for referring expression understanding | |
| CN109376309A (en) | Document recommendation method and device based on semantic label | |
| CN106571139A (en) | Artificial intelligence based voice search result processing method and device | |
| CN110929498B (en) | Method and device for calculating similarity of short text and readable storage medium | |
| CN106294635B (en) | Application program searching method, the training method of deep neural network model and device | |
| CN110990555B (en) | End-to-end retrieval type dialogue method and system and computer equipment | |
| CN107122492A (en) | Lyric generation method and device based on picture content | |
| CN112800239B (en) | Training method of intention recognition model, and intention recognition method and device | |
| CN110222654A (en) | Text segmenting method, device, equipment and storage medium | |
| CN108959559A (en) | Question and answer are to generation method and device | |
| CN108959556A (en) | Entity answering method, device and terminal neural network based | |
| CN107203611A (en) | Information providing method and device based on picture content | |
| He et al. | Image captioning with text-based visual attention | |
| WO2019118257A1 (en) | Assertion-based question answering | |
| CN112200031A (en) | Network model training method and equipment for generating image corresponding word description | |
| CN106844340A (en) | News in brief generation and display methods, apparatus and system based on artificial intelligence | |
| CN103914569B (en) | Input creation method, the device of reminding method, device and dictionary tree-model | |
| CN106951413A (en) | Segmenting method and device based on artificial intelligence | |
| Tseng et al. | An innovative bert-based readability model | |
| Wang et al. | Gated hierarchical attention for image captioning | |
| CN107025312A (en) | Information providing method and device based on video content | |
| CN109189882A (en) | Answer type recognition methods, device, server and the storage medium of sequence content |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |