CN104281690B - A kind of label-cloud generation method and device - Google Patents
A kind of label-cloud generation method and device Download PDFInfo
- Publication number
- CN104281690B CN104281690B CN201410534723.XA CN201410534723A CN104281690B CN 104281690 B CN104281690 B CN 104281690B CN 201410534723 A CN201410534723 A CN 201410534723A CN 104281690 B CN104281690 B CN 104281690B
- Authority
- CN
- China
- Prior art keywords
- label
- article
- matrix
- cloud
- text set
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/31—Indexing; Data structures therefor; Storage structures
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application provides a kind of label-cloud generation method and device,The label-cloud that text set information is carried by receiving generates request,For each label in every article in text set corresponding with text set information,Calculate weighted value generation story label matrix of the label in its affiliated article,And by carrying out singular value decomposition to story label matrix,Generate the second matrix of the weight of the first matrix and indicating label of weight of the indicative character vector in text set in characteristic vector,And then utilize the first matrix,Second matrix and the create-rule pre-set realize the generation of label-cloud,The application to story label matrix by carrying out singular value decomposition,And then utilize the matrix generation label-cloud after decomposing,When avoiding index of the label-cloud of prior art generation as article set key content,Semantic coverage indicated by each tag element is excessively wide in range,The problem of key content of embodiment article set is not accurate enough.
Description
Technical field
The application is related to label-cloud technical field, more particularly to a kind of label-cloud generation method and device.
Background technology
Label-cloud is generally by indexing some article set, and then utilize the higher label of the frequency of occurrences in this article set
Generation.By the way that the label-cloud is shown with visual means, user can be allowed intuitively to understand letter important in this article set
Breath, and the label-cloud is alternatively arranged as the index of article set key content, when user clicks on any one label in the label-cloud
During element, the information related to the tag element can be found out from article set immediately, facilitates user to consult information.
But because traditional label-cloud is that the basis of frequency statistics is directly being carried out to each label in article set
On, meet what each label of preset requirement generated as a tag element in label-cloud by the use of the frequency, therefore, it will usually
Because the tag element form for forming label-cloud is single (i.e.：Each tag element is only made up of a label) and cause label-cloud
During index as article set key content, semantic coverage indicated by each tag element is excessively wide in range, embodies article collection
The problem of key content of conjunction is not accurate enough.
The content of the invention
In view of this, the application provides a kind of label-cloud generation method and device, to avoid the label that prior art generates
During index of the cloud as article set key content, semantic coverage indicated by each tag element is excessively wide in range, embodies article
The problem of key content of set is not accurate enough.
To achieve these goals, technical scheme provided in an embodiment of the present invention is as follows：
A kind of label-cloud generation method, including：
Label-cloud generation request is received, wherein carrying text set information；
For each label in every article in text set corresponding with the text set information, calculate the label and exist
Weighted value in its affiliated article；
Utilize the label and the weight of the label in each piece article corresponding with the text set information, every article
Value, generate article-label matrix；
Singular value decomposition, power of the generation indicative character vector in the text set are carried out to the article-label matrix
Second matrix of weight of the first matrix and indicating label of weight in the characteristic vector；
Using first matrix, the second matrix and the create-rule pre-set, label-cloud is generated.
Preferably, each label in for every article corresponding with the text set information, calculates the label and exists
During weighted value in its affiliated article, calculated using equation below：
S(i)(Wk)=【Ssource(Wk)-Pos(Wk)*λ(Ssource(Wk))】*idf(Wk)*Sattributes(Wk), wherein, institute
State S(i)(Wk) it is k-th of label W in i-th history articlekThe first weighted value in the history article, the Ssource(Wk)
For label WkSource parameter, the Pos (Wk) it is label WkLocation parameter, the λ (Ssource(Wk)) it is because of label WkPosition
Put introduced punishment parameter, the idf (Wk) it is the label WkSignificance level in internet, the Sattributes
(Wk) it is the label WkPart of speech parameter.
Preferably, it is described to utilize each piece article corresponding with the text set information, the label in every article and described
The weighted value of label, article-label matrix is generated, including：
For every article, weighted value meets the label of the first threshold scope pre-set in acquisition this article；
Obtain the union of each label；
Article-label matrix is generated using each label that is described and concentrating, wherein, it is every in the article-label matrix
Row represents an article in each label that is described and concentrating, and each column represents all texts corresponding to a label that is described and concentrating
Chapter, and the element in this article-label matrix is the weighted value of label.
Preferably, it is described using first matrix, the second matrix and the create-rule pre-set, label-cloud is generated,
Including：
Obtain and meet each first element for pre-setting Second Threshold scope in first matrix；
For row corresponding with each first element respectively in second matrix, obtain and meet to set in advance in the row
Label is as a tag element in label-cloud corresponding to each second element for the 3rd threshold range put.
Preferably, in addition to：Show the label-cloud being made up of each tag element.
Preferably, in addition to：When the quantity of label in the tag element is more than preset value, deleted according to what is pre-set
Except the part labels in tag element described in redundant rule elimination.
A kind of label-cloud generating means, including：
Receiving unit, for receiving label-cloud generation request, wherein carrying text set information；
Computing unit, for for each mark in every article in text set corresponding with the text set information
Label, calculate weighted value of the label in its affiliated article；
First generation unit, for utilizing the label in each piece article corresponding with the text set information, every article
And the weighted value of the label, generate article-label matrix；
Second generation unit, for carrying out singular value decomposition to the article-label matrix, generation indicative character vector exists
Second matrix of weight of the first matrix and indicating label of the weight in the text set in the characteristic vector；
3rd generation unit, for utilizing first matrix, the second matrix and the create-rule pre-set, generation mark
Sign cloud.
Preferably, the 3rd generation unit includes：
Acquiring unit, meet each first yuan that pre-sets Second Threshold scope in first matrix for obtaining
Element；
3rd generation subelement, for for row corresponding with each first element respectively in second matrix,
Label corresponding to each second element for the 3rd threshold range for meeting to pre-set in the row is obtained as one in label-cloud
Individual tag element.
Preferably, in addition to：
Display unit, for showing the label-cloud being made up of each tag element.
Preferably, in addition to：
Unit is deleted, for when the quantity of label in the tag element is more than preset value, being deleted according to what is pre-set
Except the part labels in tag element described in redundant rule elimination.
The application provides a kind of label-cloud generation method and device, and the label-cloud that text set information is carried by receiving generates
Request, for each label in every article in text set corresponding with text set information, calculates the label belonging to it
Weighted value generation article-label matrix in article, and by carrying out singular value decomposition, generation instruction to article-label matrix
Second matrix of weight of the first matrix and indicating label of weight of the characteristic vector in text set in characteristic vector, and then
The generation of label-cloud is realized using the first matrix, the second matrix and the create-rule that pre-sets, the application passes through to article-mark
Sign matrix and carry out singular value decomposition, and then label-cloud is generated using the matrix after decomposing, avoid the label of prior art generation
During index of the cloud as article set key content, semantic coverage indicated by each tag element is excessively wide in range, embodies article
The problem of key content of set is not accurate enough.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the required accompanying drawing used in technology description to be briefly described, it should be apparent that, drawings in the following description are only this
The embodiment of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can also basis
The accompanying drawing of offer obtains other accompanying drawings.
Fig. 1 is a kind of label-cloud generation method flow chart that the embodiment of the present application one provides；
Fig. 2 is a kind of structural representation for label-cloud generating means that the embodiment of the present application two provides.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other under the premise of creative work is not made
Embodiment, belong to the scope of protection of the invention.
Embodiment one：
Fig. 1 is a kind of label-cloud generation method flow chart that the embodiment of the present application one provides.
As shown in figure 1, this method includes：
S101, label-cloud generation request is received, wherein carrying text set information.
S102, for each label in every article in text set corresponding with text set information, calculate the label
Weighted value in its affiliated article.
In the embodiment of the present application, it is preferred that the corresponding text set for including at least one article of text set information, and every
Carry at least one label in article, the source of the label when can be article generation it is user-defined (such as：Work as preservation
During certain article, be its set a label be " sweet apple ") or process word segmentation processing obtain (such as：By user
Customized label " sweet apple " passes through word segmentation processing, and two labels " banana " of generation and " apple "/when preserving article,
Using the more word of occurrence number in article as participle label).
Preferably, after the label-cloud for carrying text set information is received, need first for corresponding with text set information
Each label in every article in text set, weighted value of the label in its affiliated article is calculated, wherein, specific meter
It is as follows to calculate formula：
S(i)(Wk)=【Ssource(Wk)-Pos(Wk)*λ(Ssource(Wk))】*idf(Wk)*Sattributes(Wk), wherein, S(i)
(Wk) it is k-th of label W in i-th articlekWeighted value in this article, Ssource(Wk) it is label WkSource parameter, Pos
(Wk) it is label WkLocation parameter, λ (Ssource(Wk)) it is because of label WkThe introduced punishment parameter in position, idf (Wk) it is mark
Sign WkSignificance level in internet, Sattributes(Wk) it is label WkPart of speech parameter.
In the embodiment of the present application, it is preferred that Ssource(Wk) it is label WkSource parameter, wherein, the source of label refers to
It is customized label/participle label to show the label, and preferably, pre-sets the S when label is customized labelsource(Wk)
Value for participle label when 8~20 times.
In the embodiment of the present application, it is preferred that Pos (Wk) it is label WkLocation parameter, wherein, the position instruction of label
The position in each label of source identical in article of the label belonging to it, and preferably, text of the label belonging to it
Which position, the Pos (W of the label are arranged in each label of source identical in chapterk) value is several, such as：When carrying 5 in some article
Individual label, wherein 3 are participle label, this 3 participle labels are followed successively by " banana ", " apple ", " pear ", then, label " duck
Pos (the W of pears "k) value be 3.
In the embodiment of the present application, it is preferred that λ (Ssource(Wk)) it is because of label WkThe introduced punishment parameter in position,
Wherein, punishment parameter is different because the source of label is different, it is preferred that the λ (S pre-setsource(Wk)) value 0.08~
Between 0.11, and Ssource(Wk)-Pos(Wk)*λ(Ssource(Wk)) value be more than or equal to 0.5.
In the embodiment of the present application, it is preferred that idf (Wk) it is label WkSignificance level in internet, wherein, calculate
The process of the significance level of some label is prior art, refers to prior art, is not described in detail herein.
In the embodiment of the present application, it is preferred that Sattributes(Wk) it is label WkPart of speech parameter, wherein, it is preferred that mark
The part of speech of label be proper noun, noun, verb, adjective, adverbial word, and when part of speech be proper noun, noun, verb, adjective,
During adverbial word, S is followed successively byattributes(Wk) it is entered as 10,9,5,4,4.
S103, utilize the label and the weighted value of label in each piece article corresponding with text set information, every article, life
Into article-label matrix.
In the embodiment of the present application, it is preferred that when in every article in for text set corresponding with text set information
Each label, after weighted value of the label in its affiliated article is calculated, need to utilize corresponding with text set information each
The weighted value of piece article, the label in every article and label, generates article-label matrix, specifically generates article-label square
The process of battle array is as follows：
1st, for every article, the label of first threshold scope that weighted value in this article meets to pre-set is obtained.
In the embodiment of the present application, it is preferred that be previously provided with first threshold scope, for every article, obtain this article
Weighted value meets the label of the first threshold scope pre-set in chapter.
2nd, the union of each label is obtained.
In the embodiment of the present application, it is preferred that the label of repetition is there may be between article, as article A carry label 1,
Also label 1 is carried in article B.
Specifically, after the label for the first threshold scope that weighted value satisfaction is pre-set is got for every article,
The each label that need to be directed in the every article got, obtain the union of each label.
3rd, utilization and each label generation article-label matrix concentrated, wherein, often row represents in article-label matrix
One article and each label for concentrating, all articles corresponding to the label that each column is represented and concentrated, and this article-mark
Sign the weighted value that the element in matrix is label.
In the embodiment of the present application, it is preferred that after the union of each label is got, this need to be utilized and concentrated each
Label generate article-label matrix, wherein, in article-label matrix often row represent an article and concentrate each label,
All articles corresponding to the label that each column is represented and concentrated, and the element in this article-label matrix is the weight of label
Value.
Specifically, the article list that text set corresponding with text set information includes is { D1, D2..., Dn, every text
Chapter DtEach label TmAnd label TmWeighted value W in this articlemThe list of labels of composition is { (T1, w1) ...,
(Tm, wm), " it is more than or equal to θ ", for each label T by pre-setting first threshold scopemIf its weighted value Wm>=θ,
Then determine label TmTo obtain result, thus from every article DtIn each label TmIn filter out and meet to pre-set
All labels of first threshold scope are { T1, w1) ..., (Tp, wp), what each tag computation filtered out using this was obtained
Article-label matrix is as follows：
Wherein, n is article quantity, p is that each label takes the quantity after union, often row represent an article and concentrate
Each label, all articles corresponding to the label that each column is represented and concentrated, and the element in this article-label matrix is
The weighted value of label.
Such as：When text set corresponding to text set information includes 2 articles, respectively carry label A, label B, label C,
Label D article 1, and label A, label B, label C, label D, the article 2 of label E are carried, by for each tag computation
Its weighted value in affiliated article, and according to the first threshold scope pre-set, it is determined that meeting the first threshold scope
Label is respectively：Label A (weighted value 1A), label B (weighted value 1B), label C (weighted value 1C) in article 1, and
Label A (weighted value 2A), label C (weighted value 2C), label D (weighted value 2D) in article 2, now, the text of generation
Chapter-label matrix is as follows：
It can be seen that in article-label matrix often row represent an article and each label for concentrating, each column represents and concentrates
A label corresponding to all articles, and the element in this article-label matrix be label weighted value, when and concentrate deposit
In label B, and article 2 and the label concentrated and when label B is not present, will in article-label matrix expression article 2 label
B element is arranged to 0.
In the embodiment of the present application, it is preferred that the element not limited in article-label matrix of the generation in row is corresponding
Order in union label, inventor can arbitrarily set according to the demand of oneself, and article-label matrix of such as generation is：
S104, singular value decomposition is carried out to article-label matrix, weight of the generation indicative character vector in text set
Second matrix of the weight of the first matrix and indicating label in characteristic vector.
In the embodiment of the present application, it is preferred that carry out singular value decomposition, generation for the article-label matrix got
Second matrix of weight of the first matrix and indicating label of weight of the indicative character vector in text set in characteristic vector.
In the embodiment of the present application, it is preferred that singular value decomposition is existing mathematical algorithm, refers to existing skill
Art, detailed restriction is not done herein.
S105, utilize the first matrix, the second matrix and the create-rule pre-set, generation label-cloud.
In the embodiment of the present application, it is preferred that when article-label matrix is carried out singular value decomposition obtain the first matrix and
After second matrix, the first matrix, the second matrix and the create-rule pre-set need to be utilized, label-cloud is generated, specifically, the mistake
Cheng Wei：
1st, each first element for meeting to pre-set Second Threshold scope in the first matrix is obtained.
2nd, for row corresponding with each first element respectively in the second matrix, meet to pre-set in the row is obtained
Label corresponding to each second element of three threshold ranges is as a tag element in label-cloud.
In the embodiment of the present application, it is preferred that weight of the first matrix indicative character vector in text set, the second matrix
Weight of the indicating label in characteristic vector.For some element in the first matrix, what it was indicated is corresponding with the element
Weight of the characteristic vector in text set, weight of the indicating label in characteristic vector corresponding with the element in the second matrix
Row is row corresponding with the element in the second matrix.
Further, in a kind of label-cloud generation method that the embodiment of the present application provides, in addition to：Display is by each mark
Sign the label-cloud of element composition.
Further, in a kind of label-cloud generation method that the embodiment of the present application provides, in addition to：When in tag element
When the quantity of label is more than preset value, the part labels in tag element are deleted according to the deletion rule pre-set.
In the embodiment of the present application, it is preferred that when the quantity of the label in some tag element is more than preset value, can press
Label in the tag element is deleted according to the weighted value order from small to large of each label, until label in the tag element
Quantity meets preset value.
The application provides a kind of label-cloud generation method, and the label-cloud that text set information is carried by receiving generates request,
For each label in every article in text set corresponding with text set information, the label is calculated in its affiliated article
Weighted value generation article-label matrix, and by carrying out singular value decomposition to article-label matrix, generation indicative character to
Measure the second matrix of the weight of the first matrix and indicating label of weight in text set in characteristic vector, and then utilize the
One matrix, the second matrix and the create-rule that pre-sets realize the generation of label-cloud, and the application passes through to article-label matrix
Singular value decomposition is carried out, and then using the matrix generation label-cloud after decomposing, avoids the label-cloud conduct of prior art generation
During the index of article set key content, the semantic coverage indicated by each tag element is excessively wide in range, embodiment article set
The problem of key content is not accurate enough.
Embodiment two：
Fig. 2 is a kind of structural representation for label-cloud generating means that the embodiment of the present application two provides.
As shown in Fig. 2 the device includes：The receiving unit 1 that is sequentially connected, computing unit 2, the first generation unit 3, second
The generation unit 5 of generation unit 4 and the 3rd, wherein：
Receiving unit 1, for receiving label-cloud generation request, wherein carrying text set information.
Computing unit 2, for for each label in every article in text set corresponding with text set information, meter
Calculate weighted value of the label in its affiliated article.
First generation unit 3, for utilizing the label and mark in each piece article corresponding with text set information, every article
The weighted value of label, generate article-label matrix.
Second generation unit 4, for carrying out singular value decomposition to article-label matrix, generation indicative character vector is in text
Second matrix of weight of the first matrix and indicating label of the weight of this concentration in characteristic vector.
3rd generation unit 5, for using the first matrix, the second matrix and the create-rule pre-set, generating label
Cloud.
In the embodiment of the present application, it is preferred that the 3rd generation unit includes：Acquiring unit, for obtaining the first matrix
Middle each first element for meeting to pre-set Second Threshold scope；3rd generation subelement, for in the second matrix points
Row not corresponding with each first element, obtain each second element pair for the 3rd threshold range for meeting to pre-set in the row
The label answered is as a tag element in label-cloud.
Further, in a kind of label-cloud generating means that the embodiment of the present application provides, in addition to：Display unit, use
In the label-cloud that display is made up of each tag element.
Further, in a kind of label-cloud generating means that the embodiment of the present application provides, in addition to：Unit is deleted, is used
In when the quantity of label in tag element is more than preset value, the portion in tag element is deleted according to the deletion rule pre-set
Minute mark label.
The application provides a kind of label-cloud generating means, and the label-cloud that the device carries text set information by receiving generates
Request, for each label in every article in text set corresponding with text set information, calculates the label belonging to it
Weighted value generation article-label matrix in article, and by carrying out singular value decomposition, generation instruction to article-label matrix
Second matrix of weight of the first matrix and indicating label of weight of the characteristic vector in text set in characteristic vector, and then
The generation of label-cloud is realized using the first matrix, the second matrix and the create-rule that pre-sets, the application passes through to article-mark
Sign matrix and carry out singular value decomposition, and then label-cloud is generated using the matrix after decomposing, avoid the label of prior art generation
During index of the cloud as article set key content, semantic coverage indicated by each tag element is excessively wide in range, embodies article
The problem of key content of set is not accurate enough.
Each embodiment is described by the way of progressive in this specification, what each embodiment stressed be and other
The difference of embodiment, between each embodiment identical similar portion mutually referring to.For device disclosed in embodiment
For, because it is corresponded to the method disclosed in Example, so description is fairly simple, related part is said referring to method part
It is bright.
It the above is only the preferred embodiment of the application, make skilled artisans appreciate that or realizing the application.It is right
A variety of modifications of these embodiments will be apparent to one skilled in the art, as defined herein general former
Reason can be realized in other embodiments in the case where not departing from spirit herein or scope.Therefore, the application will not
Be intended to be limited to the embodiments shown herein, and be to fit to it is consistent with principles disclosed herein and features of novelty most
Wide scope.
Claims (9)
- A kind of 1. label-cloud generation method, it is characterised in that including：Label-cloud generation request is received, wherein carrying text set information；For each label in every article in text set corresponding with the text set information, the label is calculated in its institute Belong to the weighted value in article；It is raw using the label and the weighted value of the label in each piece article corresponding with the text set information, every article Into article-label matrix；Singular value decomposition is carried out to the article-label matrix, weight of the generation indicative character vector in the text set Second matrix of the weight of the first matrix and indicating label in the characteristic vector；Using first matrix, the second matrix and the create-rule pre-set, label-cloud is generated；Wherein, each label in for every article corresponding with the text set information, calculates the label belonging to it During weighted value in article, calculated using equation below：S(i)(Wk)=【Ssource(Wk)-Pos(Wk)*λ(Ssource(Wk))】*idf(Wk)*Sattributes(Wk), wherein, the S(i) (Wk) it is k-th of label W in i-th history articlekThe first weighted value in the history article, the Ssource(Wk) it is label WkSource parameter, the Pos (Wk) it is label WkLocation parameter, the λ (Ssource(Wk)) it is because of label WkPosition drawn The punishment parameter entered, the idf (Wk) it is the label WkSignificance level in internet, the Sattributes(Wk) for institute State label WkPart of speech parameter.
- 2. according to the method for claim 1, it is characterised in that described to utilize each piece text corresponding with the text set information The weighted value of chapter, the label in every article and the label, article-label matrix is generated, including：For every article, weighted value meets the label of the first threshold scope pre-set in acquisition this article；Obtain the union of each label；Article-label matrix is generated using each label that is described and concentrating, wherein, every row table in the article-label matrix Show an article in each label that is described and concentrating, each column represents all articles corresponding to a label that is described and concentrating, And the element in this article-label matrix is the weighted value of label.
- 3. according to the method for claim 1, it is characterised in that it is described using first matrix, the second matrix and in advance The create-rule of setting, label-cloud is generated, including：Obtain and meet each first element for pre-setting Second Threshold scope in first matrix；For row corresponding with each first element respectively in second matrix, obtain and meet what is pre-set in the row Label corresponding to each second element of 3rd threshold range is as a tag element in label-cloud.
- 4. according to the method for claim 3, it is characterised in that also include：What display was made up of each tag element Label-cloud.
- 5. according to the method for claim 3, it is characterised in that also include：When the quantity of label in the tag element is big When preset value, the part labels in the tag element are deleted according to the deletion rule pre-set.
- A kind of 6. label-cloud generating means, it is characterised in that including：Receiving unit, for receiving label-cloud generation request, wherein carrying text set information；Computing unit, for for each label in every article in text set corresponding with the text set information, meter Calculate weighted value of the label in its affiliated article；First generation unit, for utilizing the label in each piece article corresponding with the text set information, every article and institute The weighted value of label is stated, generates article-label matrix；Second generation unit, for carrying out singular value decomposition to the article-label matrix, generation indicative character vector is described Second matrix of weight of the first matrix and indicating label of the weight in text set in the characteristic vector；3rd generation unit, for using first matrix, the second matrix and the create-rule pre-set, generating label Cloud；Wherein, each label in for every article corresponding with the text set information, calculates the label belonging to it During weighted value in article, calculated using equation below：S(i)(Wk)=【Ssource(Wk)-Pos(Wk)*λ(Ssource(Wk))】*idf(Wk)*Sattributes(Wk), wherein, the S(i) (Wk) it is k-th of label W in i-th history articlekThe first weighted value in the history article, the Ssource(Wk) it is label WkSource parameter, the Pos (Wk) it is label WkLocation parameter, the λ (Ssource(Wk)) it is because of label WkPosition drawn The punishment parameter entered, the idf (Wk) it is the label WkSignificance level in internet, the Sattributes(Wk) for institute State label WkPart of speech parameter.
- 7. device according to claim 6, it is characterised in that the 3rd generation unit includes：Acquiring unit, meet each first element for pre-setting Second Threshold scope in first matrix for obtaining；3rd generation subelement, for for row corresponding with each first element respectively in second matrix, obtaining Label corresponding to each second element for the 3rd threshold range for meeting to pre-set in the row is as a mark in label-cloud Sign element.
- 8. device according to claim 7, it is characterised in that also include：Display unit, for showing the label-cloud being made up of each tag element.
- 9. device according to claim 7, it is characterised in that also include：Unit is deleted, for when the quantity of label in the tag element is more than preset value, being advised according to the deletion pre-set Then delete the part labels in the tag element.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410534723.XA CN104281690B (en) | 2014-10-11 | 2014-10-11 | A kind of label-cloud generation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410534723.XA CN104281690B (en) | 2014-10-11 | 2014-10-11 | A kind of label-cloud generation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104281690A CN104281690A (en) | 2015-01-14 |
| CN104281690B true CN104281690B (en) | 2018-01-05 |
Family
ID=52256563
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410534723.XA Active CN104281690B (en) | 2014-10-11 | 2014-10-11 | A kind of label-cloud generation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104281690B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111026832A (en) * | 2019-11-15 | 2020-04-17 | 贝壳技术有限公司 | Method and system for generating articles |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103176961A (en) * | 2013-03-05 | 2013-06-26 | 哈尔滨工程大学 | Transfer learning method based on latent semantic analysis |
| CN103440256A (en) * | 2013-07-26 | 2013-12-11 | 中国科学院深圳先进技术研究院 | Method and device for automatically generating Chinese text label cloud |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8037066B2 (en) * | 2008-01-16 | 2011-10-11 | International Business Machines Corporation | System and method for generating tag cloud in user collaboration websites |
| US20110296345A1 (en) * | 2010-05-27 | 2011-12-01 | Alcatel-Lucent Usa Inc. | Technique For Determining And Indicating Strength Of An Item In A Weighted List Based On Tagging |
-
2014
- 2014-10-11 CN CN201410534723.XA patent/CN104281690B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103176961A (en) * | 2013-03-05 | 2013-06-26 | 哈尔滨工程大学 | Transfer learning method based on latent semantic analysis |
| CN103440256A (en) * | 2013-07-26 | 2013-12-11 | 中国科学院深圳先进技术研究院 | Method and device for automatically generating Chinese text label cloud |
Non-Patent Citations (2)
| Title |
|---|
| 基于云计算技术的文本可视化分析;张林泉等;《成都工业学院学报》;20140331;第90-92页 * |
| 相关关键词与相关图书标签云;黎邦群;《信息资源建设》;20130830;第11-15页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104281690A (en) | 2015-01-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105159962B (en) | Position recommends method and apparatus, resume to recommend method and apparatus, recruitment platform | |
| CN103970765B (en) | Correct mistakes model training method, device and text of one is corrected mistakes method, device | |
| CN104915410B (en) | A kind of mind map preserves and loading method, preservation and loading system | |
| MY188760A (en) | Search intention identifying method and device | |
| CN104462323B (en) | Semantic similarity calculation method, method for processing search results and device | |
| CN103729456B (en) | Microblog multi-modal sentiment analysis method based on microblog group environment | |
| CN107301200A (en) | A kind of article appraisal procedure and system analyzed based on Sentiment orientation | |
| CN103218412B (en) | Public feelings information processing method and device | |
| CN104965821B (en) | A kind of data mask method and device | |
| CN106909572A (en) | A kind of construction method and device of question and answer knowledge base | |
| CN106779053A (en) | The knowledge point of a kind of allowed for influencing factors and neutral net is known the real situation method | |
| CN106506327A (en) | A kind of spam filtering method and device | |
| CN109902157A (en) | A kind of training sample validation checking method and device | |
| CN106909573A (en) | A kind of method and apparatus for evaluating question and answer to quality | |
| CN108009248A (en) | A kind of data classification method and system | |
| CN105488098A (en) | Field difference based new word extraction method | |
| CN106126497A (en) | A kind of automatic mining correspondence executes leader section and the method for cited literature textual content fragment | |
| CN106557566B (en) | A kind of text training method and device | |
| CN104281690B (en) | A kind of label-cloud generation method and device | |
| CN103049629B (en) | A kind of method and device detecting noise data | |
| CN106909536A (en) | Method is recommended in a kind of scoring based on Heterogeneous Information | |
| CN104239314B (en) | A kind of method and system of query expansion word | |
| CN104090918B (en) | Sentence similarity calculation method based on information amount | |
| CN105893363A (en) | A method and a system for acquiring relevant knowledge points of a knowledge point | |
| El-Oujaji et al. | Difference of stability between two elite boxing groups: a preliminary study |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |