CN107102989A - A kind of entity disambiguation method based on term vector, convolutional neural networks - Google Patents
A kind of entity disambiguation method based on term vector, convolutional neural networks Download PDFInfo
- Publication number
- CN107102989A CN107102989A CN201710373502.2A CN201710373502A CN107102989A CN 107102989 A CN107102989 A CN 107102989A CN 201710373502 A CN201710373502 A CN 201710373502A CN 107102989 A CN107102989 A CN 107102989A
- Authority
- CN
- China
- Prior art keywords
- entity
- disambiguation
- candidate
- term vector
- knowledge base
- 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
- 239000013598 vector Substances 0.000 title claims abstract description 96
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 29
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000012549 training Methods 0.000 claims abstract description 21
- 239000011159 matrix material Substances 0.000 claims description 23
- 230000007935 neutral effect Effects 0.000 claims description 12
- 230000011218 segmentation Effects 0.000 claims description 7
- 230000004913 activation Effects 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- PXUQTDZNOHRWLI-OXUVVOBNSA-O malvidin 3-O-beta-D-glucoside Chemical compound COC1=C(O)C(OC)=CC(C=2C(=CC=3C(O)=CC(O)=CC=3[O+]=2)O[C@H]2[C@@H]([C@@H](O)[C@H](O)[C@@H](CO)O2)O)=C1 PXUQTDZNOHRWLI-OXUVVOBNSA-O 0.000 claims description 6
- 230000009849 deactivation Effects 0.000 claims description 2
- 210000005036 nerve Anatomy 0.000 claims description 2
- 230000008569 process Effects 0.000 claims description 2
- 238000013528 artificial neural network Methods 0.000 abstract description 4
- 238000003062 neural network model Methods 0.000 description 7
- 238000003780 insertion Methods 0.000 description 3
- 230000037431 insertion Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011524 similarity measure Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/247—Thesauruses; Synonyms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The present invention provides a kind of entity disambiguation method based on term vector, convolutional neural networks, including Entity recognition stage, Entity Semantics represent the four-stages such as stage, neural network learning training stage and entity classification stage.This method relies on the term vector and convolutional neural networks of word2vec training, respectively for treating candidate's entity summary info constructing semantic characteristic vector in disambiguation entity context and knowledge base.The cosine similarity of characteristic vector is calculated in the entity classification stage, the maximum candidate's entity of similarity is taken as the final goal entity for treating disambiguation entity.By the method for the present invention, the semantic expressiveness ability of entity is substantially increased, and then improve the accuracy rate of follow-up disambiguation.
Description
Technical field
The invention belongs to technical field of Internet information, and in particular to a kind of entity disambiguation method, more particularly to a kind of base
In term vector, the entity disambiguation method of convolutional neural networks.
Background technology
With the popularization of mobile Internet, microblogging, blog, mhkc, forum and major news websites, government work website
Etc. the life for greatly facilitating the people.The data overwhelming majority on these platforms is all with unstructured or semi-structured shape
Formula is present, and causes to there is substantial amounts of Ambiguity in these data.If it is accurate that these can be carried out containing ambiguous entity
Disambiguation, it will for later-stage utilization produce great convenience.
The entity disambiguation algorithm underlying model of current main flow is to be based on bag of words mostly, the intrinsic limitation of bag of words,
Cause these algorithms to be unable to make full use of the semantic information of context, cause entity disambiguation effect also to have greatly improved sky
Between.Word insertion is the focus of machine learning in recent years, and the core concept of word insertion is exactly to construct a distribution for each word
Formula represents that this avoid the wide gap between vocabulary and vocabulary.Convolutional neural networks are a branches of neural network model, can
Effectively to catch local feature, then global modeling.If can be modeled using convolutional neural networks to word insertion, then energy
Access semantic feature more more effective than bag of words.And based on local sensing and the shared thought of weights, convolutional Neural net
Network Model Parameter greatly reduces, and training speed is very fast, and the AlphaGo of Google core is exactly two convolutional neural networks.
The present invention combines term vector and convolutional neural networks, for treating disambiguation entity context and knowledge base entity
Summary info, respectively constructing semantic represent that training convolutional neural networks are predicted.Substantially increase the language of entity context
Adopted descriptive power.
The content of the invention
Goal of the invention：The present invention is difficult by the present situation of the semantic information of context for existing entity disambiguation method, carries
For a kind of entity disambiguation method based on term vector, convolutional neural networks, it is intended to catch context semantic information to help entity
Disambiguation.
Technical scheme：
A kind of entity disambiguation method based on term vector, convolutional neural networks, including step：
Step 1：According to application scenarios collect comprising treating the text set of disambiguation entity, and text set is pre-processed,
Determine that each in text set treats disambiguation entity and its contextual feature；
Step 2：The knowledge base of disambiguation entity is treated according to what domain knowledge was built, and searches for knowledge base, determines that each is treated
The Expressive Features of each candidate's entity in candidate's entity sets of disambiguation entity and set；
Step 3：The term vector of the noun in centered on treating disambiguation entity fixed size window is taken to be constituted one
Term vector matrix, is used as the context semantic feature for treating disambiguation entity；The summary info of the entity of each in knowledge base is taken in meter
The term vector for calculating larger preceding 20 nouns of weight ratio after TFIDF constitutes term vector matrix, is used as the semanteme of knowledge base entity
Feature；
Step 4：By known unambiguously entity joint knowledge base target entity and candidate's entity composing training in text
Gather, and be input to convolutional neural networks model and be trained, the parameter in adjustment model；
Step 5：The sample for entity and knowledge base candidate the entity sets composition for treat disambiguation to each, is input to step 4
Obtained convolutional neural networks model, respectively obtains and treats that each knowledge base is real in disambiguation entity and knowledge base candidate's entity sets
The semantic feature vector of body；
Step 6：Based on semantic feature vector, each entity in disambiguation entity and knowledge base candidate's entity sets is treated in calculating
Cosine similarity；It is the final goal entity for treating disambiguation entity to take the maximum candidate's entity of similarity.
Pretreatment in the step 1 is to carry out part-of-speech tagging to text set with Chinese Academy of Sciences Chinese word segmentation program ICTCLAS
Participle, then filters out stop words according to deactivation vocabulary, and for some proper nouns and the indiscernible physical name of comparison
Create a name dictionary.
Chinese word segmentation program ICTCLAS is called to carry out part-of-speech tagging point to the entity description in knowledge base in the step 2
Word, stop words is filtered out according to vocabulary is disabled.
The term vector of the noun in a fixed size window in the step 3 centered on treating disambiguation entity is constituted one
Individual term vector matrix is specially：
1) Google deep learning program word2vec is called to be trained Chinese wikipedia corpus, so as to obtain word
Vector table L, the length of term vector therein is 200 dimensions, and often one-dimensional is all a real number；
2) disambiguation entity e context conetxt is treatede={ w1,w2,…,wi,…,wKIn each noun wiInquiry
Term vector table L, obtains the term vector v of each nouni；
3) according to the term vector for the context words for treating disambiguation entity e, the context term vector square for treating disambiguation entity e is built
Battle array [v1,v2,v3,…vi,…,vK]；
4) terminate.
Take the summary info of the entity of each in the knowledge base weight ratio after TFIDF is calculated larger in the step 3
The term vector of preceding 20 nouns constitutes term vector matrix：
1) to candidate entity sets E={ e1,e2,…,enIn each candidate's entity eiExpressive Features in each name
Word wiQuery word vector table L, obtains the term vector v of each nouni；
2) according to the term vector of the noun of each in Expressive Features, the term vector matrix of entity description is built；
3) terminate.
The step 4 convolutional neural networks learning training is specially：
1) each treat that the semantic feature of disambiguation and the semantic feature of candidate's entity sets, as a training sample, are input to
Neural network model；
2) semantic feature for treating disambiguation carries out convolution, sets convolution kernel feature map number as 200, sets
Convolution kernel feature map size is [2,200], i.e., a length of 2, a width of 200 matrix；
3) result of each convolution kernel convolution uses 1-max ponds, obtains the feature of each convolution kernel；
4) 200 convolution kernel feature composition intermediate result, then full articulamentum is input to, full articulamentum size is 50, finally
Obtain the semantic feature vector of one 50 dimension；
5) semantic feature of candidate's entity sets, first asks and adds and average, a full articulamentum is then input to again, size is same
Sample is 50, finally gives the semantic feature vector of one 50 dimension；
6) the loss function Loss of the training sample of each in neutral neteIt is defined as：
Losse=max (0,1-sim (e, eε)+sim(e,e′))
Wherein：eεExpression treats that any other candidates in disambiguation entity e target entity, e ' expression candidate's entity setses are real
Body, means the difference for maximizing target entity and any other candidate's Entity Semantics characteristic vector similarities；
Whole loss function is defined as：Loss=∑s Losse；
7) parameter in neutral net, which is used, is uniformly distributed U (- 0.01,0.01) initialization；
8) activation primitive in neutral net uses tanh tanh activation primitives；
9) parameter in neutral net is updated using stochastic gradient descent；
10) terminate.
The step 6 entity classification stage is specially：
1) the semantic feature vector a for treating disambiguation entity e is read from file system；
2) candidate entity sets E={ e are read from file system1,e2,…,enIn the vectorial set B=of semantic feature
{b1,b2,…,bn}；
3) candidate's entity sets is traveled through, the cosine similarity of each characteristic vector in e and E is calculated
4) the maximum entity of similarity is chosenAs finally predicting the outcome；
5) terminate.
Beneficial effect：Entity disambiguation method of the invention based on term vector, convolutional neural networks treats disambiguation entity respectively
With knowledge base candidate's entity structure semantic expressiveness.Using training set training neural network model, in entity disambiguation, it will wait to disappear
Discrimination entity is input to the neural network model trained, and output treats that most like candidate's entity of disambiguation entity is real as final goal
Body.
Brief description of the drawings
In order to illustrate more clearly of the present invention, the accompanying drawing used needed for the present invention will be briefly described below：
Fig. 1 is the entity disambiguation method based on term vector, convolutional neural networks of the invention.
Fig. 2 is the structure chart of convolutional neural networks model.
Fig. 3 is the flow chart in entity classification stage.
Embodiment
The present invention is further described below in conjunction with the accompanying drawings.
The flow chart based on term vector, the entity disambiguation method of convolutional neural networks of the present invention is as shown in Figure 1.
Step 0 is the initial state of the entity disambiguation method of the present invention；
In the Entity recognition stage (step 1-6)：
Step 1 is to be collected according to application scenarios comprising the text set for treating disambiguation entity；
Step 2 is the knowledge base for treating disambiguation entity built according to domain knowledge；
Step 3 is to call Chinese Academy of Sciences Chinese word segmentation program ICTCLAS to carry out part-of-speech tagging participle to text set, then basis
Disable vocabulary and filter out stop words, and for some proper nouns and comparison one noun word of indiscernible entity name creation
Allusion quotation；
Step 4 is to call Chinese word segmentation program ICTCLAS to carry out part-of-speech tagging participle, root to the entity description in knowledge base
Stop words is filtered out according to vocabulary is disabled；
Step 5 be according to application scenarios determine each pay close attention to treat disambiguation entity and its contextual feature；
Step 6 is the generation of candidate's entity, searches for knowledge base, compares and treats that disambiguation entity is censured in item and knowledge base in text
Whether entity censures identical, if identical, it is to treat that disambiguation entity censures candidate's entity of item in text that these entities, which are regarded, really
The Expressive Features of each candidate's entity during each is treated candidate's entity sets of disambiguation entity and gathered calmly；
The stage (step 7-10) is represented in Entity Semantics：
Step 7 is that the term vector for taking the noun in a fixed size window centered on treating disambiguation entity is constituted one
Term vector matrix；(being carried out to text set after part-of-speech tagging word segmentation processing, the word with/n mark), window size is 10；
1) Google deep learning program word2vec is called to be trained Chinese wikipedia corpus, so as to obtain word
Vector table L, the length of term vector therein is 200 dimensions, and often one-dimensional is all a real number；
2) disambiguation entity e context conetxt is treatede={ w1,w2,…,wi,…,wKIn each noun wiInquiry
Term vector table L, obtains the term vector v of each nouni；
3) according to the term vector for the context words for treating disambiguation entity e, the context term vector square for treating disambiguation entity e is built
Battle array [v1,v2,v3,…vi,…,vK]；
4) terminate.
Step 8 is the summary info for taking the entity of each in knowledge base 20 before weight ratio after calculating TFIDF is larger
The term vector of individual noun constitutes term vector matrix；If inadequate 20 nouns, take existing all nouns；
1) to candidate entity sets E={ e1,e2,…,enIn each candidate's entity eiExpressive Features in each name
Word wiQuery word vector table L, obtains the term vector v of each nouni；
2) according to the term vector of the noun of each in Expressive Features, the term vector matrix of entity description is built；
3) terminate.
Step 9 is to regard the term vector matrix of step 7 as the context semantic feature for treating disambiguation entity；
Step 10 be using the term vector matrix of step 8 as knowledge base entity semantic feature；
In the neural network learning training stage (step 11-12)：
Step 11 is by known unambiguously entity joint knowledge base entity composing training set in text；
The training set that step 12 is directed in step 11 is input to convolutional neural networks model and is trained, in adjustment model
Parameter；
1) each treat that the semantic expressiveness of disambiguation and the semantic feature of candidate's entity sets, as a training sample, are input to
Neural network model；
2) semantic feature for treating disambiguation carries out convolution, sets convolution kernel feature map number as 200, sets
Convolution kernel feature map size is [2,200], i.e., a length of 2, a width of 200 matrix；
3) result of each convolution kernel convolution uses 1-max ponds, obtains the feature of each convolution kernel；
4) 200 convolution kernel feature composition intermediate result, then full articulamentum is input to, full articulamentum size is 50, finally
Obtain the semantic feature vector of one 50 dimension；
5) semantic feature of candidate's entity sets, first asks and adds and average, a full articulamentum is then input to again, size is same
Sample is 50, finally gives the semantic feature vector of one 50 dimension；
6) the loss function Loss of the training sample of each in neutral neteIt is defined as：
Losse=max (0,1-sim (e, eε)+sim(e,e′))
Wherein：eεExpression treats that any other candidates in disambiguation entity e target entity, e ' expression candidate's entity setses are real
Body, means the difference for maximizing target entity and any other candidate's Entity Semantics characteristic vector similarities；
Whole loss function is defined as：Loss=∑s Losse；
7) parameter in neutral net, which is used, is uniformly distributed U (- 0.01,0.01) initialization；
8) activation primitive in neutral net uses tanh tanh activation primitives；
9) parameter in neutral net is updated using stochastic gradient descent；
10) terminate.
In the entity classification stage (step 13-14)：
Step 13 is to read the sample set that disambiguation entity and knowledge base candidate's entity are treated in text；
The sample set that step 14 traversal step 13 is read, the convolution that step 12 training is obtained is input to by each sample
Neural network model, and output category result；
Step 15 is the end step based on term vector, the entity disambiguation method of convolutional neural networks of the present invention；
Fig. 2 is the detailed overview diagram to the neural network structure of the step 12 in the neural network learning training stage in Fig. 1,
Including following several parts：
Term vector matrix：Treat the term vector matrix of disambiguation entity context and the term vector square of knowledge base entity description feature
Battle array as convolutional neural networks input；
Convolutional layer：Disambiguation entity context term vector matrix is treated, convolution is carried out by 200 different convolution kernels, obtains
To the feature of each convolution kernel；
1-max ponds layer：1-max ponds are carried out to the feature that convolutional layer is exported, the intermediate result of one 200 dimension is obtained；
Full articulamentum：The full articulamentum that a size is 50 is connected to above-mentioned intermediate result, to knowledge base candidate's entity
Term vector adds and is averaged also one size of connection to be 50 full articulamentum, so as to obtain the semantic features vector of two 50 dimensions；
Similarity Measure：Calculate the cosine similarity of two semantic feature vectors；
Fig. 3 is the detailed process description to the step 14 in the entity classification stage in Fig. 1：
Step 16 is Fig. 3 initial state figure；
Step 17 is to read the neural network model trained in file system；
Step 18 is to read the sample set that disambiguation entity and knowledge base candidate's entity are treated in text；
Step 19 is that sample set is input into convolutional neural networks model, obtains after semantic feature vector, travels through knowledge
Storehouse candidate's entity sets, calculates the cosine similarity for the semantic feature vector for treating disambiguation entity and each candidate's entity；
Step 20 is output similarity highest entity as final goal entity；
Step 21 is Fig. 3 done state figure；
Specifically：1) the semantic feature vector a for treating disambiguation entity e is read from file system；
2) candidate entity sets E={ e are read from file system1,e2,…,enIn the vectorial set B=of semantic feature
{b1,b2,…,bn}；
3) candidate's entity sets is traveled through, the cosine similarity of each characteristic vector in e and E is calculated
4) the maximum entity of similarity is chosenAs finally predicting the outcome；
5) terminate.
In summary, present invention comprehensive utilization term vector, the method for convolutional neural networks, are treated under disambiguation physically respectively
Text and knowledge base candidate's entity summary info construct term vector matrix, are input to convolutional neural networks model.Training convolutional nerve
Parameter in network model, adjustment model.In forecast period, export most like entity and be used as target entity.Solve tradition
Bag of words in there is vocabulary wide gap, so as to the problem of semantic expressiveness scarce capacity, further increase the standard of entity disambiguation
True rate.
Described above is only the preferred embodiment of the present invention, it should be pointed out that：For the ordinary skill people of the art
For member, under the premise without departing from the principles of the invention, some improvements and modifications can also be made, these improvements and modifications also should
It is considered as protection scope of the present invention.
Claims (7)
1. a kind of entity disambiguation method based on term vector, convolutional neural networks, it is characterised in that：Including step：
Step 1：According to application scenarios collect comprising treating the text set of disambiguation entity, and text set is pre-processed, it is determined that
Each in text set treats disambiguation entity and its contextual feature；
Step 2：The knowledge base of disambiguation entity is treated according to what domain knowledge was built, and searches for knowledge base, determines that each treats disambiguation
The Expressive Features of each candidate's entity in candidate's entity sets of entity and set；
Step 3：Take the noun in centered on treating disambiguation entity fixed size window term vector constituted a word to
Moment matrix, is used as the context semantic feature for treating disambiguation entity；The summary info of the entity of each in knowledge base is taken to calculate
The term vector of larger preceding 20 nouns of weight ratio constitutes term vector matrix after TFIDF, special as the semanteme of knowledge base entity
Levy；
Step 4：By known unambiguously entity joint knowledge base target entity and candidate's entity composing training set in text,
And be input to convolutional neural networks model and be trained, the parameter in adjustment model；
Step 5：The sample for entity and knowledge base candidate the entity sets composition for treat disambiguation to each, is input to step 4 and obtains
Convolutional neural networks model, respectively obtain and treat each knowledge base entity in disambiguation entity and knowledge base candidate's entity sets
Semantic feature vector；
Step 6：Based on semantic feature vector, calculate and treat the remaining of each entity in disambiguation entity and knowledge base candidate's entity sets
String similarity；It is the final goal entity for treating disambiguation entity to take the maximum candidate's entity of similarity.
2. entity disambiguation method according to claim 1, it is characterised in that：Pretreatment in the step 1 is section in using
Institute Chinese word segmentation program ICTCLAS carries out part-of-speech tagging participle to text set, then filters out stop words according to deactivation vocabulary, and
And for some proper nouns and comparison one noun dictionary of indiscernible entity name creation.
3. entity disambiguation method according to claim 1, it is characterised in that：Chinese word segmentation program is called in the step 2
ICTCLAS carries out part-of-speech tagging participle to the entity description in knowledge base, and stop words is filtered out according to vocabulary is disabled.
4. entity disambiguation method according to claim 1, it is characterised in that：To treat disambiguation entity in the step 3
The term vector of noun in one fixed size window of the heart constitutes a term vector matrix：
1) Google deep learning program word2vec is called to be trained Chinese wikipedia corpus, so as to obtain term vector
Table L, the length of term vector therein is 200 dimensions, and often one-dimensional is all a real number；
2) disambiguation entity e context conetxt is treatede={ w1,w2,…,wi,…,wKIn each noun wiQuery word to
Scale L, obtains the term vector v of each nouni；
3) according to the term vector for the context words for treating disambiguation entity e, the context term vector matrix for treating disambiguation entity e is built
[v1,v2,v3,…vi,…,vK]；
4) terminate.
5. entity disambiguation method according to claim 1, it is characterised in that：Each in knowledge base is taken in the step 3
It is specific that the summary info of entity term vector of larger preceding 20 nouns of weight ratio after TFIDF is calculated constitutes term vector matrix
For：
1) to candidate entity sets E={ e1,e2,…,enIn each candidate's entity eiExpressive Features in each noun wi
Query word vector table L, obtains the term vector v of each nouni；
2) according to the term vector of the noun of each in Expressive Features, the term vector matrix of entity description is built；
3) terminate.
6. entity disambiguation method according to claim 1, it is characterised in that：The step 4 convolutional neural networks study instruction
Practice detailed process as follows：
1) each treat that the semantic feature of disambiguation and the semantic feature of candidate's entity sets, as a training sample, are input to nerve
Network model；
2) semantic feature for treating disambiguation carries out convolution, sets convolution kernel feature map number as 200, sets convolution
Core feature map size is [2,200], i.e., a length of 2, a width of 200 matrix；
3) result of each convolution kernel convolution uses 1-max ponds, obtains the feature of each convolution kernel；
4) 200 convolution kernel feature composition intermediate result, then be input to full articulamentum, full articulamentum size is 50, is finally given
The semantic feature vector of one 50 dimension；
5) semantic feature of candidate's entity sets, first asks and adds and average, a full articulamentum is then input to again, size is similarly
50, finally give the semantic feature vector of one 50 dimension；
6) the loss function Loss of the training sample of each in neutral neteIt is defined as：
Losse=max (0,1-sim (e, eε)+sim(e,e′))
Wherein：eεAny other candidate's entities in disambiguation entity e target entity, e ' expression candidate's entity setses, meaning are treated in expression
Think of is the difference for maximizing target entity and any other candidate's Entity Semantics characteristic vector similarities；
Whole loss function is defined as：Loss=∑s Losse；
7) parameter in neutral net, which is used, is uniformly distributed U (- 0.01,0.01) initialization；
8) activation primitive in neutral net uses tanh tanh activation primitives；
9) parameter in neutral net is updated using stochastic gradient descent；
10) terminate.
7. entity disambiguation method according to claim 1, it is characterised in that：The step 6 entity classification stage specific mistake
Journey is as follows：
1) the semantic feature vector a for treating disambiguation entity e is read from file system；
2) candidate entity sets E={ e are read from file system1,e2,…,enIn semantic feature vector set B={ b1,
b2,…,bn}；
3) candidate's entity sets is traveled through, the cosine similarity of each characteristic vector in e and E is calculated
4) the maximum entity of similarity is chosenAs finally predicting the outcome；
5) terminate.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710373502.2A CN107102989B (en) | 2017-05-24 | 2017-05-24 | Entity disambiguation method based on word vector and convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710373502.2A CN107102989B (en) | 2017-05-24 | 2017-05-24 | Entity disambiguation method based on word vector and convolutional neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107102989A true CN107102989A (en) | 2017-08-29 |
| CN107102989B CN107102989B (en) | 2020-09-29 |
Family
ID=59670296
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710373502.2A Active CN107102989B (en) | 2017-05-24 | 2017-05-24 | Entity disambiguation method based on word vector and convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107102989B (en) |
Cited By (54)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107562729A (en) * | 2017-09-14 | 2018-01-09 | 云南大学 | The Party building document representation method strengthened based on neutral net and theme |
| CN107729509A (en) * | 2017-10-23 | 2018-02-23 | 中国电子科技集团公司第二十八研究所 | The chapter similarity decision method represented based on recessive higher-dimension distributed nature |
| CN107730002A (en) * | 2017-10-13 | 2018-02-23 | 国网湖南省电力公司 | A kind of communication network shutdown remote control parameter intelligent fuzzy comparison method |
| CN108280061A (en) * | 2018-01-17 | 2018-07-13 | 北京百度网讯科技有限公司 | Text handling method based on ambiguity entity word and device |
| CN108304552A (en) * | 2018-02-01 | 2018-07-20 | 浙江大学 | A kind of name entity link method that knowledge based planting modes on sink characteristic extracts |
| CN108335731A (en) * | 2018-02-09 | 2018-07-27 | 辽宁工程技术大学 | A kind of invalid diet's recommendation method based on computer vision |
| CN108399230A (en) * | 2018-02-13 | 2018-08-14 | 上海大学 | A kind of Chinese financial and economic news file classification method based on convolutional neural networks |
| CN108446269A (en) * | 2018-03-05 | 2018-08-24 | 昆明理工大学 | A kind of Word sense disambiguation method and device based on term vector |
| CN108563766A (en) * | 2018-04-19 | 2018-09-21 | 天津科技大学 | The method and device of food retrieval |
| CN108573047A (en) * | 2018-04-18 | 2018-09-25 | 广东工业大学 | A kind of training method and device of Module of Automatic Chinese Documents Classification |
| CN108647191A (en) * | 2018-05-17 | 2018-10-12 | 南京大学 | It is a kind of based on have supervision emotion text and term vector sentiment dictionary construction method |
| CN108647785A (en) * | 2018-05-17 | 2018-10-12 | 普强信息技术（北京）有限公司 | A kind of neural network method for automatic modeling, device and storage medium |
| CN108804595A (en) * | 2018-05-28 | 2018-11-13 | 中山大学 | A kind of short text representation method based on word2vec |
| CN108805290A (en) * | 2018-06-28 | 2018-11-13 | 国信优易数据有限公司 | A kind of determination method and device of entity class |
| CN108921213A (en) * | 2018-06-28 | 2018-11-30 | 国信优易数据有限公司 | A kind of entity classification model training method and device |
| CN108920467A (en) * | 2018-08-01 | 2018-11-30 | 北京三快在线科技有限公司 | Polysemant lexical study method and device, search result display methods |
| CN108959242A (en) * | 2018-05-08 | 2018-12-07 | 中国科学院信息工程研究所 | A kind of target entity recognition methods and device based on Chinese character part of speech feature |
| CN109101579A (en) * | 2018-07-19 | 2018-12-28 | 深圳追科技有限公司 | customer service robot knowledge base ambiguity detection method |
| CN109214007A (en) * | 2018-09-19 | 2019-01-15 | 哈尔滨理工大学 | A kind of Chinese sentence meaning of a word based on convolutional neural networks disappears qi method |
| CN109241294A (en) * | 2018-08-29 | 2019-01-18 | 国信优易数据有限公司 | A kind of entity link method and device |
| CN109299462A (en) * | 2018-09-20 | 2019-02-01 | 武汉理工大学 | Short text similarity calculating method based on multidimensional convolution feature |
| CN109325108A (en) * | 2018-08-13 | 2019-02-12 | 北京百度网讯科技有限公司 | Inquiry processing method, device, server and storage medium |
| CN109614615A (en) * | 2018-12-04 | 2019-04-12 | 联想(北京)有限公司 | Methodology for Entities Matching, device and electronic equipment |
| CN109635114A (en) * | 2018-12-17 | 2019-04-16 | 北京百度网讯科技有限公司 | Method and apparatus for handling information |
| CN109740728A (en) * | 2018-12-10 | 2019-05-10 | 杭州世平信息科技有限公司 | A kind of measurement of penalty calculation method based on a variety of neural network ensembles |
| CN109933788A (en) * | 2019-02-14 | 2019-06-25 | 北京百度网讯科技有限公司 | Type determines method, apparatus, equipment and medium |
| CN110019792A (en) * | 2017-10-30 | 2019-07-16 | 阿里巴巴集团控股有限公司 | File classification method and device and sorter model training method |
| CN110555208A (en) * | 2018-06-04 | 2019-12-10 | 北京三快在线科技有限公司 | ambiguity elimination method and device in information query and electronic equipment |
| CN110569506A (en) * | 2019-09-05 | 2019-12-13 | 清华大学 | Medical named entity recognition method based on medical dictionary |
| CN110598846A (en) * | 2019-08-15 | 2019-12-20 | 北京航空航天大学 | Hierarchical recurrent neural network decoder and decoding method |
| CN110674304A (en) * | 2019-10-09 | 2020-01-10 | 北京明略软件系统有限公司 | Entity disambiguation method and device, readable storage medium and electronic equipment |
| CN110705295A (en) * | 2019-09-11 | 2020-01-17 | 北京航空航天大学 | Entity name disambiguation method based on keyword extraction |
| CN110705292A (en) * | 2019-08-22 | 2020-01-17 | 成都信息工程大学 | Entity name extraction method based on knowledge base and deep learning |
| CN110826331A (en) * | 2019-10-28 | 2020-02-21 | 南京师范大学 | Intelligent construction method of place name labeling corpus based on interactive and iterative learning |
| CN110852106A (en) * | 2019-11-06 | 2020-02-28 | 腾讯科技（深圳）有限公司 | Named entity processing method and device based on artificial intelligence and electronic equipment |
| CN110852108A (en) * | 2019-11-11 | 2020-02-28 | 中山大学 | Joint training method, apparatus and medium for entity recognition and entity disambiguation |
| CN111241298A (en) * | 2020-01-08 | 2020-06-05 | 腾讯科技（深圳）有限公司 | Information processing method, apparatus and computer readable storage medium |
| CN111241824A (en) * | 2020-01-09 | 2020-06-05 | 中国搜索信息科技股份有限公司 | Method for identifying Chinese metaphor information |
| CN111310481A (en) * | 2020-01-19 | 2020-06-19 | 百度在线网络技术（北京）有限公司 | Speech translation method, device, computer equipment and storage medium |
| CN111597804A (en) * | 2020-05-15 | 2020-08-28 | 腾讯科技（深圳）有限公司 | Entity recognition model training method and related device |
| CN111709243A (en) * | 2020-06-19 | 2020-09-25 | 南京优慧信安科技有限公司 | Knowledge extraction method and device based on deep learning |
| WO2020228376A1 (en) * | 2019-05-16 | 2020-11-19 | 华为技术有限公司 | Text processing method and model training method and apparatus |
| CN112069826A (en) * | 2020-07-15 | 2020-12-11 | 浙江工业大学 | Vertical domain entity disambiguation method fusing topic model and convolutional neural network |
| CN112100356A (en) * | 2020-09-17 | 2020-12-18 | 武汉纺织大学 | Knowledge base question-answer entity linking method and system based on similarity |
| CN112257443A (en) * | 2020-09-30 | 2021-01-22 | 华泰证券股份有限公司 | MRC-based company entity disambiguation method combined with knowledge base |
| CN112464669A (en) * | 2020-12-07 | 2021-03-09 | 宁波深擎信息科技有限公司 | Stock entity word disambiguation method, computer device and storage medium |
| CN112580351A (en) * | 2020-12-31 | 2021-03-30 | 成都信息工程大学 | Machine-generated text detection method based on self-information loss compensation |
| CN112966117A (en) * | 2020-12-28 | 2021-06-15 | 成都数之联科技有限公司 | Entity linking method |
| CN113010633A (en) * | 2019-12-20 | 2021-06-22 | 海信视像科技股份有限公司 | Information interaction method and equipment |
| CN113283236A (en) * | 2021-05-31 | 2021-08-20 | 北京邮电大学 | Entity disambiguation method in complex Chinese text |
| CN113704416A (en) * | 2021-10-26 | 2021-11-26 | 深圳市北科瑞声科技股份有限公司 | Word sense disambiguation method and device, electronic equipment and computer-readable storage medium |
| CN113761218A (en) * | 2021-04-27 | 2021-12-07 | 腾讯科技（深圳）有限公司 | Entity linking method, device, equipment and storage medium |
| WO2023202170A1 (en) * | 2022-04-21 | 2023-10-26 | 北京沃东天骏信息技术有限公司 | Product word disambiguation method and apparatus |
| CN113761218B (en) * | 2021-04-27 | 2024-05-10 | 腾讯科技（深圳）有限公司 | Method, device, equipment and storage medium for entity linking |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104572892A (en) * | 2014-12-24 | 2015-04-29 | 中国科学院自动化研究所 | Text classification method based on cyclic convolution network |
| CN106295796A (en) * | 2016-07-22 | 2017-01-04 | 浙江大学 | Entity link method based on degree of depth study |
| CN106547735A (en) * | 2016-10-25 | 2017-03-29 | 复旦大学 | The structure and using method of the dynamic word or word vector based on the context-aware of deep learning |
| CN106570170A (en) * | 2016-11-09 | 2017-04-19 | 武汉泰迪智慧科技有限公司 | Text classification and naming entity recognition integrated method and system based on depth cyclic neural network |
-
2017
- 2017-05-24 CN CN201710373502.2A patent/CN107102989B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104572892A (en) * | 2014-12-24 | 2015-04-29 | 中国科学院自动化研究所 | Text classification method based on cyclic convolution network |
| CN106295796A (en) * | 2016-07-22 | 2017-01-04 | 浙江大学 | Entity link method based on degree of depth study |
| CN106547735A (en) * | 2016-10-25 | 2017-03-29 | 复旦大学 | The structure and using method of the dynamic word or word vector based on the context-aware of deep learning |
| CN106570170A (en) * | 2016-11-09 | 2017-04-19 | 武汉泰迪智慧科技有限公司 | Text classification and naming entity recognition integrated method and system based on depth cyclic neural network |
Non-Patent Citations (1)
| Title |
|---|
| 杜婧君等: "基于中文维基百科的命名实体消歧方法", 《杭州电子科技大学学报》 * |
Cited By (81)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107562729A (en) * | 2017-09-14 | 2018-01-09 | 云南大学 | The Party building document representation method strengthened based on neutral net and theme |
| CN107562729B (en) * | 2017-09-14 | 2020-12-08 | 云南大学 | Party building text representation method based on neural network and theme enhancement |
| CN107730002A (en) * | 2017-10-13 | 2018-02-23 | 国网湖南省电力公司 | A kind of communication network shutdown remote control parameter intelligent fuzzy comparison method |
| CN107730002B (en) * | 2017-10-13 | 2020-06-02 | 国网湖南省电力公司 | Intelligent fuzzy comparison method for remote control parameters of communication gateway machine |
| CN107729509A (en) * | 2017-10-23 | 2018-02-23 | 中国电子科技集团公司第二十八研究所 | The chapter similarity decision method represented based on recessive higher-dimension distributed nature |
| CN107729509B (en) * | 2017-10-23 | 2020-07-07 | 中国电子科技集团公司第二十八研究所 | Discourse similarity determination method based on recessive high-dimensional distributed feature representation |
| CN110019792A (en) * | 2017-10-30 | 2019-07-16 | 阿里巴巴集团控股有限公司 | File classification method and device and sorter model training method |
| CN108280061B (en) * | 2018-01-17 | 2021-10-26 | 北京百度网讯科技有限公司 | Text processing method and device based on ambiguous entity words |
| CN108280061A (en) * | 2018-01-17 | 2018-07-13 | 北京百度网讯科技有限公司 | Text handling method based on ambiguity entity word and device |
| US20190220749A1 (en) * | 2018-01-17 | 2019-07-18 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Text processing method and device based on ambiguous entity words |
| US11455542B2 (en) * | 2018-01-17 | 2022-09-27 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Text processing method and device based on ambiguous entity words |
| CN108304552A (en) * | 2018-02-01 | 2018-07-20 | 浙江大学 | A kind of name entity link method that knowledge based planting modes on sink characteristic extracts |
| CN108335731A (en) * | 2018-02-09 | 2018-07-27 | 辽宁工程技术大学 | A kind of invalid diet's recommendation method based on computer vision |
| CN108399230A (en) * | 2018-02-13 | 2018-08-14 | 上海大学 | A kind of Chinese financial and economic news file classification method based on convolutional neural networks |
| CN108446269A (en) * | 2018-03-05 | 2018-08-24 | 昆明理工大学 | A kind of Word sense disambiguation method and device based on term vector |
| CN108573047A (en) * | 2018-04-18 | 2018-09-25 | 广东工业大学 | A kind of training method and device of Module of Automatic Chinese Documents Classification |
| CN108563766A (en) * | 2018-04-19 | 2018-09-21 | 天津科技大学 | The method and device of food retrieval |
| CN108959242A (en) * | 2018-05-08 | 2018-12-07 | 中国科学院信息工程研究所 | A kind of target entity recognition methods and device based on Chinese character part of speech feature |
| CN108647785A (en) * | 2018-05-17 | 2018-10-12 | 普强信息技术（北京）有限公司 | A kind of neural network method for automatic modeling, device and storage medium |
| CN108647191A (en) * | 2018-05-17 | 2018-10-12 | 南京大学 | It is a kind of based on have supervision emotion text and term vector sentiment dictionary construction method |
| CN108804595B (en) * | 2018-05-28 | 2021-07-27 | 中山大学 | Short text representation method based on word2vec |
| CN108804595A (en) * | 2018-05-28 | 2018-11-13 | 中山大学 | A kind of short text representation method based on word2vec |
| CN110555208A (en) * | 2018-06-04 | 2019-12-10 | 北京三快在线科技有限公司 | ambiguity elimination method and device in information query and electronic equipment |
| CN108921213A (en) * | 2018-06-28 | 2018-11-30 | 国信优易数据有限公司 | A kind of entity classification model training method and device |
| CN108921213B (en) * | 2018-06-28 | 2021-06-22 | 国信优易数据股份有限公司 | Entity classification model training method and device |
| CN108805290A (en) * | 2018-06-28 | 2018-11-13 | 国信优易数据有限公司 | A kind of determination method and device of entity class |
| CN109101579B (en) * | 2018-07-19 | 2021-11-23 | 深圳追一科技有限公司 | Customer service robot knowledge base ambiguity detection method |
| CN109101579A (en) * | 2018-07-19 | 2018-12-28 | 深圳追科技有限公司 | customer service robot knowledge base ambiguity detection method |
| CN108920467A (en) * | 2018-08-01 | 2018-11-30 | 北京三快在线科技有限公司 | Polysemant lexical study method and device, search result display methods |
| CN109325108B (en) * | 2018-08-13 | 2022-05-27 | 北京百度网讯科技有限公司 | Query processing method, device, server and storage medium |
| US11216618B2 (en) | 2018-08-13 | 2022-01-04 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Query processing method, apparatus, server and storage medium |
| CN109325108A (en) * | 2018-08-13 | 2019-02-12 | 北京百度网讯科技有限公司 | Inquiry processing method, device, server and storage medium |
| CN109241294A (en) * | 2018-08-29 | 2019-01-18 | 国信优易数据有限公司 | A kind of entity link method and device |
| CN109214007A (en) * | 2018-09-19 | 2019-01-15 | 哈尔滨理工大学 | A kind of Chinese sentence meaning of a word based on convolutional neural networks disappears qi method |
| CN109299462A (en) * | 2018-09-20 | 2019-02-01 | 武汉理工大学 | Short text similarity calculating method based on multidimensional convolution feature |
| CN109614615A (en) * | 2018-12-04 | 2019-04-12 | 联想(北京)有限公司 | Methodology for Entities Matching, device and electronic equipment |
| CN109740728A (en) * | 2018-12-10 | 2019-05-10 | 杭州世平信息科技有限公司 | A kind of measurement of penalty calculation method based on a variety of neural network ensembles |
| CN109740728B (en) * | 2018-12-10 | 2019-11-01 | 杭州世平信息科技有限公司 | A kind of measurement of penalty calculation method based on a variety of neural network ensembles |
| CN109635114A (en) * | 2018-12-17 | 2019-04-16 | 北京百度网讯科技有限公司 | Method and apparatus for handling information |
| CN109933788A (en) * | 2019-02-14 | 2019-06-25 | 北京百度网讯科技有限公司 | Type determines method, apparatus, equipment and medium |
| WO2020228376A1 (en) * | 2019-05-16 | 2020-11-19 | 华为技术有限公司 | Text processing method and model training method and apparatus |
| CN110598846B (en) * | 2019-08-15 | 2022-05-03 | 北京航空航天大学 | Hierarchical recurrent neural network decoder and decoding method |
| CN110598846A (en) * | 2019-08-15 | 2019-12-20 | 北京航空航天大学 | Hierarchical recurrent neural network decoder and decoding method |
| CN110705292B (en) * | 2019-08-22 | 2022-11-29 | 成都信息工程大学 | Entity name extraction method based on knowledge base and deep learning |
| CN110705292A (en) * | 2019-08-22 | 2020-01-17 | 成都信息工程大学 | Entity name extraction method based on knowledge base and deep learning |
| CN110569506A (en) * | 2019-09-05 | 2019-12-13 | 清华大学 | Medical named entity recognition method based on medical dictionary |
| CN110705295A (en) * | 2019-09-11 | 2020-01-17 | 北京航空航天大学 | Entity name disambiguation method based on keyword extraction |
| CN110705295B (en) * | 2019-09-11 | 2021-08-24 | 北京航空航天大学 | Entity name disambiguation method based on keyword extraction |
| CN110674304A (en) * | 2019-10-09 | 2020-01-10 | 北京明略软件系统有限公司 | Entity disambiguation method and device, readable storage medium and electronic equipment |
| CN110826331B (en) * | 2019-10-28 | 2023-04-18 | 南京师范大学 | Intelligent construction method of place name labeling corpus based on interactive and iterative learning |
| CN110826331A (en) * | 2019-10-28 | 2020-02-21 | 南京师范大学 | Intelligent construction method of place name labeling corpus based on interactive and iterative learning |
| CN110852106B (en) * | 2019-11-06 | 2024-05-03 | 腾讯科技（深圳）有限公司 | Named entity processing method and device based on artificial intelligence and electronic equipment |
| CN110852106A (en) * | 2019-11-06 | 2020-02-28 | 腾讯科技（深圳）有限公司 | Named entity processing method and device based on artificial intelligence and electronic equipment |
| CN110852108A (en) * | 2019-11-11 | 2020-02-28 | 中山大学 | Joint training method, apparatus and medium for entity recognition and entity disambiguation |
| CN113010633B (en) * | 2019-12-20 | 2023-01-31 | 海信视像科技股份有限公司 | Information interaction method and equipment |
| CN113010633A (en) * | 2019-12-20 | 2021-06-22 | 海信视像科技股份有限公司 | Information interaction method and equipment |
| CN111241298A (en) * | 2020-01-08 | 2020-06-05 | 腾讯科技（深圳）有限公司 | Information processing method, apparatus and computer readable storage medium |
| CN111241298B (en) * | 2020-01-08 | 2023-10-10 | 腾讯科技（深圳）有限公司 | Information processing method, apparatus, and computer-readable storage medium |
| CN111241824A (en) * | 2020-01-09 | 2020-06-05 | 中国搜索信息科技股份有限公司 | Method for identifying Chinese metaphor information |
| CN111310481A (en) * | 2020-01-19 | 2020-06-19 | 百度在线网络技术（北京）有限公司 | Speech translation method, device, computer equipment and storage medium |
| CN111597804A (en) * | 2020-05-15 | 2020-08-28 | 腾讯科技（深圳）有限公司 | Entity recognition model training method and related device |
| CN111597804B (en) * | 2020-05-15 | 2023-03-10 | 腾讯科技（深圳）有限公司 | Method and related device for training entity recognition model |
| CN111709243B (en) * | 2020-06-19 | 2023-07-07 | 南京优慧信安科技有限公司 | Knowledge extraction method and device based on deep learning |
| CN111709243A (en) * | 2020-06-19 | 2020-09-25 | 南京优慧信安科技有限公司 | Knowledge extraction method and device based on deep learning |
| CN112069826B (en) * | 2020-07-15 | 2021-12-07 | 浙江工业大学 | Vertical domain entity disambiguation method fusing topic model and convolutional neural network |
| CN112069826A (en) * | 2020-07-15 | 2020-12-11 | 浙江工业大学 | Vertical domain entity disambiguation method fusing topic model and convolutional neural network |
| CN112100356A (en) * | 2020-09-17 | 2020-12-18 | 武汉纺织大学 | Knowledge base question-answer entity linking method and system based on similarity |
| CN112257443A (en) * | 2020-09-30 | 2021-01-22 | 华泰证券股份有限公司 | MRC-based company entity disambiguation method combined with knowledge base |
| CN112257443B (en) * | 2020-09-30 | 2024-04-02 | 华泰证券股份有限公司 | MRC-based company entity disambiguation method combined with knowledge base |
| CN112464669B (en) * | 2020-12-07 | 2024-02-09 | 宁波深擎信息科技有限公司 | Stock entity word disambiguation method, computer device, and storage medium |
| CN112464669A (en) * | 2020-12-07 | 2021-03-09 | 宁波深擎信息科技有限公司 | Stock entity word disambiguation method, computer device and storage medium |
| CN112966117A (en) * | 2020-12-28 | 2021-06-15 | 成都数之联科技有限公司 | Entity linking method |
| CN112580351B (en) * | 2020-12-31 | 2022-04-19 | 成都信息工程大学 | Machine-generated text detection method based on self-information loss compensation |
| CN112580351A (en) * | 2020-12-31 | 2021-03-30 | 成都信息工程大学 | Machine-generated text detection method based on self-information loss compensation |
| CN113761218A (en) * | 2021-04-27 | 2021-12-07 | 腾讯科技（深圳）有限公司 | Entity linking method, device, equipment and storage medium |
| CN113761218B (en) * | 2021-04-27 | 2024-05-10 | 腾讯科技（深圳）有限公司 | Method, device, equipment and storage medium for entity linking |
| CN113283236B (en) * | 2021-05-31 | 2022-07-19 | 北京邮电大学 | Entity disambiguation method in complex Chinese text |
| CN113283236A (en) * | 2021-05-31 | 2021-08-20 | 北京邮电大学 | Entity disambiguation method in complex Chinese text |
| CN113704416B (en) * | 2021-10-26 | 2022-03-04 | 深圳市北科瑞声科技股份有限公司 | Word sense disambiguation method and device, electronic equipment and computer-readable storage medium |
| CN113704416A (en) * | 2021-10-26 | 2021-11-26 | 深圳市北科瑞声科技股份有限公司 | Word sense disambiguation method and device, electronic equipment and computer-readable storage medium |
| WO2023202170A1 (en) * | 2022-04-21 | 2023-10-26 | 北京沃东天骏信息技术有限公司 | Product word disambiguation method and apparatus |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107102989B (en) | 2020-09-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107102989A (en) | A kind of entity disambiguation method based on term vector, convolutional neural networks | |
| CN107133213B (en) | Method and system for automatically extracting text abstract based on algorithm | |
| US10997370B2 (en) | Hybrid classifier for assigning natural language processing (NLP) inputs to domains in real-time | |
| CN107944559B (en) | Method and system for automatically identifying entity relationship | |
| CN110032632A (en) | Intelligent customer service answering method, device and storage medium based on text similarity | |
| CN103927302B (en) | A kind of file classification method and system | |
| CN109241283A (en) | A kind of file classification method based on multi-angle capsule network | |
| CN106599029A (en) | Chinese short text clustering method | |
| CN106815252A (en) | A kind of searching method and equipment | |
| CN104778256B (en) | A kind of the quick of field question answering system consulting can increment clustering method | |
| CN113392209B (en) | Text clustering method based on artificial intelligence, related equipment and storage medium | |
| CN106407280A (en) | Query target matching method and device | |
| WO2017193685A1 (en) | Method and device for data processing in social network | |
| CN112163425A (en) | Text entity relation extraction method based on multi-feature information enhancement | |
| CN103646099A (en) | Thesis recommendation method based on multilayer drawing | |
| CN110347776A (en) | Interest point name matching process, device, equipment and storage medium | |
| CN105893362A (en) | A method for acquiring knowledge point semantic vectors and a method and a system for determining correlative knowledge points | |
| CN109582761A (en) | A kind of Chinese intelligent Answer System method of the Words similarity based on the network platform | |
| CN107092605A (en) | A kind of entity link method and device | |
| Liu et al. | Structured alignment networks for matching sentences | |
| CN110851570A (en) | Unsupervised keyword extraction method based on Embedding technology | |
| CN113761890A (en) | BERT context sensing-based multi-level semantic information retrieval method | |
| Fahrni et al. | HITS'Monolingual and Cross-lingual Entity Linking System at TAC 2013. | |
| Lu et al. | Feature words selection for knowledge-based word sense disambiguation with syntactic parsing | |
| Feifei et al. | Bert-based Siamese Network for Semantic Similarity |
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 |