CN109522921A - Statement similarity method of discrimination and equipment - Google Patents
Statement similarity method of discrimination and equipment Download PDFInfo
- Publication number
- CN109522921A CN109522921A CN201811088787.6A CN201811088787A CN109522921A CN 109522921 A CN109522921 A CN 109522921A CN 201811088787 A CN201811088787 A CN 201811088787A CN 109522921 A CN109522921 A CN 109522921A
- Authority
- CN
- China
- Prior art keywords
- word
- sentence
- sequence
- feature
- full articulamentum
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The object of the present invention is to provide a kind of statement similarity method of discrimination and equipment, the method that the present invention proposes the Fusion Features for extracting data mining technology and stack convolutional neural networks can be improved the accuracy of semantic similarity judgement.
Description
Technical field
The present invention relates to computer field more particularly to a kind of statement similarity method of discrimination and equipment.
Background technique
In recent years, with the rapid development in deep learning field, the relevant task of more and more natural language processings
The method using deep learning gradually is turned to from traditional way, is also obviously improved in effect.Compare in these tasks
Typically have: machine translation, text generation, emotional semantic classification, intelligent answer etc..
In existing text similarity analysis task, commonly used in the base neural net to input text vector feature extraction
There are two types of networks, and one is convolutional neural networks, another kind is Recognition with Recurrent Neural Network, while also including being based on both neural networks
Other improved complicated neural networks.
Different neural networks has respective advantage in terms of feature extraction, but simultaneously also without other neural network institutes
Some advantages, existing method only simply simply merge various neural networks to extract more features, such net
Network is computationally intensive, and the feature of extraction is also limited, causes the accuracy of analysis task not high.
Summary of the invention
It is an object of the present invention to provide a kind of statement similarity method of discrimination and equipment.
According to an aspect of the invention, there is provided a kind of statement similarity method of discrimination, this method comprises:
The sequence that two sentences of input split into corresponding word perhaps word respectively is indicated the word of each sentence or
The sequence expression of word is converted into corresponding matrix sequence；
The corresponding matrix sequence of obtained each sentence is inputted into three layers of quick stack Bilstm neural network, to obtain
The feature of the Bilstm neural network of each sentence；
Using data digging method from the corresponding matrix sequence of each sentence, the similarity feature of two sentences is extracted；
The feature for the Bilstm neural network extracted and similarity feature are spliced to obtain total eigenmatrix, and
Total eigenmatrix is inputted into full articulamentum, to obtain the output of the full articulamentum；
The output of the full articulamentum is inputted into softmax classifier, the classification exported according to the softmax classifier
As a result judge whether two sentences of the input are similar.
Further, in the above method, the expression of the sequence of the word of each sentence or word is converted into corresponding matrix sequence
Column, comprising:
According to preparatory trained word or term vector model, the expression of the sequence of the word of each sentence or word is converted into pair
The matrix sequence answered.
Further, in the above method, the data digging method includes:
Method, the method for editing distance or the tf-idf feature combination wordev character representation sentence of n meta-model similarity
The method of text calculating cos similarity.
Further, in the above method, total eigenmatrix is inputted into full articulamentum, comprising:
Reduce the dimension of total eigenmatrix in the method in maximum and average pond, and total after dimension by reducing
Eigenmatrix inputs full articulamentum.
According to another aspect of the present invention, a kind of statement similarity discriminating device is additionally provided, which includes:
First device indicates for input two sentences to be split into the sequence of corresponding word or word respectively, will be every
The sequence expression of the word or word of a sentence is converted into corresponding matrix sequence；
Second device, for the corresponding matrix sequence of obtained each sentence to be inputted three layers of quick stack Bilstm mind
Through network, to obtain the feature of the Bilstm neural network of each sentence；
3rd device, for from the corresponding matrix sequence of each sentence, extracting two sentences using data digging method
Similarity feature；
4th device, the feature and similarity feature of the Bilstm neural network for will extract are spliced to obtain total
Eigenmatrix, and the total eigenmatrix is inputted into full articulamentum, to obtain the output of the full articulamentum；
5th device, for the output of the full articulamentum to be inputted softmax classifier, according to the softmax points
The classification results of class device output judge whether two sentences of the input are similar.
Further, in above equipment, the first device is used for according to preparatory trained word or term vector model,
The expression of the sequence of the word of each sentence or word is converted into corresponding matrix sequence.
Further, in above equipment, the data digging method includes:
Method, the method for editing distance or the tf-idf feature combination wordev character representation sentence of n meta-model similarity
The method of text calculating cos similarity.
Further, in above equipment, the 4th device is described total for reducing in the method in maximum and average pond
Eigenmatrix dimension, and input full articulamentum for the total eigenmatrix after dimension is reduced.
According to another aspect of the present invention, a kind of equipment based on calculating is additionally provided, wherein include:
Processor；And
It is arranged to the memory of storage computer executable instructions, the executable instruction makes the place when executed
Manage device:
The sequence that two sentences of input split into corresponding word perhaps word respectively is indicated the word of each sentence or
The sequence expression of word is converted into corresponding matrix sequence；
The corresponding matrix sequence of obtained each sentence is inputted into three layers of quick stack Bilstm neural network, to obtain
The feature of the Bilstm neural network of each sentence；
Using data digging method from the corresponding matrix sequence of each sentence, the similarity feature of two sentences is extracted；
The feature for the Bilstm neural network extracted and similarity feature are spliced to obtain total eigenmatrix, and
Total eigenmatrix is inputted into full articulamentum, to obtain the output of the full articulamentum；
The output of the full articulamentum is inputted into softmax classifier, the classification exported according to the softmax classifier
As a result judge whether two sentences of the input are similar.
According to another aspect of the present invention, a kind of computer readable storage medium is additionally provided, computer is stored thereon with
Executable instruction, wherein the computer executable instructions make processor when being executed by processor:
The sequence that two sentences of input split into corresponding word perhaps word respectively is indicated the word of each sentence or
The sequence expression of word is converted into corresponding matrix sequence；
The corresponding matrix sequence of obtained each sentence is inputted into three layers of quick stack Bilstm neural network, to obtain
The feature of the Bilstm neural network of each sentence；
Using data digging method from the corresponding matrix sequence of each sentence, the similarity feature of two sentences is extracted；
The feature for the Bilstm neural network extracted and similarity feature are spliced to obtain total eigenmatrix, and
Total eigenmatrix is inputted into full articulamentum, to obtain the output of the full articulamentum；
The output of the full articulamentum is inputted into softmax classifier, the classification exported according to the softmax classifier
As a result judge whether two sentences of the input are similar.
Compared with prior art, the present invention proposes the feature for extracting data mining technology and stack convolutional neural networks
The method of fusion can be improved the accuracy of semantic similarity judgement.
Detailed description of the invention
By reading a detailed description of non-restrictive embodiments in the light of the attached drawings below, of the invention other
Feature, objects and advantages will become more apparent upon:
Fig. 1 shows the statement similarity method of discrimination of one aspect according to the present invention and the schematic diagram of equipment.
The same or similar appended drawing reference represents the same or similar component in attached drawing.
Specific embodiment
Present invention is further described in detail with reference to the accompanying drawing.
In a typical configuration of this application, terminal, the equipment of service network and trusted party include one or more
Processor (CPU), input/output interface, network interface and memory.
Memory may include the non-volatile memory in computer-readable medium, random access memory (RAM) and/or
The forms such as Nonvolatile memory, such as read-only memory (ROM) or flash memory (flashRAM).Memory is showing for computer-readable medium
Example.
Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any method
Or technology come realize information store.Information can be computer readable instructions, data structure, the module of program or other data.
The example of the storage medium of computer includes, but are not limited to phase change memory (PRAM), static random access memory (SRAM), moves
State random access memory (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), electric erasable
Programmable read only memory (EEPROM), flash memory or other memory techniques, read-only disc read only memory (CD-ROM) (CD-ROM),
Digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or
Any other non-transmission medium, can be used for storage can be accessed by a computing device information.As defined in this article, computer
Readable medium does not include non-temporary computer readable media (transitory media), such as the data-signal and carrier wave of modulation.
As shown in Figure 1, the present invention provides a kind of statement similarity method of discrimination, comprising:
Step S1 indicates the sequence that two sentences of input split into corresponding word or word respectively, by each sentence
Word or word sequence expression be converted into corresponding matrix sequence；
Here, first having to the sequence that discrete character string is switched to a vector row；
The sequence of the word of each sentence or word can be indicated to convert according to preparatory trained word or term vector model
At corresponding matrix sequence；
By step S1, the natural language sentences text conversion that can enter text at corresponding matrix sequence, so as to
The processing of subsequent step；
Step S2, the corresponding matrix sequence of each sentence that step S1 is obtained input three layers of quick stack Bilstm mind
Through network, to obtain the feature of the Bilstm neural network of each sentence；
Here, step S2 utilizes three layers of quick stack Bilstm neural network, the front and back language of each sentence is sufficiently extracted
Adopted relationship characteristic；
Step S3, step S2 is arranged side by side with step S3, using data digging method from the corresponding matrix sequence of each sentence,
Extract the similarity feature of two sentences；
Here, the corresponding matrix sequence of each sentence obtained according to step S1, and two are extracted using data digging method
The similarity feature of a sentence；
The data digging method includes: the method for n meta-model similarity, the method for editing distance or tf-idf feature knot
Close the method that wordev character representation sentence text calculates cos similarity；
Step S4 carries out the feature for the Bilstm neural network extracted in step S2 and step S3 and similarity feature
Splicing obtains total eigenmatrix, and total eigenmatrix is inputted full articulamentum, to obtain the defeated of the full articulamentum
Out；
Here, total eigenmatrix is inputted full articulamentum, may include:
Reduce the dimension of total eigenmatrix in the method in maximum and average pond, and total after dimension by reducing
Eigenmatrix inputs full articulamentum；
The output of the full articulamentum is inputted softmax classifier by step S5, defeated according to the softmax classifier
Classification results out judge whether two sentences of the input are similar.
Here, the present invention proposes the side for the Fusion Features for extracting data mining technology and stack convolutional neural networks
Method can be improved the accuracy of semantic similarity judgement.
According to another aspect of the present invention, a kind of statement similarity discriminating device is additionally provided, which includes:
First device indicates for input two sentences to be split into the sequence of corresponding word or word respectively, will be every
The sequence expression of the word or word of a sentence is converted into corresponding matrix sequence；
Second device, for the corresponding matrix sequence of obtained each sentence to be inputted three layers of quick stack Bilstm mind
Through network, to obtain the feature of the Bilstm neural network of each sentence；
3rd device, for from the corresponding matrix sequence of each sentence, extracting two sentences using data digging method
Similarity feature；
4th device, the feature and similarity feature of the Bilstm neural network for will extract are spliced to obtain total
Eigenmatrix, and the total eigenmatrix is inputted into full articulamentum, to obtain the output of the full articulamentum；
5th device, for the output of the full articulamentum to be inputted softmax classifier, according to the softmax points
The classification results of class device output judge whether two sentences of the input are similar.
Further, in above equipment, the first device is used for according to preparatory trained word or term vector model,
The expression of the sequence of the word of each sentence or word is converted into corresponding matrix sequence.
Further, in above equipment, the data digging method includes:
Method, the method for editing distance or the tf-idf feature combination wordev character representation sentence of n meta-model similarity
The method of text calculating cos similarity.
Further, in above equipment, the 4th device is described total for reducing in the method in maximum and average pond
Eigenmatrix dimension, and input full articulamentum for the total eigenmatrix after dimension is reduced.
According to another aspect of the present invention, a kind of equipment based on calculating is additionally provided, wherein include:
Processor；And
It is arranged to the memory of storage computer executable instructions, the executable instruction makes the place when executed
Manage device:
The sequence that two sentences of input split into corresponding word perhaps word respectively is indicated the word of each sentence or
The sequence expression of word is converted into corresponding matrix sequence；
The corresponding matrix sequence of obtained each sentence is inputted into three layers of quick stack Bilstm neural network, to obtain
The feature of the Bilstm neural network of each sentence；
Using data digging method from the corresponding matrix sequence of each sentence, the similarity feature of two sentences is extracted；
The feature for the Bilstm neural network extracted and similarity feature are spliced to obtain total eigenmatrix, and
Total eigenmatrix is inputted into full articulamentum, to obtain the output of the full articulamentum；
The output of the full articulamentum is inputted into softmax classifier, the classification exported according to the softmax classifier
As a result judge whether two sentences of the input are similar.
According to another aspect of the present invention, a kind of computer readable storage medium is additionally provided, computer is stored thereon with
Executable instruction, wherein the computer executable instructions make processor when being executed by processor:
The sequence that two sentences of input split into corresponding word perhaps word respectively is indicated the word of each sentence or
The sequence expression of word is converted into corresponding matrix sequence；
The corresponding matrix sequence of obtained each sentence is inputted into three layers of quick stack Bilstm neural network, to obtain
The feature of the Bilstm neural network of each sentence；
Using data digging method from the corresponding matrix sequence of each sentence, the similarity feature of two sentences is extracted；
The feature for the Bilstm neural network extracted and similarity feature are spliced to obtain total eigenmatrix, and
Total eigenmatrix is inputted into full articulamentum, to obtain the output of the full articulamentum；
The output of the full articulamentum is inputted into softmax classifier, the classification exported according to the softmax classifier
As a result judge whether two sentences of the input are similar.
The detailed content of each equipment and storage medium embodiment of the invention, for details, reference can be made to the correspondences of each method embodiment
Part, here, repeating no more.
Obviously, those skilled in the art can carry out various modification and variations without departing from the essence of the application to the application
Mind and range.In this way, if these modifications and variations of the application belong to the range of the claim of this application and its equivalent technologies
Within, then the application is also intended to include these modifications and variations.
It should be noted that the present invention can be carried out in the assembly of software and/or software and hardware, for example, can adopt
With specific integrated circuit (ASIC), general purpose computer or any other realized similar to hardware device.In one embodiment
In, software program of the invention can be executed to implement the above steps or functions by processor.Similarly, of the invention
Software program (including relevant data structure) can be stored in computer readable recording medium, for example, RAM memory,
Magnetic or optical driver or floppy disc and similar devices.In addition, some of the steps or functions of the present invention may be implemented in hardware, example
Such as, as the circuit cooperated with processor thereby executing each step or function.
In addition, a part of the invention can be applied to computer program product, such as computer program instructions, when its quilt
When computer executes, by the operation of the computer, it can call or provide according to the method for the present invention and/or technical solution.
And the program instruction of method of the invention is called, it is possibly stored in fixed or moveable recording medium, and/or pass through
Broadcast or the data flow in other signal-bearing mediums and transmitted, and/or be stored according to described program instruction operation
In the working storage of computer equipment.Here, according to one embodiment of present invention including a device, which includes using
Memory in storage computer program instructions and processor for executing program instructions, wherein when the computer program refers to
When enabling by processor execution, method and/or skill of the device operation based on aforementioned multiple embodiments according to the present invention are triggered
Art scheme.
It is obvious to a person skilled in the art that invention is not limited to the details of the above exemplary embodiments, Er Qie
In the case where without departing substantially from spirit or essential attributes of the invention, the present invention can be realized in other specific forms.Therefore, no matter
From the point of view of which point, the present embodiments are to be considered as illustrative and not restrictive, and the scope of the present invention is by appended power
Benefit requires rather than above description limits, it is intended that all by what is fallen within the meaning and scope of the equivalent elements of the claims
Variation is included in the present invention.Any reference signs in the claims should not be construed as limiting the involved claims.This
Outside, it is clear that one word of " comprising " does not exclude other units or steps, and odd number is not excluded for plural number.That states in device claim is multiple
Unit or device can also be implemented through software or hardware by a unit or device.The first, the second equal words are used to table
Show title, and does not indicate any particular order.
Claims (10)
1. a kind of statement similarity method of discrimination, wherein this method comprises:
The sequence that two sentences of input split into corresponding word perhaps word respectively is indicated the word of each sentence or word
Sequence expression is converted into corresponding matrix sequence；
The corresponding matrix sequence of obtained each sentence is inputted into three layers of quick stack Bilstm neural network, it is each to obtain
The feature of the Bilstm neural network of sentence；
Using data digging method from the corresponding matrix sequence of each sentence, the similarity feature of two sentences is extracted；
The feature for the Bilstm neural network extracted and similarity feature are spliced to obtain total eigenmatrix, and by institute
It states total eigenmatrix and inputs full articulamentum, to obtain the output of the full articulamentum；
The output of the full articulamentum is inputted into softmax classifier, the classification results exported according to the softmax classifier
Judge whether two sentences of the input are similar.
2. according to the method described in claim 1, wherein, the expression of the sequence of the word of each sentence or word is converted into corresponding
Matrix sequence, comprising:
According to preparatory trained word or term vector model, the expression of the sequence of the word of each sentence or word is converted into corresponding
Matrix sequence.
3. according to the method described in claim 1, wherein, the data digging method includes:
Method, the method for editing distance or the tf-idf feature combination wordev character representation sentence text of n meta-model similarity
The method for calculating cos similarity.
4. according to the method described in claim 1, wherein, total eigenmatrix is inputted full articulamentum, comprising:
It reduces the dimension of total eigenmatrix in the method in maximum and average pond, and total feature after dimension will be reduced
The full articulamentum of Input matrix.
5. a kind of statement similarity discriminating device, wherein the equipment includes:
First device is indicated for input two sentences to be split into the sequence of corresponding word or word respectively, by each sentence
The sequence expression of the word or word of son is converted into corresponding matrix sequence；
Second device, for the corresponding matrix sequence of obtained each sentence to be inputted three layers of quick stack Bilstm nerve net
Network, to obtain the feature of the Bilstm neural network of each sentence；
3rd device, for from the corresponding matrix sequence of each sentence, extracting the phase of two sentences using data digging method
Like degree feature；
4th device, the feature and similarity feature of the Bilstm neural network for will extract are spliced to obtain total spy
Matrix is levied, and total eigenmatrix is inputted into full articulamentum, to obtain the output of the full articulamentum；
5th device, for the output of the full articulamentum to be inputted softmax classifier, according to the softmax classifier
The classification results of output judge whether two sentences of the input are similar.
6. equipment according to claim 5, wherein the first device, for according to preparatory trained word or word to
Model is measured, the expression of the sequence of the word of each sentence or word is converted into corresponding matrix sequence.
7. equipment according to claim 5, wherein the data digging method includes:
Method, the method for editing distance or the tf-idf feature combination wordev character representation sentence text of n meta-model similarity
The method for calculating cos similarity.
8. equipment according to claim 5, wherein the 4th device, for the method drop with maximum and average pond
The dimension of low total eigenmatrix, and total eigenmatrix after reduction dimension is inputted into full articulamentum.
9. a kind of equipment based on calculating, wherein include:
Processor；And
It is arranged to the memory of storage computer executable instructions, the executable instruction makes the processing when executed
Device:
The sequence that two sentences of input split into corresponding word perhaps word respectively is indicated the word of each sentence or word
Sequence expression is converted into corresponding matrix sequence；
The corresponding matrix sequence of obtained each sentence is inputted into three layers of quick stack Bilstm neural network, it is each to obtain
The feature of the Bilstm neural network of sentence；
Using data digging method from the corresponding matrix sequence of each sentence, the similarity feature of two sentences is extracted；
The feature for the Bilstm neural network extracted and similarity feature are spliced to obtain total eigenmatrix, and by institute
It states total eigenmatrix and inputs full articulamentum, to obtain the output of the full articulamentum；
The output of the full articulamentum is inputted into softmax classifier, the classification results exported according to the softmax classifier
Judge whether two sentences of the input are similar.
10. a kind of computer readable storage medium, is stored thereon with computer executable instructions, wherein the computer is executable
Instruction makes the processor when being executed by processor:
The sequence that two sentences of input split into corresponding word perhaps word respectively is indicated the word of each sentence or word
Sequence expression is converted into corresponding matrix sequence；
The corresponding matrix sequence of obtained each sentence is inputted into three layers of quick stack Bilstm neural network, it is each to obtain
The feature of the Bilstm neural network of sentence；
Using data digging method from the corresponding matrix sequence of each sentence, the similarity feature of two sentences is extracted；
The feature for the Bilstm neural network extracted and similarity feature are spliced to obtain total eigenmatrix, and by institute
It states total eigenmatrix and inputs full articulamentum, to obtain the output of the full articulamentum；
The output of the full articulamentum is inputted into softmax classifier, the classification results exported according to the softmax classifier
Judge whether two sentences of the input are similar.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811088787.6A CN109522921A (en) | 2018-09-18 | 2018-09-18 | Statement similarity method of discrimination and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811088787.6A CN109522921A (en) | 2018-09-18 | 2018-09-18 | Statement similarity method of discrimination and equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109522921A true CN109522921A (en) | 2019-03-26 |
Family
ID=65771041
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811088787.6A Pending CN109522921A (en) | 2018-09-18 | 2018-09-18 | Statement similarity method of discrimination and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109522921A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110134852A (en) * | 2019-05-06 | 2019-08-16 | 北京四维图新科技股份有限公司 | A kind of De-weight method of document, equipment and readable medium |
| CN114781496A (en) * | 2022-04-01 | 2022-07-22 | 北京百度网讯科技有限公司 | Optimizing sampling method and device and electronic equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105469104A (en) * | 2015-11-03 | 2016-04-06 | 小米科技有限责任公司 | Text information similarity calculating method, device and server |
| CN107291699A (en) * | 2017-07-04 | 2017-10-24 | 湖南星汉数智科技有限公司 | A kind of sentence semantic similarity computational methods |
| US20180114097A1 (en) * | 2015-10-06 | 2018-04-26 | Adobe Systems Incorporated | Font Attributes for Font Recognition and Similarity |
| CN108182275A (en) * | 2018-01-24 | 2018-06-19 | 上海互教教育科技有限公司 | A kind of mathematics variant training topic supplying system and correlating method |
| CN108509408A (en) * | 2017-02-27 | 2018-09-07 | 芋头科技（杭州）有限公司 | A kind of sentence similarity judgment method |
-
2018
- 2018-09-18 CN CN201811088787.6A patent/CN109522921A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180114097A1 (en) * | 2015-10-06 | 2018-04-26 | Adobe Systems Incorporated | Font Attributes for Font Recognition and Similarity |
| CN105469104A (en) * | 2015-11-03 | 2016-04-06 | 小米科技有限责任公司 | Text information similarity calculating method, device and server |
| CN108509408A (en) * | 2017-02-27 | 2018-09-07 | 芋头科技（杭州）有限公司 | A kind of sentence similarity judgment method |
| CN107291699A (en) * | 2017-07-04 | 2017-10-24 | 湖南星汉数智科技有限公司 | A kind of sentence semantic similarity computational methods |
| CN108182275A (en) * | 2018-01-24 | 2018-06-19 | 上海互教教育科技有限公司 | A kind of mathematics variant training topic supplying system and correlating method |
Non-Patent Citations (1)
| Title |
|---|
| 谢晓超: "基于深度神经网络的句子相似度研究与实现", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110134852A (en) * | 2019-05-06 | 2019-08-16 | 北京四维图新科技股份有限公司 | A kind of De-weight method of document, equipment and readable medium |
| CN110134852B (en) * | 2019-05-06 | 2021-05-28 | 北京四维图新科技股份有限公司 | Document duplicate removal method and device and readable medium |
| CN114781496A (en) * | 2022-04-01 | 2022-07-22 | 北京百度网讯科技有限公司 | Optimizing sampling method and device and electronic equipment |
| CN114781496B (en) * | 2022-04-01 | 2023-11-07 | 北京百度网讯科技有限公司 | Optimizing sampling method and device and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112685565B (en) | Text classification method based on multi-mode information fusion and related equipment thereof | |
| EP3926531A1 (en) | Method and system for visio-linguistic understanding using contextual language model reasoners | |
| EP3637295B1 (en) | Risky address identification method and apparatus, and electronic device | |
| US10740678B2 (en) | Concept hierarchies | |
| KR101657495B1 (en) | Image recognition method using deep learning analysis modular systems | |
| CN104735468B (en) | A kind of method and system that image is synthesized to new video based on semantic analysis | |
| CN111159412B (en) | Classification method, classification device, electronic equipment and readable storage medium | |
| CN111885000B (en) | Network attack detection method, system and device based on graph neural network | |
| CN106599317B (en) | Test data processing method, device and the terminal of question answering system | |
| JP2018190188A (en) | Summary creating device, summary creating method and computer program | |
| US10108661B2 (en) | Using synthetic events to identify complex relation lookups | |
| CN111143551A (en) | Text preprocessing method, classification method, device and equipment | |
| US20140236991A1 (en) | Parsing and Rendering Structured Images | |
| CN112199473A (en) | Multi-turn dialogue method and device in knowledge question-answering system | |
| US20200364216A1 (en) | Method, apparatus and storage medium for updating model parameter | |
| US20180018392A1 (en) | Topic identification based on functional summarization | |
| CN110674297B (en) | Public opinion text classification model construction method, public opinion text classification device and public opinion text classification equipment | |
| CN112784009A (en) | Subject term mining method and device, electronic equipment and storage medium | |
| CN110851797A (en) | Block chain-based work creation method and device and electronic equipment | |
| US10558760B2 (en) | Unsupervised template extraction | |
| CN110489559A (en) | A kind of file classification method, device and storage medium | |
| CN109522921A (en) | Statement similarity method of discrimination and equipment | |
| CN109299276A (en) | One kind converting the text to word insertion, file classification method and device | |
| US10229156B2 (en) | Using priority scores for iterative precision reduction in structured lookups for questions | |
| CN112765976A (en) | Text similarity calculation method, device and equipment and storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190326 |