CN110263321A - A kind of sentiment dictionary construction method and system - Google Patents
A kind of sentiment dictionary construction method and system Download PDFInfo
- Publication number
- CN110263321A CN110263321A CN201910372297.7A CN201910372297A CN110263321A CN 110263321 A CN110263321 A CN 110263321A CN 201910372297 A CN201910372297 A CN 201910372297A CN 110263321 A CN110263321 A CN 110263321A
- Authority
- CN
- China
- Prior art keywords
- emotion
- word
- recognition model
- emotion recognition
- weight
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/242—Dictionaries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Machine Translation (AREA)
Abstract
The present invention relates to a kind of sentiment dictionary construction method and system, the method includes the steps: the corpus of text of single statement is divided into several words;Each word being partitioned into is inputted in emotion recognition model, output obtains the weight of each word and the emotion probability value of entire sentence;The weight of each word is multiplied with the emotion probability value of entire sentence, respectively obtains the emotion score of each word, the word that emotion score is more than or equal to given threshold is added in sentiment dictionary.The present invention is a kind of method of novel building emotion word dictionary, more efficient compared to the mode of artificial constructed sentiment dictionary, solves the problems, such as that complicated artificial constructed emotion word dictionary cost is big.
Description
Technical field
The present invention relates to natural language processing technique field, in particular to a kind of sentiment dictionary construction method and system.
Background technique
Sentiment analysis refers to that text is divided into commendation, two kinds of derogatory sense by the meaning according to expressed by text and emotion information
Or several types, it is the division to text author tendentiousness and viewpoint, attitude, therefore otherwise referred to as proneness analysis.Emotion
It analyzes as a kind of special classification problem, the common problem of existing general modfel classification also has its particularity, such as emotion information
The concealment of expression, ambiguity and polarity are unobvious etc..Carrying out sentiment analysis at present, generally there are two types of modes: passing through emotion word word
Allusion quotation analysis and the method based on machine learning, wherein realization by emotion word dictionary analysis method is dependent on pre-establishing
Sentiment dictionary, and constructing sentiment dictionary majority at present is by the way of manually marking, artificial constructed dictionary needs biggish generation
Valence such as expends a large amount of human and material resources, and low efficiency.
Summary of the invention
The purpose of the present invention is to provide a kind of sentiment dictionary construction method and systems, solve artificial constructed sentiment dictionary institute
Existing low efficiency, the problem of spending human and material resources.
In order to achieve the above-mentioned object of the invention, the embodiment of the invention provides following technical schemes:
A kind of sentiment dictionary construction method, comprising the following steps:
The corpus of text of single statement is divided into several words;
Each word being partitioned into is inputted in emotion recognition model, the weight and entire sentence for obtaining each word are exported
Emotion probability value;
The weight of each word is multiplied with the emotion probability value of entire sentence, respectively obtains the emotion score of each word, it will
The word that emotion score is more than or equal to given threshold is added in sentiment dictionary as emotion word.
On the other hand, the embodiment of the invention also provides a kind of emotion words to construct system, comprises the following modules:
Single statement is divided into several words for segmenting to corpus of text by word segmentation module;
Emotion recognition module, each word for that will be partitioned into input in emotion recognition model, and output obtains each word
Weight and entire sentence emotion probability value;
Dictionary creation module respectively obtains every for the weight of each word to be multiplied with the emotion probability value of entire sentence
The word that emotion score is more than or equal to given threshold is added in sentiment dictionary the emotion score of a word as emotion word.
In another aspect, the embodiment of the present invention provides a kind of computer-readable storage including computer-readable instruction simultaneously
Medium, the computer-readable instruction make processor execute the operation in method described in the embodiment of the present invention when executed.
In another aspect, the embodiment of the present invention provides a kind of electronic equipment simultaneously, comprising: memory stores program instruction;
Processor is connected with the memory, executes the program instruction in memory, realizes in method described in the embodiment of the present invention
The step of.
Compared with prior art, the present invention is a kind of method of novel building emotion word dictionary, compared to artificial constructed
The mode of sentiment dictionary, it is more efficient, solve the problems, such as that complicated artificial constructed emotion word dictionary cost is big.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached
Figure is briefly described, it should be understood that the following drawings illustrates only certain embodiments of the present invention, therefore is not construed as pair
The restriction of range for those of ordinary skill in the art without creative efforts, can also be according to this
A little attached drawings obtain other relevant attached drawings.
Fig. 1 is a kind of flow chart of sentiment dictionary construction method described in present pre-ferred embodiments.
Fig. 2 is the flow chart of emotion recognition model training method in present pre-ferred embodiments.
Fig. 3 is the structure chart of emotion recognition model in present pre-ferred embodiments.
Fig. 4 is the schematic diagram for generating new expression in the present embodiment by Self-Attention.
Fig. 5 is that a kind of sentiment dictionary provided in the present embodiment constructs the functional block diagram of system.
Fig. 6 is the structural block diagram of a kind of electronic equipment provided in the present embodiment.
Specific embodiment
Below in conjunction with attached drawing in the embodiment of the present invention, technical solution in the embodiment of the present invention carries out clear, complete
Ground description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Usually exist
The component of the embodiment of the present invention described and illustrated in attached drawing can be arranged and be designed with a variety of different configurations herein.Cause
This, is not intended to limit claimed invention to the detailed description of the embodiment of the present invention provided in the accompanying drawings below
Range, but it is merely representative of selected embodiment of the invention.Based on the embodiment of the present invention, those skilled in the art are not doing
Every other embodiment obtained under the premise of creative work out, shall fall within the protection scope of the present invention.
Referring to Fig. 1, the present embodiment has illustratively provided a kind of sentiment dictionary construction method, comprising the following steps:
S10 segments corpus of text.In the present embodiment, corpus of text refers to single sentence, and there is example at sentence end
Such as fullstop, exclamation mark indicate the punctuation mark that sentence terminates, and will be segmented in corpus of text is exactly by complete sentence point
For several words, punctuation mark is also divided into an individual word.For example, " I is very glad today!" just it is divided into " I " " the present
It " " very " " happiness " "!" five words.If also having punctuation mark among sentence, such as comma, also the comma is divided into individually
One word.Punctuation mark is a component part of sentence, therefore in the present embodiment, by punctuation mark also as a word,
Certainly, as other embodiments, punctuation mark can also be cast out.Corpus of text is divided into word, in natural language processing
It is conventional means in technology, does not run business into particular one and state to its detailed process herein.
S20 inputs each word being partitioned into emotion recognition model, exports the weight for obtaining each word and entire language
The emotion probability value of sentence.
The weight of each word is multiplied by S30 with the emotion probability value of entire sentence, respectively obtains the emotion point of each word
The word that emotion score is more than or equal to given threshold is added in sentiment dictionary number as emotion word.The threshold value can be according to reality
Depending on situation.
Only " my lifted sentence be very glad today hereinafter!" in " happiness " word for, by emotion recognition model carry out
The weight for after emotion recognition, obtaining " happiness " is a=0.56, and the emotion probability value of entire sentence is p, then " happiness " word
Emotion score is 0.56*p.
Referring to Fig. 2, the present embodiment has illustratively provided the training method of above-mentioned emotion recognition model, this method includes
Following steps:
S101 manually marks corpus of text, and by the corpus of text after mark according to a certain percentage (as act
Example, such as 8:2) it is divided into training set and test set.It should be noted that when carrying out model training, in all corpus of text
Emotion word may be had, it is also possible to without emotion word, judged by manually whole sentence sentence be it is positive, mark into front, people
Work judges that whole sentence is that negative progress negatively marks.Front mark and negative mark herein is only two different mark shapes
Formula, to distinguish.Only as an example, for example, the sentence with positive emotional word is labeled as 1, by the language with negative emotion word
Sentence is labeled as 0.
Each sentence in training set and test set is divided into several words by S102 respectively.
S103 initializes the parameter of emotion recognition model, and whole training sets are inputted in initial emotion recognition model and are carried out
Training.
The input of whole test sets is trained obtained emotion recognition model to predict by S104 through step S103, and according to
Prediction result carries out costing bio disturbance, if loss changes greatly, is greater than given threshold, then the parameter of Optimized model, and return
Step S103, circulation execute step S103~S104;It is, for example, less than given threshold if loss variation is smaller, then training terminates, and obtains
To final emotion recognition model.
After the completion of emotion recognition model construction, emotion recognition can be carried out to corpus of text by the emotion recognition model,
Obtain the emotion probability value of the weight of each word and entire sentence in sentence.
Referring to Fig. 3, the structure of emotion recognition model can be divided into five layers from the bottom up in the present embodiment.
First layer is to receive term vector, and the term vector divides resulting several words conversion gained by sentence.For example,
" I " " today " " very " " happiness " "!" totally 5 words, 5 term vectors are obtained after conversion, input first layer.
The second layer is that the term vector (being indicated with X) of input is generated new expression by Self-Attention:
Referring to Fig. 4, only generating new expression by Self-Attention by taking this term vector of processing " I " as an example
Process is as follows:
Query by " I " as search goes to go to match with the key of words all in sentence (comprising " I ") in itself, look at
The degree of correlation is how high." I " corresponding query vector and k1 is represented with q1 and represents " I " corresponding keyvector, then is counted
The dot product for needing to calculate q1 and k1 when calculating the attention score of " I ", then carries out scaling, divided by a ruler
Scale degreeWherein dk is the dimension of a query and key vector, then recycles Softmax operation will be after scaling
As a result it is normalized to probability distribution, then just obtains weight multiplied by matrix V (all term vectors are added acquired results in sentence)
The expression Z1 of summation.
Third layer is that the term vector Z (the output result that term vector Z is the second layer) of input is passed through Self-Attention
Generate new expression:
In the present embodiment, Self-Attention twice is carried out by the term vector for dividing text sentence and is expressed,
The accuracy of expression of results can be improved.
4th layer is value layer, Z is expressed as W by Self-Attention, the power being then mapped to each Z on W
Recast is the weight a of word.For example, the weight of " I " is 0.02, the weight of " today " is 0.07, and the weight of " very " is 0.11, " high
It is emerging " weight be 0.56, "!" weight be 0.24.
As shown in figure 3, term vector Z is expressed as r by Self-Attention in third layer, in r to W process, lead to
It crosses and is added r1-r5 term vector to obtain W, similar self-attention is then done with W by r1-r5 and is operated, obtain this layer of knot
Fruit.
Layer 5 obtains the result p of the emotion recognition of entire sentence by sigmoid function.
Each word can be calculated by being multiplied with the 4th layer of obtained a in the result p of the emotion recognition of entire sentence
Emotion score is a*p, for example, the weight of " happiness " is 0.56*p.
Experimental example
Emotion recognition is carried out to certain hotel's comment data using the present embodiment above method, constructs sentiment dictionary.
As: although surface see very, the inside finishing is still very exquisite, live it is clean, it is comfortable;Go to Yixing first later
Choosing or Yixing hotel!
Above-mentioned sentence obtains following word after participle: " although ", " surface ", " seeing ", " very ", " general ", ", ",
" but ", " the inside ", " finishing ", " still ", " very ", " exquisiteness ", " ", ", ", " firmly ", " ", " clean ", ", ", " comfortable ",
";", " after ", " going ", " Yixing ", " first choice ", " still ", " Yixing ", " hotel ", "!"
After above-mentioned each word is converted to term vector, emotion recognition model is inputted, it is defeated after emotion recognition model treatment
The emotion probability value p value for obtaining entire sentence out is 0.98, and it is as shown in the table to obtain each word weight:
Finally, the word for choosing more than or equal to 0.09 is used as emotion word, i.e. selection " exquisiteness ", " clean ", " comfortable " as feelings
Feel word, is added in sentiment dictionary.
Can be seen that the above method through the invention from above-mentioned experimental example can relatively accurately extract emotion word then
Construct sentiment dictionary, and compared to artificial constructed method, efficiency can be greatlyd improve by the above method, reduce manually at
This.
Referring to Fig. 5, being based on identical inventive concept, a kind of sentiment dictionary building system is provided in the present embodiment simultaneously
It unites, the transmission direction of the arrows show data between each module shown in Fig. 5.Specifically, emotion word building system includes
With lower module:
Single statement is divided into several words for segmenting to corpus of text by word segmentation module.
Emotion recognition module, each word for that will be partitioned into input in emotion recognition model, and output obtains each word
Weight and entire sentence emotion probability value.
Dictionary creation module respectively obtains every for the weight of each word to be multiplied with the emotion probability value of entire sentence
The word that emotion score is more than or equal to given threshold is added in sentiment dictionary the emotion score of a word.
Wherein, emotion recognition module is also used to training and obtains the emotion recognition model.Emotion recognition model is specifically instructed
Practice process and please refers to flow chart and foregoing teachings shown in Fig. 3.
Aforementioned sentiment dictionary construction method is based on identical design in above-mentioned sentiment dictionary building system and the present embodiment
And propose, therefore, the related content in preceding method description is please referred to be not directed in System describe herein in place of.
As shown in fig. 6, the present embodiment provides a kind of electronic equipment simultaneously, which may include 51 He of processor
Memory 52, wherein memory 52 is coupled to processor 51.It is worth noting that, the figure is exemplary, it can also be used
The structure is supplemented or substituted to the structure of his type, realizes data extraction, report generation, communication or other function.
As shown in fig. 6, the electronic equipment can also include: input unit 53, display unit 54 and power supply 55.It is worth noting
, which is also not necessary to include all components shown in Fig. 6.In addition, electronic equipment can also include
The component being not shown in Fig. 6 can refer to the prior art.
Processor 51 is sometimes referred to as controller or operational controls, may include microprocessor or other processor devices and/
Or logic device, the processor 51 receive the operation of all parts of input and controlling electronic devices.
Wherein, memory 52 for example can be buffer, flash memory, hard disk driver, removable medium, volatile memory, it is non-easily
The property lost one of memory or other appropriate devices or a variety of, can store configuration information, the processor 51 of above-mentioned processor 51
The instruction of execution, record the information such as list data.Processor 51 can execute the program of the storage of memory 52, to realize information
Storage or processing etc..It in one embodiment, further include buffer storage in memory 52, i.e. buffer, with the intermediate letter of storage
Breath.
Input unit 53 is for example for providing each text report to processor 51.Display unit 54 is processed for showing
It is various as a result, the display unit can be for example LCD display in journey, but the present invention is not limited thereto.Power supply 55 is for being
Electronic equipment provides electric power.
The embodiment of the present invention also provides a kind of computer-readable instruction, wherein when executing described instruction in the electronic device
When, described program makes electronic equipment execute the operating procedure that the method for the present invention is included.
The embodiment of the present invention also provides a kind of storage medium for being stored with computer-readable instruction, wherein the computer can
Reading instruction makes electronic equipment execute the operating procedure that the method for the present invention is included.
Those of ordinary skill in the art may be aware that list described in conjunction with the examples disclosed in the embodiments of the present disclosure
Member and algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, in order to clearly demonstrate hardware
With the interchangeability of software, each exemplary composition and step are generally described according to function in the above description.This
A little functions are implemented in hardware or software actually, the specific application and design constraint depending on technical solution.Specially
Industry technical staff can use different methods to achieve the described function each specific application, but this realization is not
It is considered as beyond the scope of this invention.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, technical solution of the present invention is substantially
The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words
It embodies, which is stored in a storage medium, including some instructions are used so that a computer
Equipment (can be personal computer, server or the network equipment etc.) executes the complete of each embodiment the method for the present invention
Portion or part steps.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only
Memory), random access memory (RAM, RandomAccess Memory), magnetic or disk etc. are various can store journey
The medium of sequence code.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can easily think of the change or the replacement, and should all contain
Lid is within protection scope of the present invention.
Claims (7)
1. a kind of sentiment dictionary construction method, which comprises the following steps:
The corpus of text of single statement is divided into several words;
Each word being partitioned into is inputted in emotion recognition model, output obtains the weight of each word and the emotion of entire sentence
Probability value;
The weight of each word is multiplied with the emotion probability value of entire sentence, the emotion score of each word is respectively obtained, by emotion
The word that score is more than or equal to given threshold is added in sentiment dictionary as emotion word.
2. the method according to claim 1, wherein the emotion recognition model is based on Self-Attention machine
System completes emotion recognition.
3. according to the method described in claim 2, it is characterized in that, the emotion recognition model is trained by following steps
It arrives:
S101 manually marks the corpus of text of single statement, and the corpus of text after mark is drawn according to a certain percentage
It is divided into training set and test set;
Each sentence in training set and test set is divided into several words by S102 respectively;
S103 initializes the parameter of emotion recognition model, and whole training sets are inputted in initial emotion recognition model and are trained;
The input of whole test sets is trained obtained emotion recognition model to predict by S104 through step S103, and according to prediction
As a result costing bio disturbance is carried out, if loss variation is greater than given threshold, the parameter of Optimized model, and return step S103, circulation
Execute step S103~S104;If loss variation is less than given threshold, training terminates.
4. a kind of emotion word constructs system, which is characterized in that comprise the following modules:
Single statement is divided into several words for segmenting to corpus of text by word segmentation module;
Emotion recognition module, each word for that will be partitioned into input in emotion recognition model, and output obtains the power of each word
The emotion probability value of weight and entire sentence;
Dictionary creation module respectively obtains each word for the weight of each word to be multiplied with the emotion probability value of entire sentence
Emotion score, using emotion score be more than or equal to given threshold word as emotion word addition sentiment dictionary in.
5. system according to claim 4, which is characterized in that the emotion recognition module is also used to instruct by machine learning
Get the emotion recognition model.
6. a kind of computer readable storage medium including computer-readable instruction, which is characterized in that the computer-readable finger
Enable the operation for requiring processor perform claim in any the method for 1-3.
7. a kind of electronic equipment, which is characterized in that the equipment includes:
Memory stores program instruction;
Processor is connected with the memory, executes the program instruction in memory, realizes that claim 1-3 is any described
Step in method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910372297.7A CN110263321B (en) | 2019-05-06 | 2019-05-06 | Emotion dictionary construction method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910372297.7A CN110263321B (en) | 2019-05-06 | 2019-05-06 | Emotion dictionary construction method and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110263321A true CN110263321A (en) | 2019-09-20 |
CN110263321B CN110263321B (en) | 2023-06-09 |
Family
ID=67914276
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910372297.7A Active CN110263321B (en) | 2019-05-06 | 2019-05-06 | Emotion dictionary construction method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110263321B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110991167A (en) * | 2019-12-05 | 2020-04-10 | 北京理工大学 | Emotion dictionary construction method based on emotion hierarchy system |
CN111767399A (en) * | 2020-06-30 | 2020-10-13 | 平安国际智慧城市科技股份有限公司 | Emotion classifier construction method, device, equipment and medium based on unbalanced text set |
CN115796158A (en) * | 2023-02-07 | 2023-03-14 | 中国传媒大学 | Emotion dictionary construction method and device, electronic equipment and computer readable medium |
CN115796158B (en) * | 2023-02-07 | 2024-07-02 | 中国传媒大学 | Emotion dictionary construction method, emotion dictionary construction device, electronic equipment and computer readable medium |
Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20130055429A (en) * | 2011-11-18 | 2013-05-28 | 삼성전자주식회사 | Apparatus and method for emotion recognition based on emotion segment |
CN107679580A (en) * | 2017-10-21 | 2018-02-09 | 桂林电子科技大学 | A kind of isomery shift image feeling polarities analysis method based on the potential association of multi-modal depth |
JP2018073343A (en) * | 2016-11-04 | 2018-05-10 | トヨタ自動車株式会社 | Emotion estimation method |
CN108460009A (en) * | 2017-12-14 | 2018-08-28 | 中山大学 | The attention mechanism Recognition with Recurrent Neural Network text emotion analytic approach of embedded sentiment dictionary |
CN108563635A (en) * | 2018-04-04 | 2018-09-21 | 北京理工大学 | A kind of sentiment dictionary fast construction method based on emotion wheel model |
CN109243494A (en) * | 2018-10-30 | 2019-01-18 | 南京工程学院 | Childhood emotional recognition methods based on the long memory network in short-term of multiple attention mechanism |
CN109376251A (en) * | 2018-09-25 | 2019-02-22 | 南京大学 | A kind of microblogging Chinese sentiment dictionary construction method based on term vector learning model |
CN109408823A (en) * | 2018-10-31 | 2019-03-01 | 华南师范大学 | A kind of specific objective sentiment analysis method based on multi-channel model |
CN109408680A (en) * | 2018-10-08 | 2019-03-01 | 腾讯科技(深圳)有限公司 | Automatic question-answering method, device, equipment and computer readable storage medium |
CN109543722A (en) * | 2018-11-05 | 2019-03-29 | 中山大学 | A kind of emotion trend forecasting method based on sentiment analysis model |
CN109657246A (en) * | 2018-12-19 | 2019-04-19 | 中山大学 | A kind of extraction-type machine reading based on deep learning understands the method for building up of model |
CN109710761A (en) * | 2018-12-21 | 2019-05-03 | 中国标准化研究院 | The sentiment analysis method of two-way LSTM model based on attention enhancing |
-
2019
- 2019-05-06 CN CN201910372297.7A patent/CN110263321B/en active Active
Patent Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20130055429A (en) * | 2011-11-18 | 2013-05-28 | 삼성전자주식회사 | Apparatus and method for emotion recognition based on emotion segment |
JP2018073343A (en) * | 2016-11-04 | 2018-05-10 | トヨタ自動車株式会社 | Emotion estimation method |
CN107679580A (en) * | 2017-10-21 | 2018-02-09 | 桂林电子科技大学 | A kind of isomery shift image feeling polarities analysis method based on the potential association of multi-modal depth |
CN108460009A (en) * | 2017-12-14 | 2018-08-28 | 中山大学 | The attention mechanism Recognition with Recurrent Neural Network text emotion analytic approach of embedded sentiment dictionary |
CN108563635A (en) * | 2018-04-04 | 2018-09-21 | 北京理工大学 | A kind of sentiment dictionary fast construction method based on emotion wheel model |
CN109376251A (en) * | 2018-09-25 | 2019-02-22 | 南京大学 | A kind of microblogging Chinese sentiment dictionary construction method based on term vector learning model |
CN109408680A (en) * | 2018-10-08 | 2019-03-01 | 腾讯科技(深圳)有限公司 | Automatic question-answering method, device, equipment and computer readable storage medium |
CN109243494A (en) * | 2018-10-30 | 2019-01-18 | 南京工程学院 | Childhood emotional recognition methods based on the long memory network in short-term of multiple attention mechanism |
CN109408823A (en) * | 2018-10-31 | 2019-03-01 | 华南师范大学 | A kind of specific objective sentiment analysis method based on multi-channel model |
CN109543722A (en) * | 2018-11-05 | 2019-03-29 | 中山大学 | A kind of emotion trend forecasting method based on sentiment analysis model |
CN109657246A (en) * | 2018-12-19 | 2019-04-19 | 中山大学 | A kind of extraction-type machine reading based on deep learning understands the method for building up of model |
CN109710761A (en) * | 2018-12-21 | 2019-05-03 | 中国标准化研究院 | The sentiment analysis method of two-way LSTM model based on attention enhancing |
Non-Patent Citations (3)
Title |
---|
CHEN KAN 等: "AMC: Attention guided multi-modal correlation learning for image search" * |
杨鹏: "基于领域词典与机器学习的中文评论情感分析" * |
黄琼霞: "基于深度学习的文本情感分析研究" * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110991167A (en) * | 2019-12-05 | 2020-04-10 | 北京理工大学 | Emotion dictionary construction method based on emotion hierarchy system |
CN110991167B (en) * | 2019-12-05 | 2021-10-08 | 北京理工大学 | Emotion dictionary construction method based on emotion hierarchy system |
CN111767399A (en) * | 2020-06-30 | 2020-10-13 | 平安国际智慧城市科技股份有限公司 | Emotion classifier construction method, device, equipment and medium based on unbalanced text set |
CN115796158A (en) * | 2023-02-07 | 2023-03-14 | 中国传媒大学 | Emotion dictionary construction method and device, electronic equipment and computer readable medium |
CN115796158B (en) * | 2023-02-07 | 2024-07-02 | 中国传媒大学 | Emotion dictionary construction method, emotion dictionary construction device, electronic equipment and computer readable medium |
Also Published As
Publication number | Publication date |
---|---|
CN110263321B (en) | 2023-06-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Deng et al. | Deep learning in natural language processing | |
Lebret et al. | Neural text generation from structured data with application to the biography domain | |
CN106997375B (en) | Customer service reply recommendation method based on deep learning | |
Dos Santos et al. | Deep convolutional neural networks for sentiment analysis of short texts | |
CN112560479B (en) | Abstract extraction model training method, abstract extraction device and electronic equipment | |
CN108717406A (en) | Text mood analysis method, device and storage medium | |
US20130179169A1 (en) | Chinese text readability assessing system and method | |
KR101715118B1 (en) | Deep Learning Encoding Device and Method for Sentiment Classification of Document | |
Yu et al. | Combining word embeddings and feature embeddings for fine-grained relation extraction | |
Kaibi et al. | A comparative evaluation of word embeddings techniques for twitter sentiment analysis | |
CN110796160A (en) | Text classification method, device and storage medium | |
CN112668336A (en) | Task model-based word processing method | |
CN110489765A (en) | Machine translation method, device and computer readable storage medium | |
CN110263321A (en) | A kind of sentiment dictionary construction method and system | |
CN110717333B (en) | Automatic generation method and device for article abstract and computer readable storage medium | |
George et al. | Teamcen at semeval-2018 task 1: global vectors representation in emotion detection | |
CN114298031A (en) | Text processing method, computer device and storage medium | |
Duh et al. | Beyond log-linear models: Boosted minimum error rate training for n-best re-ranking | |
CN117290515A (en) | Training method of text annotation model, method and device for generating text graph | |
CN112766236A (en) | Text generation method and device, computer equipment and computer readable storage medium | |
Tannert et al. | FlowchartQA: the first large-scale benchmark for reasoning over flowcharts | |
dos Santos | Think positive: Towards Twitter sentiment analysis from scratch | |
CN114091437A (en) | New word recall method and field word vector table generating method and device | |
CN116702761A (en) | Text error correction method, device, equipment and storage medium | |
Bashynska et al. | Research and Development of a Modern Deep Learning Model for Emotional Analysis Management of Text Data |
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 |