CN112287240A - Case microblog evaluation object extraction method and device based on double-embedded multilayer convolutional neural network - Google Patents
Case microblog evaluation object extraction method and device based on double-embedded multilayer convolutional neural network Download PDFInfo
- Publication number
- CN112287240A CN112287240A CN202011006914.0A CN202011006914A CN112287240A CN 112287240 A CN112287240 A CN 112287240A CN 202011006914 A CN202011006914 A CN 202011006914A CN 112287240 A CN112287240 A CN 112287240A
- Authority
- CN
- China
- Prior art keywords
- microblog
- case
- case microblog
- field
- layer
- 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
- 238000011156 evaluation Methods 0.000 title claims abstract description 52
- 238000000605 extraction Methods 0.000 title claims abstract description 27
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 21
- 239000013598 vector Substances 0.000 claims abstract description 33
- 230000008451 emotion Effects 0.000 claims abstract description 20
- 238000004458 analytical method Methods 0.000 claims abstract description 16
- 238000012549 training Methods 0.000 claims abstract description 8
- 238000000034 method Methods 0.000 claims description 13
- 238000001914 filtration Methods 0.000 claims description 12
- 238000012216 screening Methods 0.000 claims description 12
- 238000012512 characterization method Methods 0.000 claims description 3
- 230000009193 crawling Effects 0.000 claims description 3
- 238000011176 pooling Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 2
- 238000003058 natural language processing Methods 0.000 abstract description 2
- 238000002474 experimental method Methods 0.000 description 5
- 230000002996 emotional effect Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 230000009977 dual effect Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 206010011469 Crying Diseases 0.000 description 1
- 238000002679 ablation Methods 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000001815 facial effect Effects 0.000 description 1
- 238000003780 insertion Methods 0.000 description 1
- 230000037431 insertion Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
- G06F16/355—Class or cluster creation or modification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- 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
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- 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)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to a case microblog evaluation object extraction method and device based on a double-embedded multilayer convolutional neural network, and belongs to the technical field of natural language processing. Constructing a case microblog aspect level emotion analysis data set; an embedded layer with case field characteristics is obtained by pre-training case microblog comment texts; and obtaining the representation results of evaluation objects in different fields through the embedded layers of the case microblog field and the general field characteristics, performing splicing operation to obtain a new vector, and marking the sequence by using a classifier to extract the case microblog evaluation objects. According to the case microblog aspect emotion analysis method and device, the efficiency of the case microblog object extraction task is improved, the accuracy of the case microblog object extraction task in the field of case microblog is improved, the case microblog aspect emotion analysis method and device serve as a first step of case microblog aspect emotion analysis, and a tamping support foundation is laid for the next tasks.
Description
Technical Field
The invention relates to a case microblog evaluation object extraction method and device based on a double-embedded multilayer convolutional neural network, and belongs to the technical field of natural language processing.
Background
Case microblog refers to internet microblog focusing on hot events related to cases. Compared with a common news microblog, the case microblog can send out hot topics related to the case in a short time, and social public sentiment outbreaks are caused.
The evaluation object extraction is an important research task in emotion tendency analysis and is also a precondition of emotion tendency analysis at an evaluation object level. The evaluation object represents the object of which the emotion is commented and discussed in the text. When the evaluation objects in the traditional comment corpus are extracted, the positions of the evaluation objects in the sentences are obvious, and the difficulty in identification and extraction is not very high.
However, the comment objects of case microblog comment texts are complex and are not limited to a certain meaning in different fields. The evaluation object may be composed of single or multiple words, and a sequence labeling problem can be modeled for the extraction task of case microblog evaluation objects. The related information of the public sentiment events is processed in time, an evaluation object is extracted from the acquired public sentiment data, the facial emotional tendency is analyzed, the public sentiment is guided correctly, and the negative influence caused by the public sentiment events can be effectively reduced.
Disclosure of Invention
The invention provides a case microblog evaluation object extraction method and device based on a double-embedded multilayer convolutional neural network, which are used for extracting case microblog evaluation objects and solving the problems of low case microblog evaluation object extraction accuracy, insufficient single word embedding recognition capability and the like.
In a first aspect, the invention provides a case microblog evaluation object extraction method based on a double-embedded multilayer convolutional neural network, which comprises the following steps:
step1, constructing a case microblog aspect level emotion analysis data set;
step2, pre-training a case microblog comment text to obtain an embedded layer with case field characteristics;
step3, obtaining the characterization results of the evaluation objects in different fields through the embedding layers of the case microblog field and the general field characteristics, performing splicing operation to obtain new vectors, and marking the sequences by using a classifier to extract the case microblog evaluation objects.
As a further scheme of the invention, the specific steps of Step1 are as follows:
step1.1, crawling the microblog comments of the relevant cases from the Xinlang microblog by using a crawler based on a Scapy frame;
step1.2, filtering and screening case microblog comments to construct a case microblog data set, and finally obtaining a case microblog data set;
step1.3, filtering and screening case microblog comments to construct a case microblog data set, and finally obtaining a case microblog database;
step1.4, manually marking comment data of the case microblog data set, and marking contained data by taking one microblog comment as a unit; the marked content is an evaluation object in the case microblog comment sentence and the corresponding emotion polarity.
In a further embodiment of the present invention, in Step1.2, the filtration and screening method is as follows:
step1.2.1, dividing the microblog messages according to a forwarding relation '//', and ensuring that comments below the forwarded microblog are analyzed based on the original microblog;
step1.2.2, delete the structure of "@ + username + reply" in the microblog comment, and delete irrelevant hyperlink advertisement;
step1.2.3, replacing the continuously appearing multiple punctuations by adopting first punctuations, and removing emoticons in the contents of the microblog comments;
step1.2.4, filtering and screening out comment data with less than seven characters, and ensuring the completeness and the usability of comment contents.
As a further scheme of the invention, the specific steps of Step3 are as follows:
step3.1, inputting the viewpoint sentence into an embedding layer in the case microblog field to obtain word vector representation in the case microblog field;
step3.2, inputting the viewpoint sentence into an embedding layer of the general field to obtain word vector representation of the general field;
step3.3, after the expression of the two field word vectors is obtained, splicing the two field word vectors to be used as the input of a convolutional neural network, and extracting local information of the comments in different ranges by convolutional kernels with different sizes;
step3.4, after the operation of the convolutional layer, in order to ensure that the words in the sentence can be extracted to the features, the words are directly input into the fully-connected layer and the softmax layer without the pooling layer, the fully-connected layer can further quantize the features after the sampling of the convolutional layer, and the softmax layer converts the results of the fully-connected layer into the probability of each label.
In a second aspect, the invention provides a case microblog evaluation object extraction device based on a double-embedded multilayer convolutional neural network, which comprises modules for executing the method of the first aspect.
The invention has the beneficial effects that:
1. the word embedding of case microblog comments in different fields is utilized to obtain the representation of evaluation objects in different fields, so that the evaluation objects are extracted from the acquired public sentiment data, and support is provided for the subsequent aspect-level sentiment tendency analysis;
2. the invention provides a double-embedded innovation, and the multilayer convolutional neural network is used as a main model, so that the efficiency of a case microblog object extraction task is greatly improved, the accuracy of an evaluation object extraction task in the case microblog field is improved, and as a first step of case microblog aspect level emotion analysis, a tamping support basis is laid for the next task.
Drawings
Fig. 1 is a schematic diagram of a specific structure of the recognition model in the present invention.
Detailed Description
Example 1: as shown in fig. 1, in a first aspect, the present invention provides a case microblog evaluation object extraction method based on a dual-embedded multilayer convolutional neural network, where the method includes:
step1, constructing a case microblog aspect level emotion analysis data set; the method comprises the following specific steps:
step1.1, crawling the microblog comments of the relevant cases from the Xinlang microblog by using a crawler based on a Scapy frame;
step1.2, filtering and screening case microblog comments to construct a case microblog data set, and finally obtaining a case microblog data set; the mode of filtration screening is as follows:
step1.2.1, dividing the microblog messages according to a forwarding relation '//', and ensuring that comments below the forwarded microblog are analyzed based on the original microblog;
step1.2.2, delete the structure of "@ + username + reply" in the microblog comment, and delete irrelevant hyperlink advertisement;
step1.2.3, replacing the continuously appearing multiple punctuations by adopting first punctuations, and removing emoticons in the contents of the microblog comments;
step1.2.4, filtering and screening out comment data with less than seven characters, and ensuring the completeness and the usability of comment contents.
Step1.3, filtering and screening case microblog comments to construct a case microblog data set, and finally obtaining a case microblog database;
step1.4, manually marking comment data of the case microblog data set, and marking contained data by taking one microblog comment as a unit; the marked content is an evaluation object in the case microblog comment sentence and the corresponding emotion polarity.
Data annotation examples:
the 'original text' has high right maintaining cost, low illegal cost, no tearing when trying to cry and no crying complaint. The term "n", "emotion" [ "NEG" ], [ { "aspect1_ content" ] "[" dimension cost "," aspect1_ start ": 0", "aspect1_ end": 3"," aspect _ polarity ": NEG" }, { "as-aspect 2_ content": illegal cost "," aspect2_ start ": 6", "aspect2_ end": 9"," aspect _ polarity ": NEG" } not known
Annotation of the label:
"aspect _ content": evaluating the object text;
"aspect _ start": a start position of an evaluation target;
"aspect _ end": an end position of the evaluation target;
"aspect _ polarity": evaluating the emotional polarity of the object;
"original text": case microblog comment text;
"emotion": case microblog comment text emotion polarity.
Taking this comment as an example, the sentence includes two evaluation objects, which are "right maintenance cost" and "illegal cost", respectively, the emotional polarities of the two evaluation objects are "NEG", respectively, and the emotional polarity of the whole comment is [ "NEG" ].
For example, for two cases of a # bus case in a certain city # and a # driver maintenance right #, case microblog comments are crawled and screened from a microblog and a case microblog data set is constructed. The data set related information is shown in table 1 below.
TABLE 1 Emotion analysis data set distribution of bus case and Benz case evaluation object level
Step2, pre-training a case microblog comment text to obtain an embedded layer with case field characteristics: training all obtained case microblog comments by using word2vec to obtain word embedding with case field characteristics;
step3, obtaining the characterization results of the evaluation objects in different fields through the embedding layers of the case microblog field and the general field characteristics, performing splicing operation to obtain new vectors, and marking the sequences by using a classifier to extract the case microblog evaluation objects. The method comprises the following specific steps:
step3.1, inputting the viewpoint sentence into an embedding layer in the case microblog field to obtain word vector representation in the case microblog field;
step3.2, inputting the viewpoint sentence into an embedding layer of the general field to obtain word vector representation of the general field;
given a term consisting of a sequence of m words, it can be expressed as: w ═ W1,w2,...wm}. Assigning a label to each word in the sentence to determine whether the word belongs to a part of the evaluation object is a key of the task of extracting the evaluation object. The obtained word sequence is used as the input of two independent pre-training embedding layers and is respectively processed by a word vector matrix
Obtaining each word w in the general fieldiCorresponding dGDimension word vector
Sum word vector matrix
Acquiring each word w in case microblog fieldiCorresponding dCDimension word vector
Wherein VCRepresenting a dictionary of words in a matrix of word vectors,
is representative of the word wiThe encoded vector of (2). After the expression of the two field word vectors is obtained, the two field word vectors are spliced to be used as the input of the convolutional neural network.
The two embedded layers in the model are fixed and not trained with the model.
Step3.3, after the expression of the two field word vectors is obtained, splicing the two field word vectors to be used as the input of a convolutional neural network, and extracting local information of the comments in different ranges by convolutional kernels with different sizes;
the core of the convolutional layer is a filter, local information of sentences is extracted in different ranges by convolutional kernels with different sizes, each convolutional layer is provided with a sliding window with a fixed size, and only information in the window is processed each time. The size of the window is defined as k ', where k ' is 2a +1(k ' is an odd number), and the convolution operation is shown in equation 4:
wherein wuWeight parameter representing convolution kernel, b 'representing bias parameter, f' being an activation function, UiIs the output of the convolution corresponding word.
And Step3.4, obtaining a feature vector of each word in the sentence through the convolutional layer, and calculating to obtain the probability of the label corresponding to each word. The feature vectors are directly input to the fully-connected layer and the softmax layer without pooling, and the fully-connected layer can further quantize the features after being sampled by the convolutional layer, as shown in formula (5):
Qi＝wlabUi+blab (5)
wherein wlabRepresenting the connection weight of the fully connected layer, blabIs the bias term.
The softmax layer is used for converting the result of the full connection layer into the probability of each label, and the calculation formulas for the quantization process and the normalization of the softmax function are respectively as follows:
wherein WjAnd WlTo represent the connection weight of a fully connected layer, p (y ═ j/x) represents the likelihood that vector x belongs to label j, and x represents the probability that vector x belongs to label jTAnd (4) transposing the features of the word vector, wherein L is the total number of label categories, and the label probability corresponding to the maximum probability of each word selection is predicted.
In a second aspect, the invention provides a case microblog evaluation object extraction device based on a double-embedded multilayer convolutional neural network, which comprises modules for executing the method of the first aspect.
For example, the case microblog evaluation object extraction device based on the dual-embedded multilayer convolutional neural network comprises:
the case microblog aspect level emotion analysis data set construction module comprises: the method is used for constructing case microblog aspect level emotion analysis data sets;
an embedded layer acquisition module of case field characteristics: the embedded layer with case field characteristics is obtained by pre-training case microblog comment texts;
case microblog evaluation object extraction module: the method is used for obtaining the representation results of evaluation objects in different fields through the embedded layer of the case microblog field and the general field characteristics, carrying out splicing operation to obtain a new vector, and marking the sequence by using the classifier to extract the case microblog evaluation objects.
To illustrate the effect of the present invention, the experiment was set up with 2 sets of comparative experiments under two data sets. The first set of experiments validated the performance comparison of the model herein and the 6 baseline models, as shown in table 2. The second set of experiments compared the effect of different generic domain word insertions on the model herein.
TABLE 2 comparison of bus case and Benz case data set model accuracy rates
Table 2 the experimental results show that: the model obtains the highest F1 value on two data sets, better accords with the characteristic that a convolutional neural network filter is good at extracting local features, a case microblog viewpoint sentence consisting of word sequences simultaneously passes through two embedding layers to obtain word vectors corresponding to each word, then the word vectors are spliced to obtain new word vectors, and after multilayer convolution, larger feature values can be obtained from different layers, namely more related information in the field of case microblog and in the final feature representation of a sentence.
To verify the effectiveness of dual embedding, the text was characterized by different word embedding on the two datasets, and the evaluation object extraction was performed, with the results shown in table 3.
TABLE 3 ablation experiment
Analysis of table 3 shows that the dual embedding can achieve the best performance under the same data set. Compared with single-field word embedding, the word embedding effect in the general field is better improved by 2.21% and 2.53% respectively in a # Chongqing bus Jiangjiang case # and a # Benz girl driver right maintaining case #, and the word embedding performance is proved to be related to the data scale during training. Compared with the general field and double embedding, the double embedding is respectively promoted by 3.61% and 3.14% in a # Chongqing bus Jiangjiang case # and a # Benz driver Weiright case #, and the double embedding method is proved to be effective in the field of case-related microblogs.
Through the experimental data and analysis, the method extracts the characteristics related to the case through the convolution layer by utilizing word embedding and splicing of the general field and the case microblog field, so that the evaluation object of the microblog comment is extracted. The experimental result shows that the dual-embedding application improves the performance of the extraction of the evaluation object aiming at the specific task of the extraction of the case microblog comment object.
While the present invention has been described in detail with reference to the embodiments shown in the drawings, the present invention is not limited to the embodiments, and various changes can be made without departing from the spirit of the present invention within the knowledge of those skilled in the art.
Claims (5)
1. The case microblog evaluation object extraction method based on the double-embedded multilayer convolutional neural network is characterized in that,
the method comprises the following steps:
step1, constructing a case microblog aspect level emotion analysis data set;
step2, pre-training a case microblog comment text to obtain an embedded layer with case field characteristics;
step3, obtaining the characterization results of the evaluation objects in different fields through the embedding layers of the case microblog field and the general field characteristics, performing splicing operation to obtain new vectors, and marking the sequences by using a classifier to extract the case microblog evaluation objects.
2. The case microblog evaluation object extraction method based on the double-embedded multilayer convolutional neural network as claimed in claim 1, characterized in that: the specific steps of Step1 are as follows:
step1.1, crawling the microblog comments of the relevant cases from the Xinlang microblog by using a crawler based on a Scapy frame;
step1.2, filtering and screening case microblog comments to construct a case microblog data set, and finally obtaining a case microblog data set;
step1.3, filtering and screening case microblog comments to construct a case microblog data set, and finally obtaining a case microblog database;
step1.4, manually marking comment data of the case microblog data set, and marking contained data by taking one microblog comment as a unit; the marked content is an evaluation object in the case microblog comment sentence and the corresponding emotion polarity.
3. The case microblog evaluation object extraction method based on the double-embedded multilayer convolutional neural network as claimed in claim 1, characterized in that: in Step1.2, the mode of filtration screening is as follows:
step1.2.1, dividing the microblog messages according to a forwarding relation '//', and ensuring that comments below the forwarded microblog are analyzed based on the original microblog;
step1.2.2, delete the structure of "@ + username + reply" in the microblog comment, and delete irrelevant hyperlink advertisement;
step1.2.3, replacing the continuously appearing multiple punctuations by adopting first punctuations, and removing emoticons in the contents of the microblog comments;
step1.2.4, filtering and screening out comment data with less than seven characters, and ensuring the completeness and the usability of comment contents.
4. The case microblog evaluation object extraction method based on the double-embedded multilayer convolutional neural network as claimed in claim 1, characterized in that: the specific steps of Step3 are as follows:
step3.1, inputting the viewpoint sentence into an embedding layer in the case microblog field to obtain word vector representation in the case microblog field;
step3.2, inputting the viewpoint sentence into an embedding layer of the general field to obtain word vector representation of the general field;
step3.3, after the expression of the two field word vectors is obtained, splicing the two field word vectors to be used as the input of a convolutional neural network, and extracting local information of the comments in different ranges by convolutional kernels with different sizes;
step3.4, after the operation of the convolutional layer, in order to ensure that the words in the sentence can be extracted to the features, the words are directly input into the fully-connected layer and the softmax layer without the pooling layer, the fully-connected layer can further quantize the features after the sampling of the convolutional layer, and the softmax layer converts the results of the fully-connected layer into the probability of each label.
5. Case microblog evaluation object extraction device based on double-embedded multilayer convolutional neural network, characterized by comprising modules for executing the method according to any one of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011006914.0A CN112287240A (en) | 2020-09-23 | 2020-09-23 | Case microblog evaluation object extraction method and device based on double-embedded multilayer convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011006914.0A CN112287240A (en) | 2020-09-23 | 2020-09-23 | Case microblog evaluation object extraction method and device based on double-embedded multilayer convolutional neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112287240A true CN112287240A (en) | 2021-01-29 |
Family
ID=74421334
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011006914.0A Pending CN112287240A (en) | 2020-09-23 | 2020-09-23 | Case microblog evaluation object extraction method and device based on double-embedded multilayer convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112287240A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112800229A (en) * | 2021-02-05 | 2021-05-14 | 昆明理工大学 | Knowledge graph embedding-based semi-supervised aspect-level emotion analysis method for case-involved field |
| CN113111269A (en) * | 2021-05-10 | 2021-07-13 | 网易（杭州）网络有限公司 | Data processing method and device, computer readable storage medium and electronic equipment |
| CN113298366A (en) * | 2021-05-12 | 2021-08-24 | 北京信息科技大学 | Tourism performance service value evaluation method |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111008274A (en) * | 2019-12-10 | 2020-04-14 | 昆明理工大学 | Case microblog viewpoint sentence identification and construction method of feature extended convolutional neural network |
-
2020
- 2020-09-23 CN CN202011006914.0A patent/CN112287240A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111008274A (en) * | 2019-12-10 | 2020-04-14 | 昆明理工大学 | Case microblog viewpoint sentence identification and construction method of feature extended convolutional neural network |
Non-Patent Citations (1)
| Title |
|---|
| HU XU等: "Double Embeddings and CNN-based Sequence Labeling for Aspect Extraction", 《HTTPS://ARXIV.ORG/ABS/1805.04601》, 11 May 2018 (2018-05-11), pages 1 - 7 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112800229A (en) * | 2021-02-05 | 2021-05-14 | 昆明理工大学 | Knowledge graph embedding-based semi-supervised aspect-level emotion analysis method for case-involved field |
| CN112800229B (en) * | 2021-02-05 | 2022-12-20 | 昆明理工大学 | Knowledge graph embedding-based semi-supervised aspect-level emotion analysis method for case-involved field |
| CN113111269A (en) * | 2021-05-10 | 2021-07-13 | 网易（杭州）网络有限公司 | Data processing method and device, computer readable storage medium and electronic equipment |
| CN113298366A (en) * | 2021-05-12 | 2021-08-24 | 北京信息科技大学 | Tourism performance service value evaluation method |
| CN113298366B (en) * | 2021-05-12 | 2023-12-12 | 北京信息科技大学 | Travel performance service value assessment method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107092596B (en) | Text emotion analysis method based on attention CNNs and CCR | |
| Abdullah et al. | SEDAT: sentiment and emotion detection in Arabic text using CNN-LSTM deep learning | |
| CN109446404B (en) | Method and device for analyzing emotion polarity of network public sentiment | |
| CN108763353B (en) | Baidu encyclopedia relation triple extraction method based on rules and remote supervision | |
| CN111008274B (en) | Case microblog viewpoint sentence identification and construction method of feature extended convolutional neural network | |
| CN110287323B (en) | Target-oriented emotion classification method | |
| CN110489523B (en) | Fine-grained emotion analysis method based on online shopping evaluation | |
| CN112287240A (en) | Case microblog evaluation object extraction method and device based on double-embedded multilayer convolutional neural network | |
| CN111160031A (en) | Social media named entity identification method based on affix perception | |
| CN104268160A (en) | Evaluation object extraction method based on domain dictionary and semantic roles | |
| CN110348227B (en) | Software vulnerability classification method and system | |
| CN109815485B (en) | Method and device for identifying emotion polarity of microblog short text and storage medium | |
| CN106126619A (en) | A kind of video retrieval method based on video content and system | |
| CN111767725A (en) | Data processing method and device based on emotion polarity analysis model | |
| CN112434164B (en) | Network public opinion analysis method and system taking topic discovery and emotion analysis into consideration | |
| CN112016320A (en) | English punctuation adding method, system and equipment based on data enhancement | |
| CN111460162A (en) | Text classification method and device, terminal equipment and computer readable storage medium | |
| CN111339772B (en) | Russian text emotion analysis method, electronic device and storage medium | |
| CN112784602A (en) | News emotion entity extraction method based on remote supervision | |
| CN114417851A (en) | Emotion analysis method based on keyword weighted information | |
| CN116245110A (en) | Multi-dimensional information fusion user standing detection method based on graph attention network | |
| Jia | Sentiment classification of microblog: A framework based on BERT and CNN with attention mechanism | |
| Du et al. | A convolutional attentional neural network for sentiment classification | |
| CN114742071A (en) | Chinese cross-language viewpoint object recognition and analysis method based on graph neural network | |
| CN114416969A (en) | LSTM-CNN online comment sentiment classification method and system based on background enhancement |
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 |
Application publication date: 20210129 |
|
| RJ01 | Rejection of invention patent application after publication |