CN116579343A - Named entity identification method for Chinese text travel class - Google Patents

Named entity identification method for Chinese text travel class Download PDF

Info

Publication number
CN116579343A
CN116579343A CN202310560194.XA CN202310560194A CN116579343A CN 116579343 A CN116579343 A CN 116579343A CN 202310560194 A CN202310560194 A CN 202310560194A CN 116579343 A CN116579343 A CN 116579343A
Authority
CN
China
Prior art keywords
representation
chinese
chinese text
cnn
inputting
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
Application number
CN202310560194.XA
Other languages
Chinese (zh)
Other versions
CN116579343B (en
Inventor
秦智
杜自豪
刘恩洋
张仕斌
昌燕
胡贵强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chengdu University of Information Technology
Original Assignee
Chengdu University of Information Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Chengdu University of Information Technology filed Critical Chengdu University of Information Technology
Priority to CN202310560194.XA priority Critical patent/CN116579343B/en
Publication of CN116579343A publication Critical patent/CN116579343A/en
Application granted granted Critical
Publication of CN116579343B publication Critical patent/CN116579343B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • G06F40/279—Recognition of textual entities
    • G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
    • G06F40/295—Named entity recognition
    • 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
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
    • 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
    • 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
    • G06N3/0464—Convolutional networks [CNN, ConvNet]
    • 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
    • G06N3/048—Activation functions
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • 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)
  • Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Character Discrimination (AREA)

Abstract

The invention discloses a method for identifying a named entity of a Chinese text travel class, which comprises the following steps: s1, acquiring Chinese text travel text data, and inputting the data into a character embedding layer to obtain character vector representation; s2, inputting the character vector representation into a two-way long-short-term memory network layer to obtain a context representation; s3, inputting the context representation into a CNN layer to obtain a multi-scale local context feature fusion representation; s4, inputting the multi-scale local context feature fusion representation to a CRF layer, and performing sequence labeling through the CRF layer to complete the named entity identification of Chinese travel. According to the invention, the problem of less attention to the recognition and research of the named entity of the Chinese travel class is considered, the network construction is carried out aiming at the text data of the Chinese travel class, the multi-scale local context feature fusion representation is learned by utilizing the second CNN module in the CNN layer, the correlation between the semantics is enhanced, and the feature representation beneficial to Chinese recognition is improved.

Description

Named entity identification method for Chinese text travel class
Technical Field
The invention belongs to the technical field of information extraction, and particularly relates to a named entity identification method for Chinese travel classes.
Background
Named Entity Recognition (NER) is a basic information extraction task that can be applied to many downstream tasks in Natural Language Processing (NLP), such as information extraction, social media analysis, search engines, machine translation, knowledge graph. The goal of NER is to extract some predefined specific entities from sentences and identify their correct type, such as person, place, organization. Early named entity recognition was divided into two categories, rule-based and statistical-based. As deep learning becomes more powerful, research into NER has made a tremendous progress. The related fields are various, such as medical field, financial field, news field, etc. However, studies on the recognition of named entities in the travel class are very rare, and studies on the recognition of named entities in the travel class are not focused on.
Based on the differences between languages, studies on NER methods for a specific language, such as english, arabic, indian, and other languages, have been also many, and many researchers have focused mainly on the study of NER in english. However, chinese is an important international general language, and has its own characteristics when compared with English, but research on Chinese NER is far less than that of English NER, and many researches on Chinese NER do not have to be conducted according to the characteristics of Chinese.
Disclosure of Invention
Aiming at the defects in the prior art, the named entity identification method for the Chinese travel class solves the problem that the current named entity identification research has less attention to the Chinese travel class.
In order to achieve the aim of the invention, the invention adopts the following technical scheme: a method for identifying a named entity of a Chinese hotel class comprises the following steps:
s1, acquiring Chinese text travel text data, and inputting the data into a character embedding layer to obtain character vector representation;
s2, inputting the character vector representation into a two-way long-short-term memory network layer to obtain a context representation;
s3, inputting the context representation into a CNN layer to obtain a multi-scale local context feature fusion representation;
s4, inputting the multi-scale local context feature fusion representation to a CRF layer, and performing sequence labeling through the CRF layer to complete the named entity identification of Chinese travel.
Further: in the step S1, a character embedding layer comprises a ChineseBert module and a first CNN module which are parallel;
the step S1 comprises the following sub-steps:
s11, acquiring Chinese text travel text data;
s12, inputting the Chinese text data into a Chinese text module to obtain word embedded vector representation of each word in the Chinese text data;
s13, inputting Chinese text travel text data to a first CNN module to obtain a part first-level embedded representation;
and S14, splicing the character embedding vector representation and the radical level embedding representation to obtain a character vector representation.
Further: the step S12 is specifically as follows:
inputting the Chinese text data to a Chinese Bert module, and encoding and representing the input Chinese text data by the Chinese Bert module to obtain feature vectors, and generating word embedded vector representations of each word in the Chinese text data according to the feature vectors;
wherein the feature vector includes marker embedding, position embedding, and segment embedding.
Further: in the S13, a radical level embedded representation M is obtained 2 The expression of (2) is specifically:
M 2 ＝A 1 (b 1 +C 1 (x))
wherein x is the radical level characteristic of Chinese characters, C 1 (. CNN) is the first CNN module, a 1 B is the first activation function 1 Is the bias of the first CNN module.
Further: in the S14, a character vector representation Z is obtained concat The expression of (2) is specifically:
Z concat ＝M 1 +M 2
wherein M is 1 A vector representation is embedded for the word.
The beneficial effects of the above-mentioned further scheme are: the character vector representation obtained by splicing the word embedding vector representation and the radical level embedding representation can obtain more semantic features, so that the model can better identify Chinese meanings in the text.
Further: in the step S2, the two-way long-short-term memory network layer comprises a first LSTM unit, a second LSTM unit, a third LSTM unit, a fourth LSTM unit, a fifth LSTM unit and a sixth LSTM unit, wherein the first LSTM unit and the sixth LSTM unit are used for positively processing the input character vector representation, and the seventh LSTM unit and the twelfth LSTM unit are used for reversely processing the input character vector representation;
the method for obtaining the context representation comprises the following steps:
and splicing according to the output results of the first to twelfth LSTM units to obtain the context representation. Further:
further: in S2, the expression for obtaining the context representation H is specifically:
H＝{h 1 ,...,h ti ,...,h D }
in the formula, h ti Splicing the output results of the first to twelfth LSTM units, wherein ti is the sequence number of the splice, and ti=1, …, D and D are the dimensions represented by the character vectors;
the first to twelfth LSTM units each comprise an input gate i t Output door o t And forget door f t The expression is specifically as follows:
i t ＝σ(W xi x t +W hi h t-1 +W ci c t-1 +b i )
f t ＝σ(W xf x t +W hf h t-1 +W cf c t-1 +b f )
c t ＝f t ⊙c t-1 +i t ⊙tanh(W xc x t +W hc h t-1 +b c )
o t ＝σ(W xo x t +W ho h t-1 +W co c t +b o )
h t ＝o t ⊙tanh(c t )
wherein sigma (& gt) is a sigmoid function per element, tan h (& gt) is a hyperbolic tangent function, and wt. is a multiplication function per element xi 、W hi 、W ci 、W xf 、W hf 、W cf 、W xc 、W hc 、W xo 、W ho And W is co All are weight parameters, b i 、b f 、b c And b o Are all bias parameters, c t Is a memory cell, h t To output the result.
Further: in the step S3, a second CNN module is arranged on the CNN layer to obtain a multi-scale local context feature fusion representation M 3 The expression of (2) is specifically:
M 3 ＝A 2 (b 2 +C 2 (H))
wherein H is represented by the following formula, C 2 (. CNN) is the second CNN module, a 2 B is a second activation function 2 Is the bias of the second CNN module.
The beneficial effects of the above-mentioned further scheme are: and inputting the context representation to a second CNN module, so that the correlation between semantics can be enhanced, and a multi-scale local context feature fusion representation is generated.
The beneficial effects of the invention are as follows: the method for identifying the named entity of the Chinese travel class solves the problem of less attention to the named entity identification research of the Chinese travel class, carries out network construction aiming at text data of the Chinese travel class, learns radical-level embedded representation based on Chinese by utilizing a first CNN module at a character embedding layer to obtain character vector representation favorable for Chinese identification, learns multi-scale local context feature fusion representation by utilizing a second CNN module at a CNN layer, strengthens the correlation between semantics, and further improves feature representation favorable for Chinese identification.
Drawings
FIG. 1 is a flow chart of a method for identifying a named entity of a Chinese travel class according to the present invention.
Fig. 2 is a schematic diagram of the overall network structure of the present invention.
Fig. 3 is a schematic structural diagram of the chinese bert module of the present invention.
Fig. 4 is a schematic structural diagram of a first CNN module according to the present invention.
Fig. 5 is a schematic structural diagram of a second CNN module according to the present invention.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and all the inventions which make use of the inventive concept are protected by the spirit and scope of the present invention as defined and defined in the appended claims to those skilled in the art.
As shown in fig. 1, in one embodiment of the present invention, a method for identifying named entities of chinese travel class includes the steps of:
s1, acquiring Chinese text travel text data, and inputting the data into a character embedding layer to obtain character vector representation;
s2, inputting the character vector representation into a two-way long-short-term memory network layer to obtain a context representation;
s3, inputting the context representation into a CNN layer to obtain a multi-scale local context feature fusion representation;
s4, inputting the multi-scale local context feature fusion representation to a CRF layer, and performing sequence labeling through the CRF layer to complete the named entity identification of Chinese travel.
In this embodiment, the invention provides a method for identifying a named entity of a chinese text, which aims at the characteristics of chinese characters and is represented by the fusion of radical level features and multi-scale local context features in the application field of text class data, wherein the specific structure of a network is shown in fig. 2.
In the step S1, a character embedding layer comprises a ChineseBert module and a first CNN module which are parallel;
the step S1 comprises the following sub-steps:
s11, acquiring Chinese text travel text data;
s12, inputting the Chinese text data into a Chinese text module to obtain word embedded vector representation of each word in the Chinese text data;
s13, inputting Chinese text travel text data to a first CNN module to obtain a part first-level embedded representation;
and S14, splicing the character embedding vector representation and the radical level embedding representation to obtain a character vector representation.
In this embodiment, the structure of the chinese bert module is shown in fig. 3, where the chinese bert module is a pre-training model obtained by pre-training chinese corpus, and is specifically directed to processing chinese text data.
The step S12 is specifically as follows:
inputting the Chinese text data to a Chinese Bert module, and encoding and representing the input Chinese text data by the Chinese Bert module to obtain feature vectors, and generating word embedded vector representations of each word in the Chinese text data according to the feature vectors;
wherein the feature vector includes marker embedding, position embedding, and segment embedding.
In the S13, a radical level embedded representation M is obtained 2 The expression of (2) is specifically:
M 2 ＝A 1 (b 1 +C 1 (x))
wherein x is the radical level characteristic of Chinese characters, C 1 (. CNN) is the first CNN module, a 1 B is the first activation function 1 Is the bias of the first CNN module.
In this embodiment, the Radical level embedding representation radial-level Representation is performed on the input chinese text data using CNN, the result is a Radical level embedded representation, wherein a first CNN module performs Radical Representaion a structural diagram of the input data is shown in fig. 4.
In the S14, a character vector representation Z is obtained concat The expression of (2) is specifically:
Z concat ＝M 1 +M 2
wherein M is 1 A vector representation is embedded for the word.
The character vector representation obtained by splicing the word embedding vector representation and the radical level embedding representation can obtain more semantic features, so that the model can better identify Chinese meanings in the text.
In the step S2, the two-way long-short-term memory network layer comprises a first LSTM unit, a second LSTM unit, a third LSTM unit, a fourth LSTM unit, a fifth LSTM unit and a sixth LSTM unit, wherein the first LSTM unit and the sixth LSTM unit are used for positively processing the input character vector representation, and the seventh LSTM unit and the twelfth LSTM unit are used for reversely processing the input character vector representation;
the method for obtaining the context representation comprises the following steps:
and splicing according to the output results of the first to twelfth LSTM units to obtain the context representation.
In this embodiment, the context representation block obtained by the two-way long-short-term memory network layer can promote semantic representation from the forward direction and the reverse direction, and can better identify the semantics in the paragraph.
In S2, the expression for obtaining the context representation H is specifically:
H＝{h 1 ,...,h ti ,...,h D }
in the formula, h ti Splicing the output results of the first to twelfth LSTM units, wherein ti is the sequence number of the splice, and ti=1, …, D and D are the dimensions represented by the character vectors;
the first to twelfth LSTM units each comprise an input gate i t Output door o t And forget door f t The expression is specifically as follows:
i t ＝σ(W xi x t +W hi h t-1 +W ci c t-1 +b i )
f t ＝σ(W xf x t +W hf h t-1 +W cf c t-1 +b f )
c t ＝f t ⊙c t-1 +i t ⊙tanh(W xc x t +W hc h t-1 +b c )
o t ＝σ(W xo x t +W ho h t-1 +W co c t +b o )
h t ＝o t ⊙tanh(c t )
wherein sigma (& gt) is a sigmoid function per element, tan h (& gt) is a hyperbolic tangent function, and wt. is a multiplication function per element xi 、W hi 、W ci 、W xf 、W hf 、W cf 、W xc 、W hc 、W xo 、W ho And W is co All are weight parameters, b i 、b f 、b c And b o Are all bias parameters, c t Is a memory cell, h t To output the result.
In the step S3, a second CNN module is arranged on the CNN layer to obtain a multi-scale local context feature fusion representation M 3 The expression of (2) is specifically:
M 3 ＝A 2 (b 2 +C 2 (H))
wherein H is represented by the following formula, C 2 (. CNN) is the second CNN module, a 2 B is a second activation function 2 Is the bias of the second CNN module.
In this embodiment, the structure of the second CNN module is shown in fig. 5, and the context representation is input to the second CNN module, so that the correlation between semantics can be enhanced, and a multi-scale local context feature fusion representation is generated.
And inputting the multi-scale local context feature fusion representation to a CRF layer to finish the task of sequence labeling so as to finish the named entity identification of Chinese text traveling.
The beneficial effects of the invention are as follows: the method for identifying the named entity of the Chinese travel class solves the problem of less attention to the named entity identification research of the Chinese travel class, carries out network construction aiming at text data of the Chinese travel class, learns radical-level embedded representation based on Chinese by utilizing a first CNN module at a character embedding layer to obtain character vector representation favorable for Chinese identification, learns multi-scale local context feature fusion representation by utilizing a second CNN module at a CNN layer, strengthens the correlation between semantics, and further improves feature representation favorable for Chinese identification.
In the description of the present invention, it should be understood that the terms "center," "thickness," "upper," "lower," "horizontal," "top," "bottom," "inner," "outer," "radial," and the like indicate or are based on the orientation or positional relationship shown in the drawings, merely to facilitate description of the present invention and to simplify the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be configured and operated in a particular orientation, and thus should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be interpreted as indicating or implying a relative importance or number of technical features indicated. Thus, a feature defined as "first," "second," "third," or the like, may explicitly or implicitly include one or more such feature.

Claims (8)

1. A method for identifying a named entity of a Chinese text travel class is characterized by comprising the following steps:
s1, acquiring Chinese text travel text data, and inputting the data into a character embedding layer to obtain character vector representation;
s2, inputting the character vector representation into a two-way long-short-term memory network layer to obtain a context representation;
s3, inputting the context representation into a CNN layer to obtain a multi-scale local context feature fusion representation;
s4, inputting the multi-scale local context feature fusion representation to a CRF layer, and performing sequence labeling through the CRF layer to complete the named entity identification of Chinese travel.
2. The method for identifying a named entity of a chinese text traveling class according to claim 1, wherein in S1, the character embedding layer includes a chinese bert module and a first CNN module in parallel;
the step S1 comprises the following sub-steps:
s11, acquiring Chinese text travel text data;
s12, inputting the Chinese text data into a Chinese text module to obtain word embedded vector representation of each word in the Chinese text data;
s13, inputting Chinese text travel text data to a first CNN module to obtain a part first-level embedded representation;
and S14, splicing the character embedding vector representation and the radical level embedding representation to obtain a character vector representation.
3. The method for identifying a named entity of a chinese text travel class according to claim 2, wherein S12 is specifically:
inputting the Chinese text data to a Chinese Bert module, and encoding and representing the input Chinese text data by the Chinese Bert module to obtain feature vectors, and generating word embedded vector representations of each word in the Chinese text data according to the feature vectors;
wherein the feature vector includes marker embedding, position embedding, and segment embedding.
4. The method for recognizing named entities of chinese text class according to claim 2, wherein in S13, a radical level embedded representation M is obtained 2 The expression of (2) is specifically:
M 2 ＝A 1 (b 1 +C 1 (x))
wherein x is the radical level characteristic of Chinese characters, C 1 (. CNN) is the first CNN module, a 1 B is the first activation function 1 Is the bias of the first CNN module.
5. The method for recognizing named entities in chinese text class according to claim 4, wherein in S14, a character vector representation Z is obtained concat The expression of (2) is specifically:
Z concat ＝M 1 +M 2
wherein M is 1 A vector representation is embedded for the word.
6. The method for recognizing a named entity of a chinese text traveling class according to claim 1, wherein in S2, the bidirectional long-short-term memory network layer includes first to twelfth LSTM units, the first to sixth LSTM units forward process the inputted character vector representation, and the seventh to twelfth LSTM units reverse process the inputted character vector representation;
the method for obtaining the context representation comprises the following steps:
and splicing according to the output results of the first to twelfth LSTM units to obtain the context representation.
7. The method for identifying a named entity of a chinese text travel class according to claim 6, wherein in S2, the expression for obtaining the context representation H is specifically:
H＝{h 1 ,...,h ti ,...,h D }
in the formula, h ti Splicing the output results of the first to twelfth LSTM units, wherein ti is the sequence number of the splice, and ti=1, …, D and D are the dimensions represented by the character vectors;
the first to twelfth LSTM units each comprise an input gate i t Output door o t And forget door f t The expression is specifically as follows:
i t ＝σ(W xi x t +W hi h t-1 +W ci c t-1 +b i )
f t ＝σ(W xf x t +W hf h t-1 +W cf c t-1 +b f )
c t ＝f t ⊙c t-1 +i t ⊙tanh(W xc x t +W hc h t-1 +b c )
o t ＝σ(W xo x t +W ho h t-1 +W co c t +b o )
h t ＝o t ⊙tanh(c t )
wherein sigma (& gt) is a sigmoid function per element, tan h (& gt) is a hyperbolic tangent function, and wt. is a multiplication function per element xi 、W hi 、W ci 、W xf 、W hf 、W cf 、W xc 、W hc 、W xo 、W ho And W is co All are weight parameters, b i 、b f 、b c And b o Are all bias parameters, c t Is a memory cell, h t To output the result.
8. The method for identifying a named entity of a chinese text class according to claim 1, wherein in S3, a CNN layer is provided with a second CNN module, so as to obtain a multiscale local context feature fusion representation M 3 The expression of (2) is specifically:
M 3 ＝A 2 (b 2 +C 2 (H))
wherein H is represented by the following formula, C 2 (. CNN) is the second CNN module, a 2 B is a second activation function 2 Is the bias of the second CNN module.
CN202310560194.XA 2023-05-17 2023-05-17 Named entity identification method for Chinese text travel class Active CN116579343B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310560194.XA CN116579343B (en) 2023-05-17 2023-05-17 Named entity identification method for Chinese text travel class

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310560194.XA CN116579343B (en) 2023-05-17 2023-05-17 Named entity identification method for Chinese text travel class

Publications (2)

Publication Number Publication Date
CN116579343A true CN116579343A (en) 2023-08-11
CN116579343B CN116579343B (en) 2024-06-04

Family

ID=87543867

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310560194.XA Active CN116579343B (en) 2023-05-17 2023-05-17 Named entity identification method for Chinese text travel class

Country Status (1)

Country Link
CN (1) CN116579343B (en)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021114745A1 (en) * 2019-12-13 2021-06-17 华南理工大学 Named entity recognition method employing affix perception for use in social media
US20210216862A1 (en) * 2020-01-15 2021-07-15 Beijing Jingdong Shangke Information Technology Co., Ltd. System and method for semantic analysis of multimedia data using attention-based fusion network
CN113408289A (en) * 2021-06-29 2021-09-17 广东工业大学 Multi-feature fusion supply chain management entity knowledge extraction method and system
CN114118099A (en) * 2021-11-10 2022-03-01 浙江工业大学 Chinese automatic question-answering method based on radical characteristics and multi-layer attention mechanism
CN114781380A (en) * 2022-03-21 2022-07-22 哈尔滨工程大学 Chinese named entity recognition method, equipment and medium fusing multi-granularity information
CN115455955A (en) * 2022-10-18 2022-12-09 昆明理工大学 Chinese named entity recognition method based on local and global character representation enhancement
CN115600597A (en) * 2022-10-18 2023-01-13 淮阴工学院（Cn） Named entity identification method, device and system based on attention mechanism and intra-word semantic fusion and storage medium
CN115688782A (en) * 2022-10-26 2023-02-03 成都理工大学 Named entity recognition method based on global pointer and countermeasure training

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021114745A1 (en) * 2019-12-13 2021-06-17 华南理工大学 Named entity recognition method employing affix perception for use in social media
US20210216862A1 (en) * 2020-01-15 2021-07-15 Beijing Jingdong Shangke Information Technology Co., Ltd. System and method for semantic analysis of multimedia data using attention-based fusion network
CN113408289A (en) * 2021-06-29 2021-09-17 广东工业大学 Multi-feature fusion supply chain management entity knowledge extraction method and system
CN114118099A (en) * 2021-11-10 2022-03-01 浙江工业大学 Chinese automatic question-answering method based on radical characteristics and multi-layer attention mechanism
CN114781380A (en) * 2022-03-21 2022-07-22 哈尔滨工程大学 Chinese named entity recognition method, equipment and medium fusing multi-granularity information
CN115455955A (en) * 2022-10-18 2022-12-09 昆明理工大学 Chinese named entity recognition method based on local and global character representation enhancement
CN115600597A (en) * 2022-10-18 2023-01-13 淮阴工学院（Cn） Named entity identification method, device and system based on attention mechanism and intra-word semantic fusion and storage medium
CN115688782A (en) * 2022-10-26 2023-02-03 成都理工大学 Named entity recognition method based on global pointer and countermeasure training

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
昌燕 等: "路径语义和特征提取相结合的负样本推荐方法", 小型微型计算机系统 *

Also Published As

Publication number Publication date
CN116579343B (en) 2024-06-04

Similar Documents

Publication Publication Date Title
CN110765775B (en) Self-adaptive method for named entity recognition field fusing semantics and label differences
Yousfi et al. Contribution of recurrent connectionist language models in improving LSTM-based Arabic text recognition in videos
CN111160031A (en) Social media named entity identification method based on affix perception
CN112541356B (en) Method and system for recognizing biomedical named entities
CN110647612A (en) Visual conversation generation method based on double-visual attention network
Alsaaran et al. Classical Arabic named entity recognition using variant deep neural network architectures and BERT
Mehmood et al. A precisely xtreme-multi channel hybrid approach for roman urdu sentiment analysis
CN111950283B (en) Chinese word segmentation and named entity recognition system for large-scale medical text mining
Alsaaran et al. Arabic named entity recognition: A BERT-BGRU approach
Sait et al. Deep Learning with Natural Language Processing Enabled Sentimental Analysis on Sarcasm Classification.
Chang et al. Application of Word Embeddings in Biomedical Named Entity Recognition Tasks.
Kim et al. Construction of machine-labeled data for improving named entity recognition by transfer learning
CN117010387A (en) Roberta-BiLSTM-CRF voice dialogue text naming entity recognition system integrating attention mechanism
CN112507717A (en) Medical field entity classification method fusing entity keyword features
Han et al. MAF‐CNER: A Chinese Named Entity Recognition Model Based on Multifeature Adaptive Fusion
Heo et al. Hybrid sense classification method for large-scale word sense disambiguation
Shu et al. Investigating lstm with k-max pooling for text classification
CN116227594A (en) Construction method of high-credibility knowledge graph of medical industry facing multi-source data
Bach et al. Question analysis towards a Vietnamese question answering system in the education domain
Han et al. An attention-based neural framework for uncertainty identification on social media texts
Seeha et al. ThaiLMCut: Unsupervised pretraining for Thai word segmentation
He et al. A Chinese named entity recognition model of maintenance records for power primary equipment based on progressive multitype feature fusion
Zhang et al. Sjtu-nlp at semeval-2018 task 9: Neural hypernym discovery with term embeddings
CN117057350B (en) Chinese electronic medical record named entity recognition method and system
CN117131932A (en) Semi-automatic construction method and system for domain knowledge graph ontology based on topic model

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