CN116153446A - Perovskite solar cell HTL material prediction method based on NLP (non-linear liquid phase) processing - Google Patents

Perovskite solar cell HTL material prediction method based on NLP (non-linear liquid phase) processing Download PDF

Info

Publication number
CN116153446A
CN116153446A CN202310156361.4A CN202310156361A CN116153446A CN 116153446 A CN116153446 A CN 116153446A CN 202310156361 A CN202310156361 A CN 202310156361A CN 116153446 A CN116153446 A CN 116153446A
Authority
CN
China
Prior art keywords
solar cell
perovskite solar
htl
nlp
perovskite
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
Application number
CN202310156361.4A
Other languages
Chinese (zh)
Inventor
张磊
李深越
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.)
Nanjing University of Information Science and Technology
Original Assignee
Nanjing University of Information Science and 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 Nanjing University of Information Science and Technology filed Critical Nanjing University of Information Science and Technology
Priority to CN202310156361.4A priority Critical patent/CN116153446A/en
Publication of CN116153446A publication Critical patent/CN116153446A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C60/00—Computational materials science, i.e. ICT specially adapted for investigating the physical or chemical properties of materials or phenomena associated with their design, synthesis, processing, characterisation or utilisation
    • 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
    • G06F16/903—Querying
    • G06F16/90335—Query processing
    • G06F16/90344—Query processing by using string matching techniques
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/70—Machine learning, data mining or chemometrics
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/90—Programming languages; Computing architectures; Database systems; Data warehousing
    • 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
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E10/00—Energy generation through renewable energy sources
    • Y02E10/50—Photovoltaic [PV] energy
    • Y02E10/549—Organic PV cells

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Data Mining & Analysis (AREA)
  • Software Systems (AREA)
  • Chemical & Material Sciences (AREA)
  • Crystallography & Structural Chemistry (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Computational Linguistics (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Photovoltaic Devices (AREA)

Abstract

The invention discloses a perovskite solar cell HTL material prediction method based on NLP processing. The invention adopts NLP9 (natural language) processing method to explore perovskite solar cell materials, automatically extracts scientific information and predicts candidate materials using new data types. Prototype metal halide perovskite, ETL and HTL materials and additive materials were analyzed in detail using word2 vec's natural language processing techniques; the material composition of the perovskite solar cell was predicted based on the results of NLP. The HTL candidate ferroferric oxide predicted by the NLP model has not been implemented as a suitable HTL material in existing databases. The candidate material is then structurally and electronically characterized by first principles calculation, andbuild advanced Fe 3 O 4 /CH 3 NH 3 PbI 3 Heterostructures to reveal their optoelectronic properties. The invention applies the natural language processing method to the research of material science, and provides a platform for analyzing and predicting perovskite solar cell materials by applying the natural language processing method.

Description

Perovskite solar cell HTL material prediction method based on NLP (non-linear liquid phase) processing
Technical Field
The invention belongs to the field of machine learning prediction materials, and particularly relates to a perovskite solar cell HTL material prediction method based on NLP processing.
Background
Perovskite is a crystal material with a molecular general formula of ABX3, is in an octahedral shape and has excellent structural characteristics. The perovskite crystal has simple preparation process and high photoelectric conversion efficiency, and is widely applied to the fields of photovoltaics, LEDs and the like. However, the problems of high-quality perovskite thin film repeated preparation difficulty, low light output coupling efficiency, lead pollution and the like are still to be solved, and a certain distance is still provided for commercial application at present. The use of computer automation to extract potential scientific knowledge from literature is highly desirable for future materials and chemical research in the computer artificial intelligence era, however a large number of perovskite-related scientific literature is not effectively utilized.
Disclosure of Invention
The invention aims at:
the technical scheme is that scientific information is automatically extracted and HTL candidate materials with possible application value are predicted in the perovskite solar cell field by adopting an NLP (natural language) processing method.
In order to achieve the above purpose, the present invention provides the following technical solutions: the perovskite solar cell HTL material prediction method based on NLP treatment comprises the following steps:
s1, constructing a literature database with perovskite keywords;
s2, extracting material names and corresponding chemical formulas in a literature database;
s3, constructing and training a hole transport layer material prediction model by taking the material name and the chemical formula thereof as input and the candidate materials related to the hole transport layer material as output;
s4, inputting the material name and the corresponding chemical formula into the hole transport layer material prediction model obtained in the step S3, performing DFT calculation on the candidate materials related to the obtained hole transport layer material, and verifying the atomic structure and the photoelectric performance of the candidate materials output by the hole transport layer material prediction model.
In step S1, the document abstract using perovskite as a keyword is crawled from the springlink by using the API to construct a document database.
The step S2 further comprises the following substeps:
s201, deleting sentences and punctuation marks which are irrelevant to material names in a literature by using an NLTK toolkit method;
s202, performing material name and corresponding chemical formula identification extraction by using the combination of chemdataextraction and regular expression.
The step S3 further includes the following sub-steps:
s301, inputting the material names and the chemical formulas thereof into Word2vec of each material name and the chemical formula thereof in a database, extracting Word vectors by adopting a lattice-skipping method, and adjusting optimal parameters to perform unsupervised training;
s302, sorting the cosine similarity of each word vector and the applied perovskite solar cell to obtain candidate materials related to the hole transport layer materials.
The step S4 further comprises the following substeps:
s401, spin polarization calculation is carried out through the first principle of interest, and structure and electronic characterization are carried out on candidate materials;
s402, by constructing CH 3 NH 3 PbX 3 And the optical performance of the candidate material is verified by heterostructures with the candidate material.
Further in the step S401, the first principle calculation uses cast, the cut-off energy is 430eV, and the density functional is perdu-beck-enchoff.
Compared with the prior art, the invention has the following beneficial effects:
perovskite solar cell materials were successfully simulated by natural language processing methods. The technical scheme is that scientific information is automatically extracted and HTL candidate materials with possible application value are predicted in the perovskite solar cell field by adopting an NLP (natural language) processing method. Thereby realizing the problems of high repeated preparation difficulty, low photoelectric conversion efficiency and the like existing in the perovskite solar cell at present. And gradually perfecting the application of the artificial intelligence in the field of perovskite materials, and establishing a related database further promotes the development and progress of the interdisciplinary disciplines of the artificial intelligence and the materials. The multi-modal fusion of experimental data, first sexual principle calculation data and natural language processing data in the material field is realized. And a platform is provided for analysis and prediction of perovskite solar cell materials using natural language processing methods.
Through NLP model, various chemical information such as periodic table and automatic classification of perovskite/ETL/HTL materials is obtained. The model shows that iodine based CH3NH3PbI3 acts as the primary metal halide perovskite material for perovskite solar cells, as compared to Cs-, cl-, and Br-based CH3NH3PbI 3. In addition, the model also shows that tin oxide, cuSCN, and LiCO3 are ETL, HTL, and additive materials that are highly correlated with perovskite solar cells.
Drawings
Fig. 1 is a flow chart of the present invention.
FIG. 2 is an analytical graph of perovskite, HTL, ETL in word2vec model; in the figure, (a) is a distribution diagram of perovskite, ETL and HTL materials, and (b) is an extraction diagram of material relationships.
FIG. 3 is a plot of PDOS spectra and UV-visible spectra of three Fe3O4/CH3NH3PbI3 perovskite/HTL interfaces; in the figure, (a) is a PDOS spectrum of MAI-O, (b) is a PDOS spectrum of PbI2-O, (c) is a PDOS spectrum of PbI2-Fe, and (d) is an ultraviolet-visible spectrum of MAI-O, pbI-O and PbI 2-Fe.
Fig. 4 is a graph of the predicted results of hole transport layer materials output using the training model of the present invention.
Detailed Description
For a better understanding of the technical content of the present invention, specific examples are set forth below, along with the accompanying drawings.
Aspects of the invention are described herein with reference to the drawings, in which there are shown many illustrative embodiments. The embodiments of the present invention are not limited to the embodiments described in the drawings. It is to be understood that this invention is capable of being carried out by any of the various concepts and embodiments described above and as such described in detail below, since the disclosed concepts and embodiments are not limited to any implementation. Additionally, some aspects of the disclosure may be used alone or in any suitable combination with other aspects of the disclosure.
As shown in fig. 1, the flow chart of the invention is that the perovskite solar cell hole transport layer material prediction method based on natural language processing comprises the following steps:
s1, constructing a literature database with perovskite keywords;
s2, extracting material names and corresponding chemical formulas in a literature database;
s3, constructing and training a hole transport layer material prediction model by taking the material name and the chemical formula thereof as input and the candidate materials related to the hole transport layer material as output;
s4, inputting the material name and the corresponding chemical formula into the hole transport layer material prediction model obtained in the step S3, performing DFT calculation on the candidate materials related to the obtained hole transport layer material, and verifying the atomic structure and the photoelectric performance of the candidate materials output by the hole transport layer material prediction model.
In step S1, a crawler program or a web crawler plug-in and a database API (or a search engine of the database) are utilized to crawl and collect document summaries taking perovskite as keywords in 1997 to 2021 in springer link.
(no time constraint is imposed, 20,000 or more) to obtain an initial training corpus.
Step S2 comprises the following sub-steps:
s201, deleting meaningless conjunctions, brackets, punctuations and the like which are irrelevant to material names in a literature, deleting news categories, credits, conference category and book evaluation category literature summaries and the like with little reference significance by using an NLTK toolkit method, and carrying out '[ ]' () 'and'; ' these extraneous symbols are deleted;
s202, performing material name and corresponding chemical formula identification extraction by using the combination of chemdataextraction and regular expression.
The step S3 specifically comprises the following steps:
s301, inputting the material names and the chemical formulas of the material names into Word2vec, extracting Word vectors by using a skip-gram method, and adjusting optimal parameters to perform unsupervised training.
The following parameters were used in the training: cutting off the dictionary when the word frequency is less than 2; the feature dimension is 100, the window size representing the maximum distance between the current word and the predicted word is 5, and the random downsampling threshold for high frequency words is 1 x 10-4.Word2vec tab model provides a projected probability distribution for words that are close to the input by moving a window over the text data to train the hidden layer neural network based on a given input Word. To go from the projection layer to the hidden layer, the word is represented by a single hot code; the projection weights are then converted to word embeddings. Thus, if the hidden layer contains 100 neurons, this network will provide 100-dimensional (or 200-dimensional) word embedding. In contrast, the continuous word bag (CBOW) model uses an average of a large number of input context words to predict a center word, rather than a single word in the case of a skip. S302, sorting the cosine similarity of each word vector and the applied perovskite solar cell to obtain candidate materials related to the hole transport layer materials. The codes in Word2vec and the Python loop sentence are called to find out the chemical formula with the highest correlation coefficient with the keyword hole transport layer material (HTL materials), and the ranking of the correlation coefficients reflects the probability of the hole transport layer material after machine learning and unsupervised training to a certain extent, and the higher the ranking, the higher the probability.
As shown in fig. 2, where (a) shows that perovskite solar cells, perovskite, ETL, and hole transport layer material HTL are distributed in three distinct regions after main component analysis (PCA) dimension reduction, the ability of the machine learning model to automatically identify perovskite solar cell materials is demonstrated. (b) The material relation extraction graph has the advantages that the distribution directions of chemical elements are consistent, the oxide is distributed in the other direction, and the application of the material is consistent in the other direction. Specifically, typical ETL materials of perovskite solar cells, titanium dioxide, tin oxide, and zinc oxide, are in close proximity to one another; the perovskite hole transport layer material HTL comprises very close proximity of copper iodide, copper sulfide, tungsten trioxide and copper oxide; perovskite such as typical CH 3 NH 3 PbX 3 And CsPbCl 3 Adjacent to each other. Thus, a relatively unusual hole transport layer material, ferroferric oxide, was predicted, as shown in FIG. 4, to be heavy with other hole transport layer materialsAnd (3) stacking.
S4, inputting the material name and the corresponding chemical formula into the hole transport layer material prediction model obtained in the step S3, performing DFT calculation on the candidate materials related to the obtained hole transport layer material, and verifying the atomic structure and the photoelectric performance of the candidate materials output by the hole transport layer material prediction model.
S401, spin polarization calculation is carried out through the first principle of interest, and structure and electronic characterization are carried out on candidate materials;
s402, by constructing CH 3 NH 3 PbX 3 And the optical performance of the candidate material is verified by heterostructures with the candidate material.
Wherein, the first sexual principle calculation adopts CASTEP, the cut-off energy is 430eV, and the density functional is Perot-Berk-Enzehough (PBE). Due to ferroferric oxide Fe 3 O 4 Spin polarization calculations were performed. CH (CH) 3 NH 3 PbX 3 The halide perovskite is focused on building an advanced perovskite/HTL interface, as it represents the majority of perovskite light absorbing materials. The perovskite surface along the (001) direction was focused and the different ends were studied to build the heterostructure. The heterostructure unit cell was 8.8Ax8.8Ax40.08A and a vacuum layer of 15A thickness was interposed to avoid unwanted plate interactions. Properties such as projection state density (PDOS) were obtained using a 4×4×1 k-point set. Ferroferric oxide Fe 3 O 4 Is a strongly related oxide, the system is treated with dft+u format, where u=3.8 is the Fe-3d electron. CH (CH) 3 NH 3 PbX 3 And ferroferric oxide Fe 3 O 4 The lattice mismatch of the heterostructure of (c) is within 5%. Simulate CH 3 NH 3 PbX 3 And different ends of ferroferric oxide: the halide perovskite layer comprises a Methyl Ammonium (MAI) end surface and a lead iodide end surface; for the ferroferric oxide surface, iron termination and oxygen termination are included. This results in a combination of several heterostructures; however, in the geometry optimization phase, several heterostructures consisting of perovskite layers and HTL layers fail to converge in the geometry optimization phase, CH 3 NH 3 PbX 3 /Fe 3 O 4 Three possible heterostructures of the system. The light absorption coefficient in the uv-vis absorption spectrum is determined by the real part of the dielectric function, which is derived from the imaginary part by the kreimer-kroneger relationship. The kramers-kroneger relationship is expressed as a momentum matrix element between an occupied electron state and an unoccupied electron state. Thereby verifying the correctness of the natural language processing analysis result.
As shown in fig. 3, in the stage of verifying the prediction result, theoretical verification is performed by using a first principle calculation method, and CH is revealed by PDOS spectrum and uv-vis absorption spectrum 3 NH 3 PbX 3 /Fe 3 O 4 Optoelectronic properties of the system. FIG. 3 (a) shows the PDOS spectrum of MAI-O, (b) shows the PDOS spectrum of PbI2-O, (c) shows the PDOS spectrum of PbI2-Fe, and (d) shows the UV-visible spectrum of MAI-O, pbI2-O and PbI 2-Fe. For three CH 3 NH 3 PbX 3 /Fe 3 O 4 Heterostructures, the valence band, is mainly contributed by the perovskite material, while ferroferric oxide contributes to both conduction bands. Due to ferroferric oxide Fe 3 O 4 The heterostructures exhibit balanced light absorption capability in the uv-visible and infrared regions. And typically CH 3 NH 3 PbX 3 In contrast, the ultraviolet-visible absorption spectrum anomalies of the three systems are due to ferroferric oxide Fe 3 O 4 Additional light absorption by the layer. From a simulation point of view the material can indeed be used as hole transport layer for perovskite solar cells.
While the invention has been described in terms of preferred embodiments, it is not intended to be limiting. Those skilled in the art will appreciate that various modifications and adaptations can be made without departing from the spirit and scope of the present invention. Accordingly, the scope of the invention is defined by the appended claims.

Claims (6)

1. The perovskite solar cell HTL material prediction method based on NLP treatment is characterized by comprising the following steps of:
s1, constructing a literature database with perovskite keywords;
s2, extracting material names and corresponding chemical formulas in a literature database;
s3, constructing and training a hole transport layer material prediction model by taking the material name and the chemical formula thereof as input and the candidate materials related to the hole transport layer material as output;
s4, inputting the material name and the corresponding chemical formula into the hole transport layer material prediction model obtained in the step S3, performing DFT calculation on the candidate materials related to the obtained hole transport layer material, and verifying the atomic structure and the photoelectric performance of the candidate materials output by the hole transport layer material prediction model.
2. The method for predicting the HTL material of the perovskite solar cell based on the NLP process according to claim 1, wherein in step S1, the document abstract construction document database using perovskite as a keyword is crawled from the springlink by using the API.
3. The method for predicting the HTL material of the perovskite solar cell based on NLP process according to claim 1, wherein the step S2 comprises the following sub-steps:
s201, deleting sentences and punctuation marks which are irrelevant to material names in a literature by using an NLTK toolkit method;
s202, performing material name and corresponding chemical formula identification extraction by using the combination of chemdataextraction and regular expression.
4. A method for predicting the HTL material of a perovskite solar cell based on NLP process according to claim 3, wherein the step S3 comprises the following sub-steps:
s301, inputting the material names and the chemical formulas thereof into Word2vec of each material name and the chemical formula thereof in a database, extracting Word vectors by adopting a lattice-skipping method, and adjusting optimal parameters to perform unsupervised training;
s302, sorting the cosine similarity of each word vector and the applied perovskite solar cell to obtain candidate materials related to the hole transport layer materials.
5. A perovskite solar cell HTL material prediction method based on NLP processing according to claim 3, wherein step S4 comprises the sub-steps of:
s401, spin polarization calculation is carried out through the first principle of interest, and structure and electronic characterization are carried out on candidate materials;
s402, by constructing CH 3 NH 3 PbX 3 And the optical performance of the candidate material is verified by heterostructures with the candidate material.
6. The method for predicting the HTL material of the perovskite solar cell based on the NLP process according to claim 5, wherein in step S401, the first principle calculation is performed by using a cast, the cutoff energy is 430eV, and the density functional is perdu-beck-enchoff.
CN202310156361.4A 2023-02-23 2023-02-23 Perovskite solar cell HTL material prediction method based on NLP (non-linear liquid phase) processing Pending CN116153446A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310156361.4A CN116153446A (en) 2023-02-23 2023-02-23 Perovskite solar cell HTL material prediction method based on NLP (non-linear liquid phase) processing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310156361.4A CN116153446A (en) 2023-02-23 2023-02-23 Perovskite solar cell HTL material prediction method based on NLP (non-linear liquid phase) processing

Publications (1)

Publication Number Publication Date
CN116153446A true CN116153446A (en) 2023-05-23

Family

ID=86359796

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310156361.4A Pending CN116153446A (en) 2023-02-23 2023-02-23 Perovskite solar cell HTL material prediction method based on NLP (non-linear liquid phase) processing

Country Status (1)

Country Link
CN (1) CN116153446A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116825227A (en) * 2023-08-31 2023-09-29 桑若(厦门)光伏产业有限公司 Perovskite component proportion analysis method and device based on depth generation model
CN117275634A (en) * 2023-11-20 2023-12-22 桑若(厦门)光伏产业有限公司 Perovskite solar cell design method and device based on machine learning

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116825227A (en) * 2023-08-31 2023-09-29 桑若(厦门)光伏产业有限公司 Perovskite component proportion analysis method and device based on depth generation model
CN116825227B (en) * 2023-08-31 2023-11-14 桑若(厦门)光伏产业有限公司 Perovskite component proportion analysis method and device based on depth generation model
CN117275634A (en) * 2023-11-20 2023-12-22 桑若(厦门)光伏产业有限公司 Perovskite solar cell design method and device based on machine learning
CN117275634B (en) * 2023-11-20 2024-05-28 桑若(厦门)光伏产业有限公司 Perovskite solar cell design method and device based on machine learning

Similar Documents

Publication Publication Date Title
CN116153446A (en) Perovskite solar cell HTL material prediction method based on NLP (non-linear liquid phase) processing
Liu et al. How machine learning predicts and explains the performance of perovskite solar cells
CN109344234A (en) Machine reads understanding method, device, computer equipment and storage medium
CN115659234A (en) Heterogeneous graph representation learning method integrating text attributes
Li et al. Deep learning modeling strategy for material science: from natural materials to metamaterials
Kou et al. Deep learning approach for wind speed forecasts at turbine locations in a wind farm
CN111026880A (en) Joint learning-based judicial knowledge graph construction method
CN114490954B (en) Document level generation type event extraction method based on task adjustment
CN112598039A (en) Method for acquiring positive sample in NLP classification field and related equipment
Zhang et al. Unsupervised machine learning for solar cell materials from the literature
Gilligan et al. A rule-free workflow for the automated generation of databases from scientific literature
Wei et al. Entity relationship extraction based on bi-LSTM and attention mechanism
Chan et al. Sentiment analysis by using Naïve‐Bayes classifier with stacked CARU
Zhang et al. Judicial nested named entity recognition method with MRC framework
Wang et al. Beyond word2vec: Distance-graph tensor factorization for word and document embeddings
Li et al. Text sentiment analysis based on glove model and united network
Long et al. Performance prediction based on neural architecture features
Qi et al. Cgtr: Convolution graph topology representation for document ranking
Huang et al. Hierarchical Self‐Attention Hybrid Sparse Networks for Document Classification
CN114064888A (en) Financial text classification method and system based on BERT-CNN
Zhang et al. CPPFEE: A Cascade Pointer Prediction Framework for Financial Event Extraction
Park et al. 1.5 million materials narratives generated by chatbots
Wu et al. Review of attention mechanism in electric power systems
Xu et al. SCA-ECPE: Emotion-Cause Pair Extraction Based on Sentiment Clustering Analysis
Imran et al. Leveraging FourierKAN Classification Head for Pre-Trained Transformer-based Text Classification

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