Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: May 31, 2026

Interaction between Phonological and Semantic Processes in Visual Word Recognition using Electrophysiology
05:38

Interaction between Phonological and Semantic Processes in Visual Word Recognition using Electrophysiology

Published on: June 29, 2021

Multi-dimensional text feature fusion-based BA-RILA for ancient Chinese poetry theme recognition.

Xinlu Zhang1,2, Yuemin Liu3

  • 1School of Foreign Studies, Changsha University of Science and Technology, Changsha, China.

Scientific Reports
|May 28, 2026
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Aligning climate and air quality goals demands early VOC control for China's ozone pollution.

Nature communications·2026
Same author

Identification and antialgal mechanisms of key allelochemicals from Chlorella pyrenoidosa culture filtrate.

Environmental pollution (Barking, Essex : 1987)·2026
Same author

Microcolin A, a lipopeptide from marine cyanobacteria, suppresses the non-small cell lung carcinoma by regulating the C-terminal of Hsp90α and ROS/ERK signaling pathways.

Biochemical pharmacology·2026
Same author

Improving immune function and mitochondrial health in patients undergoing hemodialysis: benefits of combining high-flux dialysis with hemoperfusion.

Frontiers in medicine·2026
Same author

Divergent microbial strategies under saline-alkali stress: a comparative study of bacterial and fungal communities in four agro-pastoral habitats.

Frontiers in microbiology·2026
Same author

Ultrasound-responsive bacteria-nanocomplex reverses tumor resistance to sensitize chemotherapy/chemodynamic therapy in triple-negative breast cancer.

Acta biomaterialia·2026

This study introduces BA-RILA, a new method for ancient poetry theme recognition using combined semantic, rhythmic, and imagery features. The framework significantly improves theme identification accuracy for digitalized traditional Chinese culture.

Area of Science:

  • Natural Language Processing (NLP)
  • Digital Humanities
  • Computational Linguistics

Background:

  • Digitalization of traditional Chinese culture necessitates accurate ancient poetry theme identification.
  • Existing NLP models struggle with limited ancient Chinese corpora and poor adaptability.
  • General-purpose models lack the specialized features required for nuanced poetry analysis.

Purpose of the Study:

  • To propose a novel ancient poetry theme recognition method, BA-RILA (BACM-Attention-Rhythm-Imagery-BiLSTM-Aided Framework).
  • To address the scarcity of ancient Chinese corpora and the limitations of current NLP models.
  • To enhance the accuracy and generalization ability of theme identification in classical Chinese poetry.

Main Methods:

  • Developed a three-dimensional text feature fusion framework integrating semantics, rhythm, and imagery.
Keywords:
Ancient poetry theme recognitionBERT-ancient Chinese pre-trained model (BACM)BiLSTMMulti-dimensional text feature fusionMulti-head attention

Related Experiment Videos

Last Updated: May 31, 2026

Interaction between Phonological and Semantic Processes in Visual Word Recognition using Electrophysiology
05:38

Interaction between Phonological and Semantic Processes in Visual Word Recognition using Electrophysiology

Published on: June 29, 2021

  • Utilized an optimized BERT ancient Chinese pre-trained model (BACM) for semantic vector extraction.
  • Employed an attention-weighting scheme to fuse heterogeneous features, followed by a two-layer BiLSTM and an 8-head multi-head attention (MHA) mechanism.

    • Main Results:

    • The proposed BA-RILA method significantly outperforms benchmark models in poetry theme recognition.
    • Experimental results demonstrate strong generalization ability across different historical dynasties (Tang and Song).
    • The fusion of semantic, rhythmic, and imagery features proved effective for capturing complex textual dependencies.

    Conclusions:

    • BA-RILA offers a robust solution for ancient poetry theme identification, crucial for digital cultural heritage.
    • The multi-dimensional feature fusion and attention mechanisms enhance model performance and adaptability.
    • This approach paves the way for more sophisticated NLP applications in classical Chinese literature analysis.