RoBERTa-GCN: A New Method for Relation Extraction in Automobile Accessory Domain
DOI:
https://doi.org/10.30564/jcsr.v6i3.6695Abstract
The automotive industry's rapid expansion has sparked increasing interest in the realm of automotive accessories. Navigating vast information landscapes to find accurate matches has become paramount. Leveraging cutting-edge information technologies, such as knowledge graphs and graph database-based question-answering systems, offers a crucial avenue for enhancing search efficiency. Addressing challenges posed by the domain's specialized terminology and intricate relationships, this paper introduces an innovative approach that combines a pre-trained model (RoBERTa) with graph convolutional networks (GCN). Initially, the text undergoes processing through the pre-trained model, yielding semantic feature vectors that enhance comprehension of industry-specific terminology. Subsequently, a graph convolutional network (GCN) is employed to process these semantic vectors, capturing a broader scope of neighboring vector node information. This approach not only strengthens the relationships between semantic information but also captures the intricate interconnections among entities. Ultimately, an automotive accessory query knowledge graph question-answering system is constructed using extracted entity relationship triplets. Experimental results demonstrate that the proposed RoBERTa-GCN model outperforms other baseline models, achieving an impressive F1 score of 83.93%. This research significantly enhances query capabilities and exhibits versatility in handling natural language inputs from diverse users.
Keywords:
Knowledge graph; Relation extraction; Pre-trained model; Graph convolutional modeReferences
[1] Chen, X., Jia, S., Xiang, Y.J.E.S.w.A., 2020. A review: Knowledge reasoning over knowledge graph. 141, 112948.
[2] Liu, Y., et al., 2019. Roberta: A robustly optimized bert pretraining approach.
[3] Pei, H., et al.,2020. Geom-gcn: Geometric graph convolutional networks.
[4] Wang, X., et al., 2023. Knowledge Graph of Urban Firefighting with Rule-Based Entity Extraction. in International Conference on Engineering Applications of Neural Networks. Springer.
[5] Wang, L.H., et al., 2023. Emergency entity relationship extraction for water diversion project based on pre-trained model and multi-featured graph convolutional network. 18(10), e0292004.
[6] Sherstinsky, A.J.P.D.N.P., 2020. Fundamentals of recurrent neural network (RNN) and long short-term memory (LSTM) network. 404, 132306.
[7] Sarker, I.H.J.S.C.S., 2021. Deep learning: a comprehensive overview on techniques, taxonomy, applications and research directions. 2(6), 420.
[8] She, D., Jia, M.J.M., 2021. A BiGRU method for remaining useful life prediction of machinery. 167, 108277.
[9] Kattenborn, T., et al., 2021. Review on Convolutional Neural Networks (CNN) in vegetation remote sensing. 173, 24–49.
[10] Barthelmy, S., et al., 1998. The GRB coordinates network (GCN): A status report. in AIP Conference Proceedings. American Institute of Physics.
[11] Devlin, J., et al., 2018. Bert: Pre-training of deep bidirectional transformers for language understanding.
[12] Liu, X., et al., 2023. GPT understands, too.
[13] Kaur, K., Kaur, P.J.P.C.S., 2023. BERT-CNN: Improving BERT for Requirements Classification using CNN. 218, 2604–2611.
[14] Di Gennaro, G., Buonanno, A., Palmieri, F.A.J.T.J.o.S., 2021. Considerations about learning Word2Vec. 1–16.
[15] Yang, Z., et al., 2019. Xlnet: Generalized autoregressive pretraining for language understanding. 32.
[16] Wu, C., et al., 2022. Machine Reading Comprehension Based on SpanBERT and Dynamic Convolutional Attention. in Proceedings of the 4th International Conference on Advanced Information Science and System.
[17] Zhang, Z., et al., 2019. ERNIE: Enhanced language representation with informative entities.
[18] Shi, W.,. Song, M., Wang, Y., 2022. Perturbation-enhanced-based RoBERTa combined with BiLSTM model for Text classification. in ICETIS 2022; 7th International Conference on Electronic Technology and Information Science. VDE.
[19] Li, Z., et al., 2021. Causality extraction based on self-attentive BiLSTM-CRF with transferred embeddings. 423, 207–219.
[20] Jin, J., Zhao, Y., Cui, R., 2021. A Comparative Study of Korean Feature Granularity Based on Hybrid Neural Network. in 2021 IEEE International Conference on Power Electronics, Computer Applications (ICPECA). IEEE.
[21] Meng, F., et al., 2022. Creating knowledge graph of electric power equipment faults based on BERT–BiLSTM–CRF model. 17(4), 2507–2516.
[22] Aslan, M.F., et al., 2021. CNN-based transfer learning–BiLSTM network: A novel approach for COVID-19 infection detection. 98, 106912.
[23] Li, W., et al., 2022. UD_BBC: Named entity recognition in social network combined BERT-BiLSTM-CRF with active learning. 116, 105460.
Downloads
How to Cite
Issue
Article Type
License
Copyright © 2024 Author(s)
This is an open access article under the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) License.