计算机科学 ›› 2025, Vol. 52 ›› Issue (6A): 240500127-8.doi: 10.11896/jsjkx.240500127
赵哲宇, 王中卿, 王红玲
ZHAO Zheyu, WANG Zhongqing, WANG Hongling
摘要： 商品属性分类任务是指对一段商品的描述文字进行属性分析并进而对多个属性进行分类的过程,其有助于人们从多个角度了解商品,为市场营销、产品管理等提供帮助。当前大语言模型的使用也愈加广泛,但在商品属性分类问题上,通用大模型由于缺乏领域知识和属性关联等信息,性能不够理想。为此,提出了一个基于双重预训练的商品属性分类方法,旨在通过使用特定的预训练方式提高大语言模型在商品属性分类任务中的性能。在T5模型的基础上,引入了领域内文本预训练和基于属性间关联性的预训练两种方法。在Clothing Fit Data数据集上的实验结果显示,使用了双重预训练的T5模型较未经过预训练的模型以及其他基准模型,在各个属性上的分类效果都取得了一定提升。实验结果证明了所提方法的有效性。
中图分类号:
| [1]WU Z,DAI X Y,YINC,et al.Improving review representations with user attention and product attention for sentiment classification[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2018. [2]ABDULLA G M,BORAR S.Size recommendation system forfashion e-commerce[C]//KDD Workshop on Machine Learning Meets Fashion.2017:1-7. [3]BHATT A,PATEL A,CHHEDA H,et al.Amazon review classification and sentiment analysis[J].International Journal of Computer Science and Information Technologies,2015,6(6):5107-5110. [4]CHENG Z J,ZHU Y H.SVM-based commodity scoring system[J].Computer Knowledge and Technology,2018,14(30):223-225. [5]HEARST M A,DUMAIS S T,OSUNA E,et al.Support vectormachines[J].IEEE Intelligent Systems and Their Applications,1998,13(4):18-28. [6]LUO H Q,LU X Y,ZHANG X B et al.Hidden parsimonious Bayes-based sentiment classification method for commodity reviews[J].Computer Engineering and Design,2017,38(1):203-208. [7]JIANG L,ZHANG H,CAI Z.A novel bayes model:Hidden naive bayes[J].IEEE Transactions on Knowledge and Data Engineering,2008,21(10):1361-1371. [8]FAYAZ M,KHAN A,RAHMAN J U,et al.Ensemble machine learning model for classification of spam product reviews[J].Complexity,2020,2020:1-10. [9]TAUD H,MAS J F.Multilayer perceptron(MLP)[M]//Geomatic Approaches for Modeling LAND Change Scenarios.Springer,2018:451-455. [10]KRISHNA K,MURTY M N.Genetic K-means algorithm[J].IEEE Transactions on Systems,Man,and Cybernetics,Part B(Cybernetics),1999,29(3):433-439. [11]RIGATTIS J.Random forest[J].Journal of Insurance Medi-cine,2017,47(1):31-39. [12]PENG S C.Research on Sentiment Classification of Commodity Reviews Based on RNN and LDA Models [D].Hangzhou:Zhejiang University of Technology,2018. [13]ZAREMBA W,SUTSKEVER I,VINYALS O.Recurrent neural network regularization[J].arXiv:1409.2329,2014. [14]BLEI D M,NG A Y,JORDAN M I.Latent dirichlet allocation[J].Journal of Machine Learning Research,2003,3(Jan):993-1022. [15]CHEN H,SUN M,TUC,et al.Neural sentiment classification with user and product attention[C]//Proceedings of the 2016 Conference on Empirical Methods in Natural Language Proces-sing.2016:1650-1659. [16]SHI X,CHEN Z,WANG H,et al.Convolutional LSTM net-work:A machine learning approach for precipitation nowcasting[J].Advances in Neural Information Processing Systems,2015,28. [17]KENTON J D M W C,TOUTANOVA L K.BERT:Pre-training of Deep Bidirectional Transformers for Language Understanding[C]//Proceedings of NAACL-HLT.2019:4171-4186. [18]XU M M.Research on commodity classification method based on BERT [D].Nanchang:Nanchang University,2023. [19]YOON K.Convolutional Neural Networks for Sentence Classifications[C]//Conference on Empirical Methods in Natural Language Processing.Association for Computational Linguistic,2014:1746-1751 [20]ZHANG S Q,MA J,ZHOU X B,et al.Commodity attribute extraction based on pre-trained language models[J].Journal of Chinese Information,2022,36(1):56-64. [21]CLARK K,LUONG M T,LEQ V,et al.Electra:Pre-trainingtext encoders as discriminators rather than generators[J].ar-Xiv:2003.10555,2020. [22]ESHEL Y,LEVI O,ROITMANH,et al.Presize:predicting size in e-commerce using transformers[C]//Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval.2021:255-264. [23]CHATTERJEE O,TEJ J R,DASARAJUN V.Incorporatingcustomer reviews in size and fit recommendation systems for fashion e-commerce[J].arXiv:2208.06261,2022. [24]SUN C,QIU X,XU Y,et al.How to fine-tune bert for text clas-sification?[C]//Chinese Computational Linguistics:18th China National Conference,CCL 2019,Kunming,China,October 18-20,2019,Proceedings 18.Springer International Publishing,2019:194-206. [25]BAO X,WANG Z,ZHOU G.Exploring graph pre-training for aspect-based sentiment analysis[C]//Findings of the Association for Computational Linguistics:EMNLP 2023.2023:3623-3634. [26]RAFFEL C,SHAZEER N,ROBERTS A,et al.Exploring the limits of transfer learning with a unified text-to-text transformer[J].The Journal of Machine Learning Research,2020,21(1):5485-5551. [27]MISRA R,WAN M,MCAULEY J.Decomposing fit semantics for product size recommendation in metric spaces[C]//Procee-dings of the 12th ACM Conference on Recommender Systems.2018:422-426. [28]PAPINENI K,ROUKOS S,WARD T,et al.Bleu:a method for automatic evaluation of machine translation[C]//Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics.2002:311-318. [29]YANG Z,DAI Z,YANGY,et al.Xlnet:Generalized autoregressive pretraining for language understanding[J].Advances in Neural Information Processing Systems,2019,32. [30]OUYANG L,WU J,JIANG X,et al.Training language models to follow instructions with human feedback[J].Advances in Neural Information Processing Systems,2022,35:27730-27744. [31]TOUVRON H,MARTIN L,STONE K,et al.Llama 2:OpenFoundation and Fine-Tuned Chat Models[J].arXiv:2307.092882023. |
|
||
首页