Volume 2,Issue 7
融合BERT 语义表征与残差特征增强的O2O食品安全风险监测研究
随着O2O 餐饮模式的快速发展，用户评论已成为识别食品安全隐患的关键数据源。然而，O2O 评论存在明显的口语化、碎片化与语义模糊等特点，传统机器学习及单一深度学习模型在微调过程中难以兼顾全局语义理解与局部细粒度特征的精准提取，易造成食品安全风险信号漏检。针对上述挑战，本文提出一种融合BERT 预训练语言模型与残差网络的食品安全风险识别模型。该模型首先利用BERT 获取文本深层双向语义表示，构建全局特征基准；随后通过堆叠残差块对隐层状态序列进行多尺度局部特征挖掘，借助跳跃连接缓解深层网络微调时的梯度退化问题，增强模型对“异物”、“发臭”、“致病”等风险关键词的感知能力；最后采用特征拼接策略融合全局与局部特征向量，实现风险评论的自动判别。基于CCF BDCI 提供的10000条真实O2O 评论数据集进行实验。结果表明：本文提出的BERT+ Residual Block 模型在验证集上F1值最高可达0.9257，整体F1值为0.9096，相较于CNN + BERT 和BiLSTM +BERT 模型更具稳定性；测试集全量准确率达到97.5%，五折交叉验证的标准差仅为0.0062，展现出较强泛化能力与鲁棒性；模型在风险类别的F1分数达到0.92，可有效降低智慧监管场景下的风险漏检率。本文验证了残差结构在提升预训练模型特征提取能力的有效性，为O2O 场景下食品安全数字化监测与预警提供一种高效技术方案。未来研究将进一步引入外部知识图谱注入与多模态融合技术，增强模型对复杂反讽语义的识别能力。
[1] 宋超玉, 方美芳.O2O 模式在大学及周边的发展研究—— 以蚌埠高校为例[J]. 山西农经,2018,(06):110-111.DOI:10.16675/j.cnki.cn14-1065/f.2018.06.083.
[2]Zhang Q, Wu A, Liu L, et al. Application of Machine Learning in Food Safety Risk Assessment[J]. Foods, 2025, 14(23): 4005.
[3]Pang B, Lee L, Vaithyanathan S. Thumbs up? Sentiment classification using machine learning techniques[C]// Proceedings of the ACL-02 Conference on Empirical Methods in Natural Language Processing. 2002: 79-86.
[4]Kim Y. Convolutional Neural Networks for Sentence Classification[C]// Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2014: 1746-1751.
[5]Tang D, Qin B, Liu T. Document Modeling with Gated Recurrent Neural Network for Sentiment Classification[C]// Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2015: 1422-1432.
[6]Devlin J, Chang M W, Lee K, et al. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding[EB/OL]. (2018-10-11). https://arxiv.org/abs/1810.04805.
[7]He K, Zhang X, Ren S, et al. Deep Residual Learning for Image Recognition[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).2016: 770-778.
[8] 梁小林, 柳映堂, 梁曌, 等. 基于偏最小二乘支持向量机模型的个人信用评估研究[J]. 湖南文理学院学报( 自然科学版),2021,33(04):6-10.
[9]DEVLIN J, CHANG M W, LEE K, et al. BERT: Pre-training of deep bidirectional transformers for language understanding[J]. arXiv preprint arXiv:1810.04805, 2018.
[10]CORTES C, VAPNIK V. Support-vector networks [J]. Machine Learning, 1995, 20(3): 273-297.
[11]CHEN T, GUESTRIN C. XGBoost: A Scalable Tree Boosting System [C]// Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 2016: 785-794.
[12]DEVLIN J, CHANG M W, LEE K, et al. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding [J]. arXiv preprint arXiv:1810.04805, 2018.
[13]KIM Y. Convolutional Neural Networks for Sentence Classification [C]// Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2014: 1746-1751.
[14]GRAVES A, SCHMIDHUBER J. Framewise phoneme classification with bidirectional LSTM and other neural network architectures [J]. Neural Networks, 2005, 18(5-6): 602-610.
[15]HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition [C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2016: 770-778.
[16]DEVLIN J, CHANG M W, LEE K, et al. BERT: Pre-training of deep bidirectional transformers for language understanding [J]. arXiv preprint arXiv:1810.04805, 2018.