上海理工大学,管理学院;浙江大学,计算机科学与技术学院;
【背景】现有的出行推荐系统大多是根据特定目标，对各种交通方式的出行方案进行同质化推荐，未能考虑用户出行情景的异质性，导致难以满足用户个性化的出行需求。【目标】提出一种基于大语言模型的多模式出行方案个性化推荐算法，通过时空上下文感知，将多维异构的个性化出行情景表征为统一空间的嵌入向量，以更精准地刻画用户的出行需求，从而为用户推荐最匹配的出行方案。【数据】基于百度地图的历史出行大数据，提取用户出行查询记录、出行方案推荐记录、点击记录和用户属性等信息。【方法】利用BERT(Bidirectional Encoder Representations from Transformers)大语言模型将用户每次查询所对应的推荐出行方案集文本描述转化为语义向量，并与出行时段、用户属性等特征进行融合，构建综合表征用户个性化出行情景时空异质性的特征向量。将个性化出行情景特征向量输入多层感知器（Multilayer Perceptron, MLP）模型，预测用户最有可能选择的交通方式，作为推荐的最佳出行方案。【结果】基于BERT语义嵌入与MLP构建的多模式出行方案个性化推荐模型能够较为精确地辨识用户的差异化出行需求，其加权F1分数高于已有研究中使用相同数据集的其他模型。【应用】应用大语言模型感知用户个性化出行情景，可以显著提升出行推荐系统的智能化和个性化水平，对于优化出行服务质量，提高用户出行满意度具有重要作用。
| 634 | 0 | 70 |
| 下载次数 | 被引频次 | 阅读次数 |
[1]孙全明,曲志坚,任崇广.基于粒子群优化和LightGBM的情景感知多式联运推荐[J].电子学报, 2021, 49(5):894-903.SUN Quanming, QU Zhijian, REN Chongguang. Contextaware multi-modal transportation recommendation based on particle swarm optimization and LightGBM[J]. Acta Electronica Sinica, 2021, 49(5):894-903.
[2] HUANG Y Q, GAO L J, NI A N, et al. Analysis of travel mode choice and trip chain pattern relationships based on multi-day GPS data:a case study in Shanghai, China[J].Journal of Transport Geography, 2021, 93:103070.
[3] LIU L, XU J, LIAO S S, et al. A real-time personalized route recommendation system for self-drive tourists based on vehicle to vehicle communication[J]. Expert Systems with Applications, 2014, 41(7):3409-3417.
[4] HERZOG D, MASSOUD H, W?RNDL W. RouteMe:a mobile recommender system for personalized, multi-modal route planning[C]//Proceedings of the 25th Conference on User Modeling, Adaptation and Personalization.Bratislava Slovakia:ACM, 2017:67-75.
[5] DU R, ZHANG N, GAO X, et al. Optimal path choice based on multi-modal public transport—a case study of the Chengdu Qinghua Road area[C]//ICTE 2015. Dailan:American Society of Civil Engineers, 2015:1682-1688.
[6] SOCHAROENTUM M, KARIMI H A. Multi-modal transportation with multi-criteria walking(MMT-MCW):personalized route recommender[J]. Computers, Environment and Urban Systems, 2016, 55:44-54.
[7] DAI J, YANG B, GUO C, et al. Personalized route recommendation using big trajectory data[C]//2015 IEEE 31st International Conference on Data Engineering. Seoul, Korea(South):IEEE, 2015:543-554.
[8]段宗涛,任国亮,康军,等.基于频繁轨迹序列模式挖掘的路径推荐方法[J].太原理工大学学报, 2022, 53(2):240-247.DUAN Zongtao, REN Guoliang, KANG Jun, et al. Route recommendation method based on frequent trajectory sequence pattern mining[J]. Journal of Taiyuan University of Technology, 2022, 53(2):240-247.
[9] CHEN Z B, SHEN H T, ZHOU X F. Discovering popular routes from trajectories[C]//2011 IEEE 27th International Conference on Data Engineering. Hannover, Germany:IEEE, 2011:900-911.
[10] LIU H, LI T, HU R J, et al. Joint representation learning for multi-modal transportation recommendation[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2019, 33(1):1036-1043.
[11] WU F, LYU C, LIU Y. A personalized recommendation system for multi-modal transportation systems[J]. Multimodal Transportation, 2022, 1(2):100016.
[12]孙全明,常磊,马铖,等.基于图嵌入和CaGBDT的多模态出行推荐[J].北京邮电大学学报, 2021, 44(5):81-87, 106.SUN Quanming, CHANG Lei, MA Cheng, et al. Multimodal transportation recommendation based on graph embedding and CaGBDT[J]. Journal of Beijing University of Posts and Telecommunications, 2021, 44(5):81-87, 106.
[13] LI W X, ZHANG Y L, CHEN Y F, et al. Multi-day activity pattern recognition based on semantic embeddings of activity chains[J]. Travel Behaviour and Society, 2024,34:100682.
[14]贺正冰.大语言模型在道路交通领域应用：创新与挑战[J/OL].交通运输工程与信息学报, 1-10(2024-12-11)[2024-12-11]. https://doi. org/10.19961/j. cnki. 1672-4747.2024.12.001.HE Zhengbing. Large language models in road transportation:innovations and challenges[J/OL]. Journal of Transportation Engineering and Information, 1-10(2024-12-11)[2024-12-11]. https://doi. org/10.19961/j. cnki.1672-4747. 2024. 12.001.
[15] JIN K, WI J, LEE E, et al. TrafficBERT:Pre-trained model with large-scale data for long-range traffic flow forecasting[J]. Expert Systems with Applications, 2021,186:115738.
[16] ZHANG K P, ZHOU F, WU L, et al. Semantic understanding and prompt engineering for large-scale traffic data imputation[J]. Information Fusion, 2024, 102:102038.
[17] URAIBI H S. VIF-regression screening ultrahigh dimensional feature space[J]. Journal of Modern Applied Statistical Methods, 2021, 19(1):3.
[18] DEVLIN J, CHANG M W, LEE K, et al. BERT:pretraining of deep bidirectional transformers for language understanding[EB/OL]. arXiv, 2018:1810.04805. https://arxiv.org/abs/1810.04805v2.
[19] CAR Z,?EGOTA S B, AN?ELI?N, et al. Modeling the spread of COVID-19 infection using a multilayer perceptron[J]. Computational and Mathematical Methods in Medicine, 2020, 2020:5714714.
[20] AYD?N Y, CAKIROGLU C, BEKDA?G, et al. Explainable ensemble learning and multilayer perceptron modeling for compressive strength prediction of ultra-high-performance concrete[J]. Biomimetics, 2024, 9(9):544.
[21] AL BATAINEH A, MANACEK S. MLP-PSO hybrid algorithm for heart disease prediction[J]. Journal of Personalized Medicine, 2022, 12(8):1208.
[22] HEIDARI A A, FARIS H, ALJARAH I, et al. An efficient hybrid multilayer perceptron neural network with grasshopper optimization[J]. Soft Computing, 2019, 23:7941-7958.
[23] GéRON A. Hands-on machine learning with ScikitLearn, Keras, and TensorFlow(2nd Edition)[M]. O'Reilly, 2022.
[24] ABEDALLA A, FADEL A, TUFFAHA I, et al. MTRecSDLT:multi-modal transport recommender system using deep learning and tree models[C]//2019 Sixth International Conference on Social Networks Analysis, Management and Security(SNAMS). Granada, Spain:IEEE,2019:274-278.
[25] ZHU M, HU J, Hao, et al. Personalized context-aware multi-modal transportation recommendation[EB/OL].arXiv, 2019:1910.12601. https://arxiv. org/abs/1910.12601v1.
[26] LIU Y, LYU C, LIU Z Y, et al. Exploring a large-scale multi-modal transportation recommendation system[J].Transportation Research Part C:Emerging Technologies, 2021, 126:103070.
基本信息:
DOI：10.19961/j.cnki.1672-4747.2024.11.022
中图分类号:TP391.3;U12
引用信息:
[1]李文翔,丁龙远,张逸文等.基于大语言模型的多模式出行方案个性化推荐[J].交通运输工程与信息学报,2025,23(01):93-103.DOI:10.19961/j.cnki.1672-4747.2024.11.022.
基金信息:
国家自然科学基金项目（72471149）; 教育部人文社会科学研究项目（24YJCZH147）; 上海市哲学社会科学规划青年课题（2023ECK003）; 上海市教育委员会“人工智能促进科研范式改革赋能学科跃升计划专项”