| 智能网联交通混合标签感知的推荐预测模型 |
| |
| 引用本文： | 李湘媛,丁飞,任素菊,张登银,康忆宁.智能网联交通混合标签感知的推荐预测模型[J].重庆邮电大学学报(自然科学版),2023,35(4):688-695. |
| |
| 作者姓名： | 李湘媛 丁飞 任素菊 张登银 康忆宁 |
| |
| 作者单位： | 南京邮电大学 江苏省宽带无线通信和物联网重点实验室,南京 210003;南京邮电大学 物联网学院,南京 210003 |
| |
| 基金项目： | 国家自然科学基金项目（61871446，61872423）；江苏省“六大人才高峰”高层次人才资助项目（DZXX-008）；中国博士后科学基金面上资助项目（2019M661900）；江苏省博士后科研资助计划（2019K026）；南京邮电大学科研基金资助项目（NY220028） |
| |
| 摘 要： | 针对智能网联车辆高速移动以及智能网联组网模式多元化导致的传统协同过滤算法有效性受到限制的问题，提出一种新型混合标签感知推荐模型（hybrid tag-aware recommender model，HTRM）。嵌入层采用Word2Vec模型对项目标签、项目评分、用户行为标签和用户评分进行向量表示；特征层引入自编码器提取项目的自相似特征，采用长短期记忆网络（long short-term memory，LSTM）提取用户行为特征；门控层联合用户和项目的特征，并输入至全连接神经网络（fully connected neural network，FCNN）进行评分预测。实验结果表明，与TCF、CCF、ACF和DSPR传统模型相比，HTRM模型设计更合理，可以获得较高的推荐预测精度。
|
| 关 键 词： | 智能网联交通 推荐系统 协同过滤算法 标签感知 全连接神经网络 |
| 收稿时间： | 2022/5/28 0:00:00 |
| 修稿时间： | 2023/4/24 0:00:00 |
Hybrid tag-aware recommender prediction model in intelligent connected transportation |
| |
| LI Xiangyuan,DING Fei,REN Suju,ZHANG Dengyin,KANG Yining.Hybrid tag-aware recommender prediction model in intelligent connected transportation[J].Journal of Chongqing University of Posts and Telecommunications,2023,35(4):688-695. |
| |
| Authors: | LI Xiangyuan DING Fei REN Suju ZHANG Dengyin KANG Yining |
| |
| Affiliation: | Jiangsu Key Laboratory of Broadband Wireless Communication and Internet of Things, Nanjing University of Posts and Telecommunications, Nanjing 210003, P. R. China;School of Internet of Things, Nanjing University of Posts and Telecommunications, Nanjing 210003, P. R. China |
| |
| Abstract: | In the intelligent networked traffic scenario, different vehicles and traffic participants conduct web service interaction based on different terminal platforms and recommendation systems. The high-speed movement of vehicles and the diversification of intelligent network networking modes limit the effectiveness of traditional collaborative filtering algorithms. This paper proposes a new hybrid tag-aware recommender model (HTRM). The embedding layer uses the Word2Vec model to represent item tags, item ratings, user behavior tags, and user ratings as vectors. The feature layer introduces an autoencoder. The self-similar features of the items are extracted, and the long short-term memory network (LSTM) is used to extract the user behavior features. The gating layer combines the features of users and items, and inputs them to the fully connected neural network (FCNN) for scoring prediction. Finally, three public datasets are used to test the recommendation of the vehicle web information service and the model function is verified. Experimental results show that the HTRM model is reasonable compared with the traditional TCF, CCF, ACF and DSPR models, and the MAE performance is optimized by 34.1%, 28.9%, 23.9% and 16.9%, respectively, which can achieve higher recommendation prediction accuracy. |
| |
| Keywords: | intelligent connected transportation recommendation system collaborative filtering algorithm tag-aware fully connected neural network |
|
| 点击此处可从《重庆邮电大学学报(自然科学版)》浏览原始摘要信息 |
|
点击此处可从《重庆邮电大学学报(自然科学版)》下载全文 |
|
