电子学报 ›› 2022, Vol. 50 ›› Issue (10): 2361-2371.DOI: 10.12263/DZXB.20210443
张帅1,2, 高旻1,2(), 文俊浩1,2, 熊庆宇1,2, 唐旭2
收稿日期:
2021-04-07
修回日期:
2021-09-02
出版日期:
2022-10-25
通讯作者:
作者简介:
基金资助:
ZHANG Shuai1,2, GAO Min1,2(), WEN Jun-hao1,2, XIONG Qing-yu1,2, TANG Xu2
Received:
2021-04-07
Revised:
2021-09-02
Online:
2022-10-25
Published:
2022-10-11
Corresponding author:
摘要:
近年来,随着推荐系统研究的不断深入,推荐系统的公平性受到越来越多关注. 流行度偏差也即流行的物品比非流行的物品更容易被推荐,是影响其公平性的重要因素之一. 流行度偏差对推荐系统的各利益相关者都有严重的影响,引起研究者的广泛关注. 相关研究主要通过推荐结果重排或学习过程中融合正则化项提升非流行物品的曝光率,而非流行物品的交互数据极度稀疏成为研究的瓶颈. 针对此问题,本文提出基于自监督学习的去流行度偏差推荐方法,解决两个难点:(1)准确学习交互数据极度稀疏的非流行物品的表征;(2)提升非流行物品曝光率的同时,兼顾不同用户对流行和非流行物品的偏好. 具体地,从用户的角度,提出流行物品和非流行物品双视图的用户偏好学习方法,准确学习用户对流行和非流行物品的真实偏好;从物品的角度,采用自监督学习,利用互信息最大化捕获非流行物品与流行物品间的潜在关系,辅助提升非流行物品嵌入学习的准确性. 最后,设计用户流行度偏好一致性、资格公平性等指标,并通过三个公开数据集的大量实验说明了本文方法在提升推荐性能的同时,能有效缓解流行度偏差问题并具有较强的通用性.
中图分类号:
张帅, 高旻, 文俊浩, 熊庆宇, 唐旭. 基于自监督学习的去流行度偏差推荐方法[J]. 电子学报, 2022, 50(10): 2361-2371.
ZHANG Shuai, GAO Min, WEN Jun-hao, XIONG Qing-yu, TANG Xu. Self-Supervised Learning for Alleviating Popularity Bias in Recommender Systems[J]. Acta Electronica Sinica, 2022, 50(10): 2361-2371.
数据集 | 用户数 | 物品数 | 反馈数 | 稀疏度 |
---|---|---|---|---|
Douban | 2 562 | 30 679 | 441 935 | 99.44% |
LastFM | 1 867 | 17 487 | 73 698 | 99.77% |
Epinions | 40 163 | 139 736 | 664 824 | 99.99% |
表1 数据集统计
数据集 | 用户数 | 物品数 | 反馈数 | 稀疏度 |
---|---|---|---|---|
Douban | 2 562 | 30 679 | 441 935 | 99.44% |
LastFM | 1 867 | 17 487 | 73 698 | 99.77% |
Epinions | 40 163 | 139 736 | 664 824 | 99.99% |
Douban | ||||||||
---|---|---|---|---|---|---|---|---|
方法 | ||||||||
MF-BPR+Top-K | 0.041 8 | 0.121 1 | 131.69% | 145.81% | 151.64% | 9.963 7 | 0.055 8 | 0.005 6 |
Random-K | 0.000 3 | 0.000 8 | -87.48% | -94.16% | -92.11% | 1.989 9 | 0.602 6 | 0.801 0 |
Poorest-K | 0 | 0 | -99.17% | -99.64% | -99.42% | 0 | 0.000 5 | 1 |
Most Popular | 0.032 7 | 0.099 7 | 573.66% | 190.37% | 294.74% | 10.000 0 | 0.001 7 | 0 |
MF-BPR+FairRec | 0.035 5 | 0.101 8 | -74.11% | -69.54% | -69.08% | 4.134 9 | 0.185 3 | 0.155 0 |
MF-BPR+xQuAD | 0.031 6 | 0.088 6 | 24.12% | 152.97% | 147.74% | 8.923 0 | 0.098 1 | 0.155 0 |
MF-CMIPS | 0.030 8 | 0.082 1 | 53.92% | 81.30% | 70.96% | 8.711 4 | 0.175 1 | 0.155 0 |
MF-CIPS | 0.032 8 | 0.085 4 | 59.98% | 90.77% | 75.49% | 8.889 2 | 0.151 8 | 0.155 0 |
MF-ESAM | 0.033 4 | 0.093 1 | 45.24% | 71.28% | 62.52% | 8.586 5 | 0.121 1 | 0.155 0 |
MF-SSLAB(本文方法) | 0.040 2 | 0.117 3 | 20.26% | 41.77% | 37.73% | 7.392 0 | 0.195 3 | 0.155 0 |
LastFM | ||||||||
方法 | ||||||||
MF-BPR+Top-K | 0.160 4 | 0.148 3 | 164.33% | 89.91% | 123.18% | 9.722 0 | 0.102 9 | 0.028 0 |
Random-K | 0.000 3 | 0.000 2 | -80.66% | -95.74% | -92.62% | 2.025 8 | 0.699 2 | 0.797 4 |
Poorest-K | 0 | 0 | -95.66% | -99.14% | -98.43% | 0 | 0.001 3 | 1 |
Most Popular | 0.060 1 | 0.062 2 | 1 557.07% | 197.77% | 431.87% | 10.000 0 | 0.001 7 | 0 |
MF-BPR+FairRec | 0.118 4 | 0.128 6 | 82.20% | 53.04% | 88.80% | 8.463 7 | 0.199 5 | 0.217 0 |
MF-BPR+xQuAD | 0.090 1 | 0.118 5 | 94.33% | 83.46% | 95.39% | 8.763 4 | 0.126 9 | 0.217 0 |
MF-CMIPS | 0.081 7 | 0.108 8 | 87.54% | 76.87% | 83.47% | 8.637 7 | 0.189 1 | 0.217 0 |
MF-CIPS | 0.105 1 | 0.129 4 | 90.19% | 81.42% | 84.66% | 8.846 6 | 0.169 5 | 0.217 0 |
MF-ESAM | 0.099 2 | 0.106 1 | 124.36% | 96.58% | 129.35% | 9.234 9 | 0.106 3 | 0.217 0 |
MF-SSLAB(本文方法) | 0.152 8 | 0.138 7 | 36.86% | 48.12% | 47.58% | 7.477 6 | 0.209 2 | 0.217 0 |
Epinions | ||||||||
方法 | ||||||||
MF-BPR+Top-K | 0.034 2 | 0.019 5 | 86.75% | 202.15% | 249.61% | 9.708 8 | 0.095 8 | 0.028 2 |
Random-K | 0.000 05 | 0.000 03 | -77.93% | -92.73% | -95.45% | 1.983 1 | 0.846 5 | 0.802 0 |
Poorest-K | 0 | 0 | -93.53% | -97.93% | -98.69% | 0 | 0.000 2 | 1 |
Most Popular | 0.020 5 | 0.015 7 | 3 379.34% | 1 003.55% | 600.06% | 10.000 0 | 0.000 2 | 0 |
MF-BPR+FairRec | 0.018 3 | 0.012 6 | 83.51% | 63.43% | 65.76% | 8.351 1 | 0.163 1 | 0.296 0 |
MF-BPR+xQuAD | 0.019 9 | 0.013 9 | 72.03% | 76.49% | 83.52% | 8.563 6 | 0.159 7 | 0.296 0 |
MF-CMIPS | 0.021 4 | 0.016 6 | 68.93% | 82.81% | 89.19% | 8.786 7 | 0.184 1 | 0.296 0 |
MF-CIPS | 0.017 2 | 0.011 3 | 85.90% | 75.14% | 93.01% | 8.962 7 | 0.175 2 | 0.296 0 |
MF-ESAM | 0.022 7 | 0.014 4 | 71.54% | 88.13% | 74.77% | 8.571 2 | 0.145 4 | 0.296 0 |
MF-SSLAB(本文方法) | 0.031 4 | 0.018 2 | 32.88% | 30.23% | 42.32% | 7.729 9 | 0.235 7 | 0.296 0 |
表2 实验结果对比
Douban | ||||||||
---|---|---|---|---|---|---|---|---|
方法 | ||||||||
MF-BPR+Top-K | 0.041 8 | 0.121 1 | 131.69% | 145.81% | 151.64% | 9.963 7 | 0.055 8 | 0.005 6 |
Random-K | 0.000 3 | 0.000 8 | -87.48% | -94.16% | -92.11% | 1.989 9 | 0.602 6 | 0.801 0 |
Poorest-K | 0 | 0 | -99.17% | -99.64% | -99.42% | 0 | 0.000 5 | 1 |
Most Popular | 0.032 7 | 0.099 7 | 573.66% | 190.37% | 294.74% | 10.000 0 | 0.001 7 | 0 |
MF-BPR+FairRec | 0.035 5 | 0.101 8 | -74.11% | -69.54% | -69.08% | 4.134 9 | 0.185 3 | 0.155 0 |
MF-BPR+xQuAD | 0.031 6 | 0.088 6 | 24.12% | 152.97% | 147.74% | 8.923 0 | 0.098 1 | 0.155 0 |
MF-CMIPS | 0.030 8 | 0.082 1 | 53.92% | 81.30% | 70.96% | 8.711 4 | 0.175 1 | 0.155 0 |
MF-CIPS | 0.032 8 | 0.085 4 | 59.98% | 90.77% | 75.49% | 8.889 2 | 0.151 8 | 0.155 0 |
MF-ESAM | 0.033 4 | 0.093 1 | 45.24% | 71.28% | 62.52% | 8.586 5 | 0.121 1 | 0.155 0 |
MF-SSLAB(本文方法) | 0.040 2 | 0.117 3 | 20.26% | 41.77% | 37.73% | 7.392 0 | 0.195 3 | 0.155 0 |
LastFM | ||||||||
方法 | ||||||||
MF-BPR+Top-K | 0.160 4 | 0.148 3 | 164.33% | 89.91% | 123.18% | 9.722 0 | 0.102 9 | 0.028 0 |
Random-K | 0.000 3 | 0.000 2 | -80.66% | -95.74% | -92.62% | 2.025 8 | 0.699 2 | 0.797 4 |
Poorest-K | 0 | 0 | -95.66% | -99.14% | -98.43% | 0 | 0.001 3 | 1 |
Most Popular | 0.060 1 | 0.062 2 | 1 557.07% | 197.77% | 431.87% | 10.000 0 | 0.001 7 | 0 |
MF-BPR+FairRec | 0.118 4 | 0.128 6 | 82.20% | 53.04% | 88.80% | 8.463 7 | 0.199 5 | 0.217 0 |
MF-BPR+xQuAD | 0.090 1 | 0.118 5 | 94.33% | 83.46% | 95.39% | 8.763 4 | 0.126 9 | 0.217 0 |
MF-CMIPS | 0.081 7 | 0.108 8 | 87.54% | 76.87% | 83.47% | 8.637 7 | 0.189 1 | 0.217 0 |
MF-CIPS | 0.105 1 | 0.129 4 | 90.19% | 81.42% | 84.66% | 8.846 6 | 0.169 5 | 0.217 0 |
MF-ESAM | 0.099 2 | 0.106 1 | 124.36% | 96.58% | 129.35% | 9.234 9 | 0.106 3 | 0.217 0 |
MF-SSLAB(本文方法) | 0.152 8 | 0.138 7 | 36.86% | 48.12% | 47.58% | 7.477 6 | 0.209 2 | 0.217 0 |
Epinions | ||||||||
方法 | ||||||||
MF-BPR+Top-K | 0.034 2 | 0.019 5 | 86.75% | 202.15% | 249.61% | 9.708 8 | 0.095 8 | 0.028 2 |
Random-K | 0.000 05 | 0.000 03 | -77.93% | -92.73% | -95.45% | 1.983 1 | 0.846 5 | 0.802 0 |
Poorest-K | 0 | 0 | -93.53% | -97.93% | -98.69% | 0 | 0.000 2 | 1 |
Most Popular | 0.020 5 | 0.015 7 | 3 379.34% | 1 003.55% | 600.06% | 10.000 0 | 0.000 2 | 0 |
MF-BPR+FairRec | 0.018 3 | 0.012 6 | 83.51% | 63.43% | 65.76% | 8.351 1 | 0.163 1 | 0.296 0 |
MF-BPR+xQuAD | 0.019 9 | 0.013 9 | 72.03% | 76.49% | 83.52% | 8.563 6 | 0.159 7 | 0.296 0 |
MF-CMIPS | 0.021 4 | 0.016 6 | 68.93% | 82.81% | 89.19% | 8.786 7 | 0.184 1 | 0.296 0 |
MF-CIPS | 0.017 2 | 0.011 3 | 85.90% | 75.14% | 93.01% | 8.962 7 | 0.175 2 | 0.296 0 |
MF-ESAM | 0.022 7 | 0.014 4 | 71.54% | 88.13% | 74.77% | 8.571 2 | 0.145 4 | 0.296 0 |
MF-SSLAB(本文方法) | 0.031 4 | 0.018 2 | 32.88% | 30.23% | 42.32% | 7.729 9 | 0.235 7 | 0.296 0 |
方法 | ||||||||
---|---|---|---|---|---|---|---|---|
NGCF-BPR+Top-K | 0.168 9 | 0.164 2 | 217.85% | 86.52% | 148.82% | 9.862 8 | 0.098 7 | 0.033 6 |
NGCF-BPR+FairRec | 0.119 5 | 0.131 2 | 89.06% | 67.56% | 86.87% | 8.578 8 | 0.208 6 | 0.217 |
NGCF-BPR+xQuAD | 0.108 3 | 0.129 0 | 71.37% | 84.32% | 64.72% | 8.493 7 | 0.219 1 | 0.217 |
NGCF-CMIPS | 0.097 2 | 0.113 2 | 78.47% | 81.38% | 70.91% | 8.633 9 | 0.212 5 | 0.217 |
NGCF-CIPS | 0.116 3 | 0.135 5 | 81.36% | 89.71% | 75.77% | 8.776 3 | 0.204 7 | 0.217 |
NGCF-ESAM | 0.102 1 | 0.108 6 | 83.24% | 75.57% | 55.09% | 8.513 1 | 0.193 0 | 0.217 |
NGCF-SSLAB(本文方法) | 0.157 4 | 0.156 7 | 36.39% | 47.82% | 33.14% | 7.315 1 | 0.231 8 | 0.217 |
表3 NGCF在LastFM数据集上的实验结果
方法 | ||||||||
---|---|---|---|---|---|---|---|---|
NGCF-BPR+Top-K | 0.168 9 | 0.164 2 | 217.85% | 86.52% | 148.82% | 9.862 8 | 0.098 7 | 0.033 6 |
NGCF-BPR+FairRec | 0.119 5 | 0.131 2 | 89.06% | 67.56% | 86.87% | 8.578 8 | 0.208 6 | 0.217 |
NGCF-BPR+xQuAD | 0.108 3 | 0.129 0 | 71.37% | 84.32% | 64.72% | 8.493 7 | 0.219 1 | 0.217 |
NGCF-CMIPS | 0.097 2 | 0.113 2 | 78.47% | 81.38% | 70.91% | 8.633 9 | 0.212 5 | 0.217 |
NGCF-CIPS | 0.116 3 | 0.135 5 | 81.36% | 89.71% | 75.77% | 8.776 3 | 0.204 7 | 0.217 |
NGCF-ESAM | 0.102 1 | 0.108 6 | 83.24% | 75.57% | 55.09% | 8.513 1 | 0.193 0 | 0.217 |
NGCF-SSLAB(本文方法) | 0.157 4 | 0.156 7 | 36.39% | 47.82% | 33.14% | 7.315 1 | 0.231 8 | 0.217 |
方法 | ||||||||
---|---|---|---|---|---|---|---|---|
LightGCN-BPR+Top-K | 0.178 1 | 0.175 9 | 96.24% | 68.53% | 79.61% | 8.748 9 | 0.148 4 | 0.037 6 |
LightGCN-BPR+FairRec | 0.127 1 | 0.153 1 | 86.37% | 58.71% | 70.74% | 8.363 4 | 0.212 6 | 0.217 |
LightGCN-BPR+xQuAD | 0.118 8 | 0.143 4 | 61.81% | 66.18% | 69.41% | 8.113 2 | 0.223 5 | 0.217 |
LightGCN-CMIPS | 0.109 7 | 0.133 6 | 87.54% | 76.87% | 83.47% | 8.486 6 | 0.207 9 | 0.217 |
LightGCN-CIPS | 0.132 4 | 0.146 5 | 90.19% | 81.42% | 84.66% | 8.631 9 | 0.198 9 | 0.217 |
LightGCN-ESAM | 0.121 6 | 0.127 5 | 89.21% | 95.52% | 84.17% | 8.678 4 | 0.218 2 | 0.217 |
LightGCN-SSLAB(本文方法) | 0.169 2 | 0.162 5 | 27.84% | 35.80% | 27.85% | 7.191 3 | 0.243 6 | 0.217 |
表4 LightGCN在LastFM数据集上的实验结果
方法 | ||||||||
---|---|---|---|---|---|---|---|---|
LightGCN-BPR+Top-K | 0.178 1 | 0.175 9 | 96.24% | 68.53% | 79.61% | 8.748 9 | 0.148 4 | 0.037 6 |
LightGCN-BPR+FairRec | 0.127 1 | 0.153 1 | 86.37% | 58.71% | 70.74% | 8.363 4 | 0.212 6 | 0.217 |
LightGCN-BPR+xQuAD | 0.118 8 | 0.143 4 | 61.81% | 66.18% | 69.41% | 8.113 2 | 0.223 5 | 0.217 |
LightGCN-CMIPS | 0.109 7 | 0.133 6 | 87.54% | 76.87% | 83.47% | 8.486 6 | 0.207 9 | 0.217 |
LightGCN-CIPS | 0.132 4 | 0.146 5 | 90.19% | 81.42% | 84.66% | 8.631 9 | 0.198 9 | 0.217 |
LightGCN-ESAM | 0.121 6 | 0.127 5 | 89.21% | 95.52% | 84.17% | 8.678 4 | 0.218 2 | 0.217 |
LightGCN-SSLAB(本文方法) | 0.169 2 | 0.162 5 | 27.84% | 35.80% | 27.85% | 7.191 3 | 0.243 6 | 0.217 |
1 | ABDOLLAHPOURI H. Popularity bias in ranking and recommendation[C]//Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and Society. Honolulu: ACM, 2019: 529-530. |
2 | VARGAS S, CASTELLS P. Improving sales diversity by recommending users to items[C]//Eighth ACM Conference on Recommender Systems. Foster City: ACM, 2014: 145-152. |
3 | MANSOURY M, ABDOLLAHPOURI H, PECHENIZKIY M, et al. Fair match: A graph-based approach for improving aggregate diversity in recommender systems[C]//Proceedings of the 28th ACM Conference on User Modeling, Adaptation and Personalization. Genoa: ACM, 2020: 154-162. |
4 | ABDOLLAHPOURI H, BURKE R, MOBASHER B. Managing popularity bias in recommender systems with personalized re-ranking[C]//Proceedings of the Thirty-Second International Florida Artificial Intelligence Research Society Conference. Sarasota: AAAI Press, 2019: 413-418. |
5 | ABDOLLAHPOURI H, BURKE R, MOBASHER B. Controlling popularity bias in learning-to-rank recommendation[C]//Proceedings of the Eleventh ACM Conference on Recommender Systems. Como: ACM, 2017: 42-46. |
6 | WANG X, HE X, WANG M, et al. Neural graph collaborative filtering[C]//Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. Paris: ACM, 2019: 165-174. |
7 | HE X, DENG K, WANG X, et al. LightGCN: Simplifying and powering graph convolution network for recommendation[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2020: 639-648. |
8 | 陈聪, 俞东进, 吴建华, 等. 基于隐式反馈数据的个性化游戏推荐[J]. 电子学报, 2018, 46(11): 2626-2632. |
CHEN C, YU D J, WU J H, et al. Personalized game recommendation based on implicit feedback[J]. Acta Electronica Sinica, 2018, 46(11): 2626-2632. (in Chinese) | |
9 | 郭景峰, 朱晓松, 李爽. 直播电视推荐系统的评分预测算法研究[J]. 电子学报, 2020, 48(9): 1735-1740. |
GUO J F, ZHU X S, LI S. Research on rating algorithm of live TV recommender systems[J]. Acta Electronica Sinica, 2020, 48(9): 1735-1740. (in Chinese) | |
10 | RENDLE S, FREUDENTHALER C, GANTNER Z, et al. BPR: Bayesian personalized ranking from implicit feedback[C]//Proceedings of the Twenty-Fifth Conference on Uncertainty in Artificial Intelligence. Montreal: AUAI Press, 2009: 452-461. |
11 | HJELM R, FEDOROV A, LAVOIE-MARCHILDON S, et al. Learning deep representations by mutual information estimation and maximization[C]//7th International Conference on Learning Representations. New Orleans: OpenReview, 2019: 233-256. |
12 | KONG L, D'AUTUME C, YU L, et al. A mutual information maximization perspective of language representation learning[C]//8th International Conference on Learning Representations. Addis Ababa: OpenReview, 2020: 1167-1178. |
13 | LINSKER R. Self-organization in a perceptual network[J]. IEEE Computer, 1988, 21(3): 105-117. |
14 | CARSE J, CAREY F, MCKENNA S. Unsupervised representation learning from pathology images with multi-directional contrastive predictive coding[C]//18th IEEE International Symposium on Biomedical Imaging. Nice: IEEE, 2021: 1254-1258. |
15 | ZHOU K, WANG H, ZHAO W, et al. S3-Rec: Self-supervised learning for sequential recommendation with mutual information maximization[C]//The 29th ACM International Conference on Information and Knowledge Management. Ireland: ACM, 2020: 1893-1902. |
16 | MA J, ZHOU C, YANG H, et al. Disentangled self-supervision in sequential recommenders[C]//The 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. CA: ACM, 2020: 483-491. |
17 | PARK Y J, TUZHILIN A. The long tail of recommender systems and how to leverage it[C]//Proceedings of the 2008 ACM Conference on Recommender Systems. Lausanne: ACM, 2008: 11-18. |
18 | ABDOLLAHPOURI H, MANSOURY M, BURKE R, et al. The unfairness of popularity bias in recommendation[C]//Proceedings of the Workshop on Recommendation in Multi-Stakeholder Environments Co-Located with the 13th ACM Conference on Recommender Systems, Copenhagen: CEUR-WS, 2019: 1257-1263. |
19 | PATRO G, BISWAS A, GANGULY N, et al. FairRec: Two-sided fairness for personalized recommendations in two-sided platforms[C]//Proceedings of The Web Conference 2020. Taipei: ACM, 2020: 1194-1204. |
20 | VARDASBI A, RIJKE M, MARKOV I. Cascade model-based propensity estimation for counterfactual learning to rank[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2020: 2089-2092. |
21 | SAITO Y, YAGINUMA S, NISHINO Y, et al. Unbiased recommender learning from missing-not-at-random implicit feedback[C]//The Thirteenth ACM International Conference on Web Search and Data Mining. Houston: ACM, 2020: 501-509. |
22 | SINHA A, GLEICH D, RAMANI K. Deconvolving feedback loops in recommender systems[C]//Advances in Neural Information Processing Systems 29: Annual Conference on Neural Information Processing Systems. Barcelona: ACM, 2017: 3243-3251. |
23 | ZHENG Y, GAO C, LI X, et al. Disentangling user interest and conformity for recommendation with causal conference embedding[C]//Proceedings of The Web 2021. Ljubljana: ACM, 2021: 2980-2991. |
24 | DEVLINJ, CHANGM, LEEK, etal. BERT: Pre-training of deepbi directional transformers for language[C]//Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics : Human Language. Minneapolis: ACL, 2019: 4171-4186. |
25 | VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Advances in Neural Information Processing Systems 30: Annual Conference on Neural Information Processing Systems. Long Beach: ACM, 2017: 5998-6008. |
26 | CHEN Z, XIAO R, LI C, et al. ESAM: Discriminative domain adaptation with non-displayed items to improve long-tail performance[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research and Development In Information Retrieval. New York: ACM, 2020: 579-588. |
[1] | 潘敏婷, 王韫博, 朱祥明, 高思宇, 龙明盛, 杨小康. 基于无标签视频数据的深度预测学习方法综述[J]. 电子学报, 2022, 50(4): 869-886. |
[2] | 郑孝遥, 罗永龙, 汪祥舜, 孙丽萍, 陈付龙, 胡桂银, 汪小寒. 基于位置服务的分布式差分隐私推荐方法研究[J]. 电子学报, 2021, 49(1): 99-110. |
[3] | 郭景峰, 朱晓松, 李爽. 直播电视推荐系统的评分预测算法研究[J]. 电子学报, 2020, 48(9): 1735-1740. |
[4] | 戚珉, 陈明. 标准模型下基于身份的混淆乐观公平交换方案[J]. 电子学报, 2020, 48(8): 1516-1527. |
[5] | 黄名选, 蒋曹清. 基于项权值排序挖掘的跨语言查询扩展[J]. 电子学报, 2020, 48(3): 568-576. |
[6] | 崔兆阳, 黄容兰, 万德焕. 一种有效改善系统性能的非正交传输方案[J]. 电子学报, 2020, 48(10): 1915-1922. |
[7] | 任开旭, 王玉龙, 刘同存, 李炜. 融合多维语义表示的概率矩阵分解模型[J]. 电子学报, 2019, 47(9): 1848-1854. |
[8] | 刘露, 胡封晔, 牛亮, 彭涛. 异质网络中基于节点影响力的相似度度量方法[J]. 电子学报, 2019, 47(9): 1929-1936. |
[9] | 肖云鹏, 孙华超, 戴天骥, 李茜, 李暾. 一种基于云模型的社交网络推荐系统评分预测方法[J]. 电子学报, 2018, 46(7): 1762-1767. |
[10] | 肖云鹏, 刘晏驰, 刘红, 刘媛妮. 一种基于动态角色标识和张量分解的推荐模型[J]. 电子学报, 2018, 46(3): 569-574. |
[11] | 鲜征征, 李启良, 黄晓宇, 陆寄远, 李磊. 融合显/隐式信任协同过滤算法的差分隐私保护[J]. 电子学报, 2018, 46(12): 3050-3059. |
[12] | 黄名选, 蒋曹清. 基于完全加权正负关联模式挖掘的越-英跨语言查询译后扩展[J]. 电子学报, 2018, 46(12): 3029-3036. |
[13] | 俞东进, 陈聪, 吴建华, 陈耀旺. 基于隐式反馈数据的个性化游戏推荐[J]. 电子学报, 2018, 46(11): 2626-2632. |
[14] | 范利云, 左万利, 王英, 王鑫. 一种基于差分隐私和时序的推荐系统模型研究[J]. 电子学报, 2017, 45(9): 2057-2064. |
[15] | 俞春花, 刘学军, 李斌, 章玮. 基于上下文相似度和社会网络的移动服务推荐方法[J]. 电子学报, 2017, 45(6): 1530-1536. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||