TY - GEN
T1 - Dual Sequential Network for Temporal Sets Prediction
AU - Sun, Leilei
AU - Bai, Yansong
AU - Du, Bowen
AU - Liu, Chuanren
AU - Xiong, Hui
AU - Lv, Weifeng
N1 - Funding Information:
This work is supported by the National Natural Science Foundation of China (51778033, 51822802, 51991395, 71901011, U1811463), the Science and Technology Major Project of Beijing (Z191100002519012) and the National Key R & D Program of China (2018YFB2101003).
Funding Information:
This work is supported by the National Natural Science Foundation of China (51778033, 51822802, 51991395, 71901011, U1811463), the Science and Technology Major Project of Beijing (Z191100002519012) and the National Key R &D Program of China (2018YFB2101003).
Publisher Copyright:
© 2020 ACM.
PY - 2020/7/25
Y1 - 2020/7/25
N2 - Many sequential behaviors such as purchasing items from time to time, selecting courses in different terms, collecting event logs periodically could be formalized as sequential sets of actions or elements, namely temporal sets. Predicting the subsequent set according to historical sequence of sets could help us make better producing, scheduling, or operating decisions. However, most of the existing methods were designed for predicting time series or temporal events, which could not be directly used for temporal sets prediction due to the difficulties of multi-level representations of items and sets, complex temporal dependencies of sets, and evolving dynamics of sequential behaviors. To address these issues, this paper provides a novel sets prediction method, called DSNTSP (Dual Sequential Network for Temporal Sets Prediction). Our model first learns both item-level representations and set-level representations of set sequences separately based on a transformer framework. Then, a co-transformer module is proposed to capture the multiple temporal dependencies of items and sets. Last, a gated neural module is designed to predict the subsequent set by fusing all the multi-level correlations and multiple temporal dependencies of items and sets. The experimental results on real-world data sets show that our methods lead to significant and consistent improvements as compared to other methods.
AB - Many sequential behaviors such as purchasing items from time to time, selecting courses in different terms, collecting event logs periodically could be formalized as sequential sets of actions or elements, namely temporal sets. Predicting the subsequent set according to historical sequence of sets could help us make better producing, scheduling, or operating decisions. However, most of the existing methods were designed for predicting time series or temporal events, which could not be directly used for temporal sets prediction due to the difficulties of multi-level representations of items and sets, complex temporal dependencies of sets, and evolving dynamics of sequential behaviors. To address these issues, this paper provides a novel sets prediction method, called DSNTSP (Dual Sequential Network for Temporal Sets Prediction). Our model first learns both item-level representations and set-level representations of set sequences separately based on a transformer framework. Then, a co-transformer module is proposed to capture the multiple temporal dependencies of items and sets. Last, a gated neural module is designed to predict the subsequent set by fusing all the multi-level correlations and multiple temporal dependencies of items and sets. The experimental results on real-world data sets show that our methods lead to significant and consistent improvements as compared to other methods.
KW - deep neural network
KW - set embedding
KW - temporal sets prediction
UR - http://www.scopus.com/inward/record.url?scp=85090150574&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85090150574&partnerID=8YFLogxK
U2 - 10.1145/3397271.3401124
DO - 10.1145/3397271.3401124
M3 - Conference contribution
AN - SCOPUS:85090150574
T3 - SIGIR 2020 - Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval
SP - 1439
EP - 1448
BT - SIGIR 2020 - Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval
PB - Association for Computing Machinery, Inc
T2 - 43rd Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2020
Y2 - 25 July 2020 through 30 July 2020
ER -