Filters








252 Hits in 2.5 sec

Large Scale Purchase Prediction with Historical User Actions on B2C Online Retail Platform [article]

Yuyu Zhang, Liang Pang, Lei Shi, Bin Wang
2015 arXiv   pre-print
With real-world user action data provided by Tmall, one of the largest B2C online retail platforms in China, this competition requires to predict future user purchases on Tmall website.  ...  We model the purchase prediction problem as standard machine learning problem, and mainly employ regression and classification methods as single models.  ...  In training, we use the last month as the target span and previous months as the feature span. In predicting, we use all available months as the feature span.  ... 
arXiv:1408.6515v3 fatcat:d5au2cpirjdmxb232mvhnmoaya

Improving Implicit Recommender Systems with View Data

Jingtao Ding, Guanghui Yu, Xiangnan He, Yuhan Quan, Yong Li, Tat-Seng Chua, Depeng Jin, Jiajie Yu
2018 Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence  
Most existing recommender systems leverage the primary feedback data only, such as the purchase records in E-commerce.  ...  We propose to model the pairwise ranking relations among purchased, viewed, and non-viewed interactions, being more effective and flexible than typical pointwise matrix factorization (MF) methods.  ...  For a specific user u, vector p u ∈ R K denotes the K-dimensional latent feature vector, and set R u denotes the set of items that are interacted by u.  ... 
doi:10.24963/ijcai.2018/464 dblp:conf/ijcai/DingY0QLCJY18 fatcat:gywcxc6mijfk7ehiz4w37iqwoy

Learning Recommender Systems from Multi-Behavior Data [article]

Chen Gao, Xiangnan He, Dahua Gan, Xiangning Chen, Fuli Feng, Yong Li, Tat-Seng Chua, Depeng Jin
2018 arXiv   pre-print
Further analysis shows that modeling multiple behaviors is particularly useful for providing recommendation for sparse users that have very few interactions.  ...  In particular, our model accounts for the cascading relationship among different types of behaviors (e.g., a user must click on a product before purchasing it).  ...  Specifically, the performance of (Purchase, 50% Carting) is worse than only using purchase, while (Purchase, 50% Viewing) is better than only using purchase. There are two major reasons.  ... 
arXiv:1809.08161v3 fatcat:6qje73dawnhfvejda2j6bh3hjm

Research on Identification Method of Anonymous Fake Reviews in E-commerce

Lizhen Liu, Xinlei Zhao, Hanshi Wang, Wei Song, Chao Du
2016 TELKOMNIKA (Telecommunication Computing Electronics and Control)  
The extracted features are also useful to identifying the fake reviews when the reviewer's identification is not accessable.  ...  The proposed method takes into account these 5 features to calculate the fake reviews content by constructing multivariate linear regression model, Experiments show that this prelimilnary work performed  ...  Reference [6] , the dataset is also from Taobao and Tmall. The click farmers' userID is obtained, tracking purchasing informationfor analysis and getting 14 features.  ... 
doi:10.12928/telkomnika.v14i4.3654 fatcat:thj4m2t4inbnbai3zfgm4spleu

Attentive Sequential Models of Latent Intent for Next Item Recommendation

Md Mehrab Tanjim, Congzhe Su, Ethan Benjamin, Diane Hu, Liangjie Hong, Julian McAuley
2020 Proceedings of The Web Conference 2020  
We use this representation to guide an attentive model to predict the next item.  ...  , add-to-favorites, purchase).  ...  Later, to make sure this latent feature captures intent, we use it for dual prediction of both the next interaction (by a feedforward network) and the next item (by a co-attentional transformer layer).  ... 
doi:10.1145/3366423.3380002 dblp:conf/www/TanjimSBHHM20 fatcat:ywr7ktpgcrblhnri46mtbfr7xi

Interest-Behaviour Multiplicative Network for Resource-limited Recommendation [article]

Qianliang Wu and Tong Zhang and Zhen Cui and Jian Yang
2020 arXiv   pre-print
Finally, mutual information is introduced to measure the similarity between the user action and fused features to predict future interaction, where the fused features come from both MRRNNs and resource-limited  ...  We test the performance on the built used car transaction dataset as well as the Tmall dataset, and the experimental results verify the effectiveness of our framework.  ...  We treat purchase actions as 'purchase' interactions while the other two kinds of actions as 'click'. We obtain the image feature using vgg16 features while the structure feature using id embedding.  ... 
arXiv:2009.13249v4 fatcat:fyisfwr665bi5k7oips335qykm

Sequential Recommender System based on Hierarchical Attention Networks

Haochao Ying, Fuzhen Zhuang, Fuzheng Zhang, Yanchi Liu, Guandong Xu, Xing Xie, Hui Xiong, Jian Wu
2018 Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence  
Specifically, the first attention layer learns user long-term preferences based on the historical purchased item representation, while the second one outputs final user representation through coupling  ...  Moreover, they have different influences on the next item to be purchased by using the different learned weights.  ...  More specifically, the relative performance improvement by HRM is 6.7% and 4.5% in terms of Recall@50 on Tmall and Gowalla datasets, respectively.  ... 
doi:10.24963/ijcai.2018/546 dblp:conf/ijcai/YingZZLXXX018 fatcat:pzi6knhz4nfazd7o3igltcbpcy

Cascading Residual Graph Convolutional Network for Multi-Behavior Recommendation [article]

Mingshi Yan, Zhiyong Cheng, Chen Gao, Jing Sun, Fan Liu, Fuming Sun, Haojie Li
2022 arXiv   pre-print
Most existing multi-behavior recommendation methods take the strategy to first extract information from different behaviors separately and then fuse them for final prediction.  ...  Specifically, the next behavior provides more specific information which can help us refine user's preference.  ...  On The Tmall platform, users can buy the item directly after page viewing, or add it to the cart before purchasing, or they may just click on the collection instead of the purchase behavior. • Beibei.  ... 
arXiv:2205.13128v1 fatcat:dioenuzmwrapjna25teukt6rfe

Non-Compensatory Psychological Models for Recommender Systems

Chen Lin, Xiaolin Shen, Si Chen, Muhua Zhu, Yanghua Xiao
2019 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
We show how this conceptual model can be realized in both pointwise rating prediction models and pair-wise ranking prediction models.  ...  models which are based on latent factor models assume the consumers follow the compensatory rules, i.e. they evaluate an item over multiple aspects and compute a weighted or/and summated score which is used  ...  Tmall-hybrid: the pairwise rankings are built by extracting purchased items in each session and all remaining items which are not purchased in the same session.  ... 
doi:10.1609/aaai.v33i01.33014304 fatcat:6n6wwkrdmbbspp4dcfovysxljq

Multi-behavioral Sequential Prediction with Recurrent Log-bilinear Model [article]

Qiang Liu, Shu Wu, Liang Wang
2017 arXiv   pre-print
Given the behavioral history of a specific user, predicting his or her next choice plays a key role in improving various online services.  ...  Moreover, considering continuous time difference in behavioral history is a key factor for dynamic prediction, we further extend RLBL and replace position-specific transition matrices with time-specific  ...  On this dataset, we aim to predict what users will purchase next.  ... 
arXiv:1608.07102v4 fatcat:ob6wlo4nazeuzbpuqnb6b6skcy

A novel approach to dynamic profiling of e-customers considering click stream data and online reviews

Houda Zaim, Adil Haddi, Mohammed Ramdani
2019 International Journal of Electrical and Computer Engineering (IJECE)  
The extracted values of the website's features are also useful to identifying the satisfaction level when the customer's rate is not available.</p><p> </p>  ...  Experiments show that this work performed well in identifying relevant customer's stream data to judge the chinese e-commerce website "Tmall".  ...  [4] integrated classifier to predict the type of purchase that a customer would make, as well as the number of visits that he/she would make during a year.  ... 
doi:10.11591/ijece.v9i1.pp602-612 fatcat:35vghcg6q5aafjvlvj45qt376m

Using graded implicit feedback for bayesian personalized ranking

Lukas Lerche, Dietmar Jannach
2014 Proceedings of the 8th ACM Conference on Recommender systems - RecSys '14  
., different types of user actions like item views, cart or purchase actions and there can exist several actions for an item over time.  ...  An empirical analysis shows that this extension can help to measurably increase the predictive accuracy of BPR on realistic e-commerce datasets.  ...  Meta-data features or social network information could for example be used as done, e.g., in [5] or [7] .  ... 
doi:10.1145/2645710.2645759 dblp:conf/recsys/LercheJ14 fatcat:kekhausw7zhtrinrcjir7kb6n4

Time-aware conversion prediction

Wendi Ji, Xiaoling Wang, Feida Zhu
2016 Frontiers of Computer Science  
The challenging question: what products should be recommended for a given time period to maximize conversion-is what has motivated us in this paper to propose a rank-based time-aware conversion prediction  ...  The "conversion" here refers to actually a more general set of pre-defined actions, including for example purchases or registrations in recommendation and advertising systems.  ...  ) with cWMM has been proposed to estimate behaviors over a specific prediction period. 4) The proposed cWMM has been evaluated using two real-world datasets.  ... 
doi:10.1007/s11704-016-5546-y fatcat:3gysgwo5vzfa7j6vtgygwtg6ku

Arbitrary Talking Face Generation via Attentional Audio-Visual Coherence Learning

Hao Zhu, Huaibo Huang, Yi Li, Aihua Zheng, Ran He
2020 Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence  
For instance, purchasing a basket of products is a common user behaviour in our daily life but is hard to precisely predict.  ...  Experiments and Evaluation Data Preparation Two real-world transaction datasets commonly used to test the performance of next-basket prediction [Guidotti and et al., 2018; Le et al., 2019] are used  ... 
doi:10.24963/ijcai.2020/323 dblp:conf/ijcai/WangHWSOC20 fatcat:h3ks7i7oovbebk75b45af3smqy

Collaborative Metric Learning with Memory Network for Multi-Relational Recommender Systems [article]

Xiao Zhou, Danyang Liu, Jianxun Lian, Xing Xie
2019 arXiv   pre-print
In everyday life, large amounts of user feedback data are created along with user-item online interactions in a variety of ways, such as browsing, purchasing, and sharing.  ...  These multiple types of user feedback provide us with tremendous opportunities to detect individuals' fine-grained preferences.  ...  employed memory networks for e-commerce recommendations using historical purchase records of users.  ... 
arXiv:1906.09882v1 fatcat:fld575kolvatpdiregabghkslq
« Previous Showing results 1 — 15 out of 252 results