arXiv subjects

Depeng Jin

Publications and source records attributed to Depeng Jin.

At least 73 records · Page 4Linked to original sources

arXiv2021

Learnable Embedding Sizes for Recommender Systems

The embedding-based representation learning is commonly used in deep learning recommendation models to map the raw sparse features to dense vectors. The traditional embedding manner that assigns a uniform size to all features has two issues. First, the numerous features inevitably lead to a gigantic embedding table that causes a high memory usage cost. Second, it is likely to cause the over-fitting problem for those features that do not require too large representation capacity. Existing works that try to address the problem always cause a significant drop in recommendation performance or suffers from the limitation of unaffordable training time cost. In this paper, we proposed a novel approach, named PEP (short for Plug-in Embedding Pruning), to reduce the size of the embedding table while avoiding the drop of recommendation accuracy. PEP prunes embedding parameter where the pruning threshold(s) can be adaptively learned from data. Therefore we can automatically obtain a mixed-dimension embedding-scheme by pruning redundant parameters for each feature. PEP is a general framework that can plug in various base recommendation models. Extensive experiments demonstrate it can efficiently cut down embedding parameters and boost the base model's performance. Specifically, it achieves strong recommendation performance while reducing 97-99% parameters. As for the computation cost, PEP only brings an additional 20-30% time cost compared with base models. Codes are available at https://github.com/ssui-liu/learnable-embed-sizes-for-RecSys.

cs.LG↗

arXiv2021

Policy-Aware Mobility Model Explains the Growth of COVID-19 in Cities

With the continued spread of coronavirus, the task of forecasting distinctive COVID-19 growth curves in different cities, which remain inadequately explained by standard epidemiological models, is critical for medical supply and treatment. Predictions must take into account non-pharmaceutical interventions to slow the spread of coronavirus, including stay-at-home orders, social distancing, quarantine and compulsory mask-wearing, leading to reductions in intra-city mobility and viral transmission. Moreover, recent work associating coronavirus with human mobility and detailed movement data suggest the need to consider urban mobility in disease forecasts. Here we show that by incorporating intra-city mobility and policy adoption into a novel metapopulation SEIR model, we can accurately predict complex COVID-19 growth patterns in U.S. cities ($R^2$ = 0.990). Estimated mobility change due to policy interventions is consistent with empirical observation from Apple Mobility Trends Reports (Pearson's R = 0.872), suggesting the utility of model-based predictions where data are limited. Our model also reproduces urban "superspreading", where a few neighborhoods account for most secondary infections across urban space, arising from uneven neighborhood populations and heightened intra-city churn in popular neighborhoods. Therefore, our model can facilitate location-aware mobility reduction policy that more effectively mitigates disease transmission at similar social cost. Finally, we demonstrate our model can serve as a fine-grained analytic and simulation framework that informs the design of rational non-pharmaceutical interventions policies.

q-bio.PE↗

arXiv2021

Disentangling User Interest and Conformity for Recommendation with Causal Embedding

Recommendation models are usually trained on observational interaction data. However, observational interaction data could result from users' conformity towards popular items, which entangles users' real interest. Existing methods tracks this problem as eliminating popularity bias, e.g., by re-weighting training samples or leveraging a small fraction of unbiased data. However, the variety of user conformity is ignored by these approaches, and different causes of an interaction are bundled together as unified representations, hence robustness and interpretability are not guaranteed when underlying causes are changing. In this paper, we present DICE, a general framework that learns representations where interest and conformity are structurally disentangled, and various backbone recommendation models could be smoothly integrated. We assign users and items with separate embeddings for interest and conformity, and make each embedding capture only one cause by training with cause-specific data which is obtained according to the colliding effect of causal inference. Our proposed methodology outperforms state-of-the-art baselines with remarkable improvements on two real-world datasets on top of various backbone models. We further demonstrate that the learned embeddings successfully capture the desired causes, and show that DICE guarantees the robustness and interpretability of recommendation.

cs.IR↗

arXiv2020

Group-Buying Recommendation for Social E-Commerce

Group buying, as an emerging form of purchase in social e-commerce websites, such as Pinduoduo, has recently achieved great success. In this new business model, users, initiator, can launch a group and share products to their social networks, and when there are enough friends, participants, join it, the deal is clinched. Group-buying recommendation for social e-commerce, which recommends an item list when users want to launch a group, plays an important role in the group success ratio and sales. However, designing a personalized recommendation model for group buying is an entirely new problem that is seldom explored. In this work, we take the first step to approach the problem of group-buying recommendation for social e-commerce and develop a GBGCN method (short for Group-Buying Graph Convolutional Network). Considering there are multiple types of behaviors (launch and join) and structured social network data, we first propose to construct directed heterogeneous graphs to represent behavioral data and social networks. We then develop a graph convolutional network model with multi-view embedding propagation, which can extract the complicated high-order graph structure to learn the embeddings. Last, since a failed group-buying implies rich preferences of the initiator and participants, we design a double-pairwise loss function to distill such preference signals. We collect a real-world dataset of group-buying and conduct experiments to evaluate the performance. Empirical results demonstrate that our proposed GBGCN can significantly outperform baseline methods by 2.69%-7.36%. The codes and the dataset are released at https://github.com/Sweetnow/group-buying-recommendation.

cs.IR↗

arXiv2020

Disruption in the Chinese E-Commerce During COVID-19

The recent outbreak of the novel coronavirus (COVID-19) has infected millions of citizens worldwide and claimed many lives. This paper examines its impact on the Chinese e-commerce market by analyzing behavioral changes seen from a large online shopping platform. We first conduct a time series analysis to identify product categories that faced the most extensive disruptions. The time-lagged analysis shows that behavioral patterns seen in shopping actions are highly responsive to epidemic development. Based on these findings, we present a consumer demand prediction method by encompassing the epidemic statistics and behavioral features for COVID-19 related products. Experiment results demonstrate that our predictions outperform existing baselines and further extend to the long-term and province-level forecasts. We discuss how our market analysis and prediction can help better prepare for future pandemics by gaining an extra time to launch preventive steps.

physics.soc-ph↗

arXiv2020

Simplify and Robustify Negative Sampling for Implicit Collaborative Filtering

Negative sampling approaches are prevalent in implicit collaborative filtering for obtaining negative labels from massive unlabeled data. As two major concerns in negative sampling, efficiency and effectiveness are still not fully achieved by recent works that use complicate structures and overlook risk of false negative instances. In this paper, we first provide a novel understanding of negative instances by empirically observing that only a few instances are potentially important for model learning, and false negatives tend to have stable predictions over many training iterations. Above findings motivate us to simplify the model by sampling from designed memory that only stores a few important candidates and, more importantly, tackle the untouched false negative problem by favouring high-variance samples stored in memory, which achieves efficient sampling of true negatives with high-quality. Empirical results on two synthetic datasets and three real-world datasets demonstrate both robustness and superiorities of our negative sampling method.

cs.LG↗

arXiv2020

Bundle Recommendation with Graph Convolutional Networks

Bundle recommendation aims to recommend a bundle of items for a user to consume as a whole. Existing solutions integrate user-item interaction modeling into bundle recommendation by sharing model parameters or learning in a multi-task manner, which cannot explicitly model the affiliation between items and bundles, and fail to explore the decision-making when a user chooses bundles. In this work, we propose a graph neural network model named BGCN (short for \textit{\textBF{B}undle \textBF{G}raph \textBF{C}onvolutional \textBF{N}etwork}) for bundle recommendation. BGCN unifies user-item interaction, user-bundle interaction and bundle-item affiliation into a heterogeneous graph. With item nodes as the bridge, graph convolutional propagation between user and bundle nodes makes the learned representations capture the item level semantics. Through training based on hard-negative sampler, the user's fine-grained preferences for similar bundles are further distinguished. Empirical results on two real-world datasets demonstrate the strong performance gains of BGCN, which outperforms the state-of-the-art baselines by 10.77\% to 23.18\%.

cs.IR↗

arXiv2020

Price-aware Recommendation with Graph Convolutional Networks

In recent years, much research effort on recommendation has been devoted to mining user behaviors, i.e., collaborative filtering, along with the general information which describes users or items, e.g., textual attributes, categorical demographics, product images, and so on. Price, an important factor in marketing --- which determines whether a user will make the final purchase decision on an item --- surprisingly, has received relatively little scrutiny. In this work, we aim at developing an effective method to predict user purchase intention with the focus on the price factor in recommender systems. The main difficulties are two-fold: 1) the preference and sensitivity of a user on item price are unknown, which are only implicitly reflected in the items that the user has purchased, and 2) how the item price affects a user's intention depends largely on the product category, that is, the perception and affordability of a user on item price could vary significantly across categories. Towards the first difficulty, we propose to model the transitive relationship between user-to-item and item-to-price, taking the inspiration from the recently developed Graph Convolution Networks (GCN). The key idea is to propagate the influence of price on users with items as the bridge, so as to make the learned user representations be price-aware. For the second difficulty, we further integrate item categories into the propagation progress and model the possible pairwise interactions for predicting user-item interactions. We conduct extensive experiments on two real-world datasets, demonstrating the effectiveness of our GCN-based method in learning the price-aware preference of users. Further analysis reveals that modeling the price awareness is particularly useful for predicting user preference on items of unexplored categories.

cs.IR↗

arXiv2020

Learning to Recommend with Multiple Cascading Behaviors

Most existing recommender systems leverage user behavior data of one type only, such as the purchase behavior in E-commerce that is directly related to the business KPI (Key Performance Indicator) of conversion rate. Besides the key behavioral data, we argue that other forms of user behaviors also provide valuable signal, such as views, clicks, adding a product to shop carts and so on. They should be taken into account properly to provide quality recommendation for users. In this work, we contribute a new solution named NMTR (short for Neural Multi-Task Recommendation) for learning recommender systems from user multi-behavior data. We develop a neural network model to capture the complicated and multi-type interactions between users and items. 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). To fully exploit the signal in the data of multiple types of behaviors, we perform a joint optimization based on the multi-task learning framework, where the optimization on a behavior is treated as a task. Extensive experiments on two real-world datasets demonstrate that NMTR significantly outperforms state-of-the-art recommender systems that are designed to learn from both single-behavior data and multi-behavior data. Further analysis shows that modeling multiple behaviors is particularly useful for providing recommendation for sparse users that have very few interactions.

cs.IR↗

arXiv2019

UrbanRhythm: Revealing Urban Dynamics Hidden in Mobility Data

Understanding urban dynamics, i.e., how the types and intensity of urban residents' activities in the city change along with time, is of urgent demand for building an efficient and livable city. Nonetheless, this is challenging due to the expanding urban population and the complicated spatial distribution of residents. In this paper, to reveal urban dynamics, we propose a novel system UrbanRhythm to reveal the urban dynamics hidden in human mobility data. UrbanRhythm addresses three questions: 1) What mobility feature should be used to present residents' high-dimensional activities in the city? 2) What are basic components of urban dynamics? 3) What are the long-term periodicity and short-term regularity of urban dynamics? In UrbanRhythm, we extract staying, leaving, arriving three attributes of mobility and use a image processing method Saak transform to calculate the mobility distribution feature. For the second question, several city states are identified by hierarchy clustering as the basic components of urban dynamics, such as sleeping states and working states. We further characterize the urban dynamics as the transform of city states along time axis. For the third question, we directly observe the long-term periodicity of urban dynamics from visualization. Then for the short-term regularity, we design a novel motif analysis method to discovery motifs as well as their hierarchy relationships. We evaluate our proposed system on two real-life datesets and validate the results according to App usage records. This study sheds light on urban dynamics hidden in human mobility and can further pave the way for more complicated mobility behavior modeling and deeper urban understanding.

cs.SI↗

arXiv2018

Sampler Design for Bayesian Personalized Ranking by Leveraging View Data

Bayesian Personalized Ranking (BPR) is a representative pairwise learning method for optimizing recommendation models. It is widely known that the performance of BPR depends largely on the quality of negative sampler. In this paper, we make two contributions with respect to BPR. First, we find that sampling negative items from the whole space is unnecessary and may even degrade the performance. Second, focusing on the purchase feedback of E-commerce, we propose an effective sampler for BPR by leveraging the additional view data. In our proposed sampler, users' viewed interactions are considered as an intermediate feedback between those purchased and unobserved interactions. The pairwise rankings of user preference among these three types of interactions are jointly learned, and a user-oriented weighting strategy is considered during learning process, which is more effective and flexible. Compared to the vanilla BPR that applies a uniform sampler on all candidates, our view-enhanced sampler enhances BPR with a relative improvement over 37.03% and 16.40% on two real-world datasets. Our study demonstrates the importance of considering users' additional feedback when modeling their preference on different items, which avoids sampling negative items indiscriminately and inefficiently.

cs.IR↗

arXiv2017

Trajectory Recovery From Ash: User Privacy Is NOT Preserved in Aggregated Mobility Data

Human mobility data has been ubiquitously collected through cellular networks and mobile applications, and publicly released for academic research and commercial purposes for the last decade. Since releasing individual's mobility records usually gives rise to privacy issues, datasets owners tend to only publish aggregated mobility data, such as the number of users covered by a cellular tower at a specific timestamp, which is believed to be sufficient for preserving users' privacy. However, in this paper, we argue and prove that even publishing aggregated mobility data could lead to privacy breach in individuals' trajectories. We develop an attack system that is able to exploit the uniqueness and regularity of human mobility to recover individual's trajectories from the aggregated mobility data without any prior knowledge. By conducting experiments on two real-world datasets collected from both mobile application and cellular network, we reveal that the attack system is able to recover users' trajectories with accuracy about 73%~91% at the scale of tens of thousands to hundreds of thousands users, which indicates severe privacy leakage in such datasets. Through the investigation on aggregated mobility data, our work recognizes a novel privacy problem in publishing statistic data, which appeals for immediate attentions from both academy and industry.

cs.CY↗

arXiv2015

Iterative Eigenvalue Decomposition and Multipath-Grouping Tx/Rx Joint Beamforming for Millimeter-Wave Communication

We investigate Tx/Rx joint beamforming in millimeter-wave communications (MMWC). As the multipath components (MPCs) have different steering angles and independent fadings, beamforming aims at achieving array gain as well as diversity gain in this scenario. A sub-optimal beamforming scheme is proposed to find the antenna weight vectors (AWVs) at Tx/Rx via iterative eigenvalue decomposition (EVD), provided that full channel state information (CSI) is available at both the transmitter and receiver. To make this scheme practically feasible in MMWC, a corresponding training approach is suggested to avoid the channel estimation and iterative EVD computation. As in fast fading scenario the training approach may be time-consuming due to frequent training, another beamforming scheme, which exploits the quasi-static steering angles in MMWC, is proposed to reduce the overhead and increase the system reliability by multipath grouping (MPG). The scheme first groups the MPCs and then concurrently beamforms towards multiple steering angles of the grouped MPCs, so that both array gain and diversity gain are achieved. Performance comparisons show that, compared with the corresponding state-of-the-art schemes, the iterative EVD scheme with the training approach achieves the same performance with a reduced overhead and complexity, while the MPG scheme achieves better performance with an approximately equivalent complexity.

cs.IT↗

arXiv2015

Understanding Mobile Traffic Patterns of Large Scale Cellular Towers in Urban Environment

Understanding mobile traffic patterns of large scale cellular towers in urban environment is extremely valuable for Internet service providers, mobile users, and government managers of modern metropolis. This paper aims at extracting and modeling the traffic patterns of large scale towers deployed in a metropolitan city. To achieve this goal, we need to address several challenges, including lack of appropriate tools for processing large scale traffic measurement data, unknown traffic patterns, as well as handling complicated factors of urban ecology and human behaviors that affect traffic patterns. Our core contribution is a powerful model which combines three dimensional information (time, locations of towers, and traffic frequency spectrum) to extract and model the traffic patterns of thousands of cellular towers. Our empirical analysis reveals the following important observations. First, only five basic time-domain traffic patterns exist among the 9,600 cellular towers. Second, each of the extracted traffic pattern maps to one type of geographical locations related to urban ecology, including residential area, business district, transport, entertainment, and comprehensive area. Third, our frequency-domain traffic spectrum analysis suggests that the traffic of any tower among the 9,600 can be constructed using a linear combination of four primary components corresponding to human activity behaviors. We believe that the proposed traffic patterns extraction and modeling methodology, combined with the empirical analysis on the mobile traffic, pave the way toward a deep understanding of the traffic patterns of large scale cellular towers in modern metropolis.

cs.NI↗

arXiv2015

Exploiting Multi-Hop Relaying to Overcome Blockage in Directional mmWave Small Cells

With vast amounts of spectrum available in the millimeter wave (mmWave) band, small cells at mmWave frequencies densely deployed underlying the conventional homogeneous macrocell network have gained considerable interest from academia, industry, and standards bodies. Due to high propagation loss at higher frequencies, mmWave communications are inherently directional, and concurrent transmissions (spatial reuse) under low inter-link interference can be enabled to significantly improve network capacity. On the other hand, mmWave links are easily blocked by obstacles such as human body and furniture. In this paper, we develop a Multi-Hop Relaying Transmission scheme, termed as MHRT, to steer blocked flows around obstacles by establishing multi-hop relay paths. InMHRT, a relay path selection algorithmis proposed to establish relay paths for blocked flows for better use of concurrent transmissions. After relay path selection, we use a multi-hop transmission scheduling algorithm to compute near-optimal schedules by fully exploiting the spatial reuse. Through extensive simulations under various traffic patterns and channel conditions, we demonstrate MHRT achieves superior performance in terms of network throughput and connection robustness compared with other existing protocols, especially under serious blockage conditions. The performance of MHRT with different hop limitations is also simulated and analyzed for a better choice of themaximum hop number in practice.

cs.NI↗

arXiv2015

Energy Efficient Scheduling for mmWave Backhauling of Small Cells in Heterogeneous Cellular Networks

Heterogeneous cellular networks with small cells densely deployed underlying the conventional homogeneous macrocells are emerging as a promising candidate for the fifth generation (5G) mobile network. When a large number of base stations are deployed, the cost-effective, flexible, and green backhaul solution becomes one of the most urgent and critical challenges. With vast amounts of spectrum available, wireless backhaul in the millimeter wave (mmWave) band is able to provide several-Gbps transmission rates. To overcome high propagation loss at higher frequencies, mmWave backhaul utilize beamforming to achieve directional transmission, and concurrent transmissions (spatial reuse) under low inter-link interference can be enabled to significantly improve network capacity. To achieve an energy efficient solution for the mmWave backhauling of small cells, we first formulate the problem of minimizing the energy consumption via concurrent transmission scheduling and power control into a mixed integer nonlinear programming problem. Then we develop an energy efficient and practical mmWave backhauling scheme, where the maximum independent set based scheduling algorithm and the power control algorithm are proposed to exploit the spatial reuse for low energy consumption and high energy efficiency. We also theoretically analyze the conditions that our scheme reduces energy consumption, and the choice of the interference threshold for energy reduction. Through extensive simulations under various traffic patterns and system parameters, we demonstrate the superior performance of our scheme in terms of energy consumption and energy efficiency, and also analyze the choice of the interference threshold under different traffic loads, BS distributions, and the maximum transmission power.

cs.NI↗

arXiv2015

Exploiting Device-to-Device Communications to Enhance Spatial Reuse for Popular Content Downloading in Directional mmWave Small Cells

With the explosive growth of mobile demand, small cells in millimeter wave (mmWave) bands underlying the macrocell networks have attracted intense interest from both academia and industry. MmWave communications in the 60 GHz band are able to utilize the huge unlicensed bandwidth to provide multiple Gbps transmission rates. In this case, device-to-device (D2D) communications in mmWave bands should be fully exploited due to no interference with the macrocell networks and higher achievable transmission rates. In addition, due to less interference by directional transmission, multiple links including D2D links can be scheduled for concurrent transmissions (spatial reuse). With the popularity of content-based mobile applications, popular content downloading in the small cells needs to be optimized to improve network performance and enhance user experience. In this paper, we develop an efficient scheduling scheme for popular content downloading in mmWave small cells, termed PCDS (popular content downloading scheduling), where both D2D communications in close proximity and concurrent transmissions are exploited to improve transmission efficiency. In PCDS, a transmission path selection algorithm is designed to establish multi-hop transmission paths for users, aiming at better utilization of D2D communications and concurrent transmissions. After transmission path selection, a concurrent transmission scheduling algorithm is designed to maximize the spatial reuse gain. Through extensive simulations under various traffic patterns, we demonstrate PCDS achieves near-optimal performance in terms of delay and throughput, and also superior performance compared with other existing protocols, especially under heavy load.

cs.NI↗

arXiv2015

Boosting Spatial Reuse via Multiple Paths Multi-Hop Scheduling for Directional mmWave WPANs

With huge unlicensed bandwidth available in most parts of the world, millimeter wave (mmWave) communications in the 60 GHz band has been considered as one of the most promising candidates to support multi-gigabit wireless services. Due to high propagation loss of mmWave channels, beamforming is likely to become adopted as an essential technique. Consequently, transmission in 60 GHz band is inherently directional. Directivity enables concurrent transmissions (spatial reuse), which can be fully exploited to improve network capacity. In this paper, we propose a multiple paths multi-hop scheduling scheme, termed MPMH, for mmWave wireless personal area networks, where the traffic across links of low channel quality is transmitted through multiple paths of multiple hops to unleash the potential of spatial reuse. We formulate the problem of multiple paths multi-hop scheduling as a mixed integer linear program (MILP), which is generally NP-hard. To enable the implementation of the multiple paths multi-hop transmission in practice, we propose a heuristic scheme including path selection, traffic distribution, and multiple paths multi-hop scheduling to efficiently solve the formulated problem. Finally, through extensive simulations, we demonstrate MPMH achieves near-optimal network performance in terms of transmission delay and throughput, and enhances the network performance significantly compared with existing protocols.

cs.NI↗