arXiv subjects

Liang Wang

Publications and source records attributed to Liang Wang.

At least 559 records · Page 31Linked to original sources

arXiv2020

Fi-GNN: Modeling Feature Interactions via Graph Neural Networks for CTR Prediction

Click-through rate (CTR) prediction is an essential task in web applications such as online advertising and recommender systems, whose features are usually in multi-field form. The key of this task is to model feature interactions among different feature fields. Recently proposed deep learning based models follow a general paradigm: raw sparse input multi-filed features are first mapped into dense field embedding vectors, and then simply concatenated together to feed into deep neural networks (DNN) or other specifically designed networks to learn high-order feature interactions. However, the simple \emph{unstructured combination} of feature fields will inevitably limit the capability to model sophisticated interactions among different fields in a sufficiently flexible and explicit fashion. In this work, we propose to represent the multi-field features in a graph structure intuitively, where each node corresponds to a feature field and different fields can interact through edges. The task of modeling feature interactions can be thus converted to modeling node interactions on the corresponding graph. To this end, we design a novel model Feature Interaction Graph Neural Networks (Fi-GNN). Taking advantage of the strong representative power of graphs, our proposed model can not only model sophisticated feature interactions in a flexible and explicit fashion, but also provide good model explanations for CTR prediction. Experimental results on two real-world datasets show its superiority over the state-of-the-arts.

cs.IR↗

arXiv2020

Deep Graph Contrastive Representation Learning

Graph representation learning nowadays becomes fundamental in analyzing graph-structured data. Inspired by recent success of contrastive methods, in this paper, we propose a novel framework for unsupervised graph representation learning by leveraging a contrastive objective at the node level. Specifically, we generate two graph views by corruption and learn node representations by maximizing the agreement of node representations in these two views. To provide diverse node contexts for the contrastive objective, we propose a hybrid scheme for generating graph views on both structure and attribute levels. Besides, we provide theoretical justification behind our motivation from two perspectives, mutual information and the classical triplet loss. We perform empirical experiments on both transductive and inductive learning tasks using a variety of real-world datasets. Experimental experiments demonstrate that despite its simplicity, our proposed method consistently outperforms existing state-of-the-art methods by large margins. Moreover, our unsupervised method even surpasses its supervised counterparts on transductive tasks, demonstrating its great potential in real-world applications.

cs.LG↗

arXiv2020

Adversarial Self-Supervised Learning for Semi-Supervised 3D Action Recognition

We consider the problem of semi-supervised 3D action recognition which has been rarely explored before. Its major challenge lies in how to effectively learn motion representations from unlabeled data. Self-supervised learning (SSL) has been proved very effective at learning representations from unlabeled data in the image domain. However, few effective self-supervised approaches exist for 3D action recognition, and directly applying SSL for semi-supervised learning suffers from misalignment of representations learned from SSL and supervised learning tasks. To address these issues, we present Adversarial Self-Supervised Learning (ASSL), a novel framework that tightly couples SSL and the semi-supervised scheme via neighbor relation exploration and adversarial learning. Specifically, we design an effective SSL scheme to improve the discrimination capability of learned representations for 3D action recognition, through exploring the data relations within a neighborhood. We further propose an adversarial regularization to align the feature distributions of labeled and unlabeled samples. To demonstrate effectiveness of the proposed ASSL in semi-supervised 3D action recognition, we conduct extensive experiments on NTU and N-UCLA datasets. The results confirm its advantageous performance over state-of-the-art semi-supervised methods in the few label regime for 3D action recognition.

cs.CV↗

arXiv2020

Multi-future Merchant Transaction Prediction

The multivariate time series generated from merchant transaction history can provide critical insights for payment processing companies. The capability of predicting merchants' future is crucial for fraud detection and recommendation systems. Conventionally, this problem is formulated to predict one multivariate time series under the multi-horizon setting. However, real-world applications often require more than one future trend prediction considering the uncertainties, where more than one multivariate time series needs to be predicted. This problem is called multi-future prediction. In this work, we combine the two research directions and propose to study this new problem: multi-future, multi-horizon and multivariate time series prediction. This problem is crucial as it has broad use cases in the financial industry to reduce the risk while improving user experience by providing alternative futures. This problem is also challenging as now we not only need to capture the patterns and insights from the past but also train a model that has a strong inference capability to project multiple possible outcomes. To solve this problem, we propose a new model using convolutional neural networks and a simple yet effective encoder-decoder structure to learn the time series pattern from multiple perspectives. We use experiments on real-world merchant transaction data to demonstrate the effectiveness of our proposed model. We also provide extensive discussions on different model design choices in our experimental section.

cs.LG↗

arXiv2020

On Learning Semantic Representations for Million-Scale Free-Hand Sketches

In this paper, we study learning semantic representations for million-scale free-hand sketches. This is highly challenging due to the domain-unique traits of sketches, e.g., diverse, sparse, abstract, noisy. We propose a dual-branch CNNRNN network architecture to represent sketches, which simultaneously encodes both the static and temporal patterns of sketch strokes. Based on this architecture, we further explore learning the sketch-oriented semantic representations in two challenging yet practical settings, i.e., hashing retrieval and zero-shot recognition on million-scale sketches. Specifically, we use our dual-branch architecture as a universal representation framework to design two sketch-specific deep models: (i) We propose a deep hashing model for sketch retrieval, where a novel hashing loss is specifically designed to accommodate both the abstract and messy traits of sketches. (ii) We propose a deep embedding model for sketch zero-shot recognition, via collecting a large-scale edge-map dataset and proposing to extract a set of semantic vectors from edge-maps as the semantic knowledge for sketch zero-shot domain alignment. Both deep models are evaluated by comprehensive experiments on million-scale sketches and outperform the state-of-the-art competitors.

cs.CV↗

arXiv2020

TFNet: Multi-Semantic Feature Interaction for CTR Prediction

The CTR (Click-Through Rate) prediction plays a central role in the domain of computational advertising and recommender systems. There exists several kinds of methods proposed in this field, such as Logistic Regression (LR), Factorization Machines (FM) and deep learning based methods like Wide&Deep, Neural Factorization Machines (NFM) and DeepFM. However, such approaches generally use the vector-product of each pair of features, which have ignored the different semantic spaces of the feature interactions. In this paper, we propose a novel Tensor-based Feature interaction Network (TFNet) model, which introduces an operating tensor to elaborate feature interactions via multi-slice matrices in multiple semantic spaces. Extensive offline and online experiments show that TFNet: 1) outperforms the competitive compared methods on the typical Criteo and Avazu datasets; 2) achieves large improvement of revenue and click rate in online A/B tests in the largest Chinese App recommender system, Tencent MyApp.

cs.IR↗

arXiv2020

Programmable In-Network Obfuscation of Traffic

Recent advances in programmable switch hardware offer a fresh opportunity to protect user privacy. This paper presents PINOT, a lightweight in-network anonymity solution that runs at line rate within the memory and processing constraints of hardware switches. PINOT encrypts a client's IPv4 address with an efficient encryption scheme to hide the address from downstream ASes and the destination server. PINOT is readily deployable, requiring no end-user software or cooperation from networks other than the trusted network where it runs. We implement a PINOT prototype on the Barefoot Tofino switch, deploy PINOT in a campus network, and present results on protecting user identity against public DNS, NTP, and WireGuard VPN services.

cs.NI↗

arXiv2020

Investigating Label Bias in Beam Search for Open-ended Text Generation

Beam search is an effective and widely used decoding algorithm in many sequence-to-sequence (seq2seq) text generation tasks. However, in open-ended text generation, beam search is often found to produce repetitive and generic texts, sampling-based decoding algorithms like top-k sampling and nucleus sampling are more preferred. Standard seq2seq models suffer from label bias due to its locally normalized probability formulation. This paper provides a series of empirical evidence that label bias is a major reason for such degenerate behaviors of beam search. By combining locally normalized maximum likelihood estimation and globally normalized sequence-level training, label bias can be reduced with almost no sacrifice in perplexity. To quantitatively measure label bias, we test the model's ability to discriminate the groundtruth text and a set of context-agnostic distractors. We conduct experiments on large-scale response generation datasets. Results show that beam search can produce more diverse and meaningful texts with our approach, in terms of both automatic and human evaluation metrics. Our analysis also suggests several future working directions towards the grand challenge of open-ended text generation.

cs.CL↗

arXiv2020

The backbone-residual model. Accurately characterising the instrumental profile of a fibre-fed echelle spectrograph

Context: Instrumental profile (IP) is the basic property of a spectrograph. Accurate IP characterisation is the prerequisite of accurate wavelength solution. It also facilitates new spectral acquisition methods such as the forward modeling and deconvolution. Aims: We investigate an IP modeling method for the fibre-fed echelle spectrograph with the emission lines of the ThAr lamp, and explore the method to evaluate the accuracy of IP characterisation. Methods: The backbone-residual (BR) model is put forward and tested on the fibre-fed High Resolution Spectrograph (HRS) at the Chinese Xinglong 2.16-m Telescope, which is the sum of the backbone function and the residual function. The backbone function is a bell-shaped function to describe the main component and the spatial variation of IP. The residual function, which is expressed as the cubic spline function, accounts for the difference between the bell-shaped function and the actual IP. The method of evaluating the accuracy of IP characterisation is based on the spectral reconstruction and Monte Carlo simulation. Results: The IP of HRS is characterised with the BR model, and the accuracy of the characterised IP reaches 0.006 of the peak value of the backbone function. This result demonstrates that the accurate IP characterisation has been achieved on HRS with the BR model, and the BR model is an excellent choice for accurate IP characterisation of fibre-fed echelle spectrographs.

astro-ph.IM↗

arXiv2020

Every Document Owns Its Structure: Inductive Text Classification via Graph Neural Networks

Text classification is fundamental in natural language processing (NLP), and Graph Neural Networks (GNN) are recently applied in this task. However, the existing graph-based works can neither capture the contextual word relationships within each document nor fulfil the inductive learning of new words. In this work, to overcome such problems, we propose TextING for inductive text classification via GNN. We first build individual graphs for each document and then use GNN to learn the fine-grained word representations based on their local structures, which can also effectively produce embeddings for unseen words in the new document. Finally, the word nodes are aggregated as the document embedding. Extensive experiments on four benchmark datasets show that our method outperforms state-of-the-art text classification methods.

cs.CL↗

arXiv2020

TAGNN: Target Attentive Graph Neural Networks for Session-based Recommendation

Session-based recommendation nowadays plays a vital role in many websites, which aims to predict users' actions based on anonymous sessions. There have emerged many studies that model a session as a sequence or a graph via investigating temporal transitions of items in a session. However, these methods compress a session into one fixed representation vector without considering the target items to be predicted. The fixed vector will restrict the representation ability of the recommender model, considering the diversity of target items and users' interests. In this paper, we propose a novel target attentive graph neural network (TAGNN) model for session-based recommendation. In TAGNN, target-aware attention adaptively activates different user interests with respect to varied target items. The learned interest representation vector varies with different target items, greatly improving the expressiveness of the model. Moreover, TAGNN harnesses the power of graph neural networks to capture rich item transitions in sessions. Comprehensive experiments conducted on real-world datasets demonstrate its superiority over state-of-the-art methods.

cs.IR↗

arXiv2020

Stochastic SIR Levy jump model with heavy tailed increments

This paper considers a general stochastic SIR epidemic model driven by a multidimensional Levy jump process with heavy tailed increments and possible correlation between noise components. In this framework, we derive new sufficient conditions for disease extinction and persistence in the mean. Our method differs from previous approaches by the use of Kunita's inequality instead of the Burkholder-Davis-Gundy inequality for continuous processes, and allows for the treatment of infinite Levy measures by the definition of new threshold values. An SIR model driven by a tempered stable process is presented as an example of application with the ability to model sudden disease outbreak, illustrated by numerical simulations. The results show that persistence and extinction are dependent not only on the variance of the processes increments, but also on the shapes of their distributions.

math.PR↗

arXiv2020

Toward a Wearable RFID System for Real-Time Activity Recognition Using Radio Patterns

Elderly care is one of the many applications supported by real-time activity recognition systems. Traditional approaches use cameras, body sensor networks, or radio patterns from various sources for activity recognition. However, these approaches are limited due to ease-of-use, coverage, or privacy preserving issues. In this paper, we present a novel wearable Radio Frequency Identification (RFID) system aims at providing an easy-to-use solution with high detection coverage. Our system uses passive tags which are maintenance-free and can be embedded into the clothes to reduce the wearing and maintenance efforts. A small RFID reader is also worn on the user's body to extend the detection coverage as the user moves. We exploit RFID radio patterns and extract both spatial and temporal features to characterize various activities. We also address the issues of false negative of tag readings and tag/antenna calibration, and design a fast online recognition system. Antenna and tag selection is done automatically to explore the minimum number of devices required to achieve target accuracy. We develop a prototype system which consists of a wearable RFID system and a smartphone to demonstrate the working principles, and conduct experimental studies with four subjects over two weeks. The results show that our system achieves a high recognition accuracy of 93.6 percent with a latency of 5 seconds. Additionally, we show that the system only requires two antennas and four tagged body parts to achieve a high recognition accuracy of 85 percent.

cs.OH↗

arXiv2020

Deep Self-Supervised Representation Learning for Free-Hand Sketch

In this paper, we tackle for the first time, the problem of self-supervised representation learning for free-hand sketches. This importantly addresses a common problem faced by the sketch community -- that annotated supervisory data are difficult to obtain. This problem is very challenging in that sketches are highly abstract and subject to different drawing styles, making existing solutions tailored for photos unsuitable. Key for the success of our self-supervised learning paradigm lies with our sketch-specific designs: (i) we propose a set of pretext tasks specifically designed for sketches that mimic different drawing styles, and (ii) we further exploit the use of a textual convolution network (TCN) in a dual-branch architecture for sketch feature learning, as means to accommodate the sequential stroke nature of sketches. We demonstrate the superiority of our sketch-specific designs through two sketch-related applications (retrieval and recognition) on a million-scale sketch dataset, and show that the proposed approach outperforms the state-of-the-art unsupervised representation learning methods, and significantly narrows the performance gap between with supervised representation learning.

cs.CV↗

arXiv2019

Pose-Guided Multi-Granularity Attention Network for Text-Based Person Search

Text-based person search aims to retrieve the corresponding person images in an image database by virtue of a describing sentence about the person, which poses great potential for various applications such as video surveillance. Extracting visual contents corresponding to the human description is the key to this cross-modal matching problem. Moreover, correlated images and descriptions involve different granularities of semantic relevance, which is usually ignored in previous methods. To exploit the multilevel corresponding visual contents, we propose a pose-guided multi-granularity attention network (PMA). Firstly, we propose a coarse alignment network (CA) to select the related image regions to the global description by a similarity-based attention. To further capture the phrase-related visual body part, a fine-grained alignment network (FA) is proposed, which employs pose information to learn latent semantic alignment between visual body part and textual noun phrase. To verify the effectiveness of our model, we perform extensive experiments on the CUHK Person Description Dataset (CUHK-PEDES) which is currently the only available dataset for text-based person search. Experimental results show that our approach outperforms the state-of-the-art methods by 15 \% in terms of the top-1 metric.

cs.CV↗

arXiv2019

The Pan-Pacific Planet Search. VIII. Complete results and the occurrence rate of planets around low-luminosity giants

Our knowledge of the populations and occurrence rates of planets orbiting evolved intermediate-mass stars lags behind that for solar-type stars by at least a decade. Some radial velocity surveys have targeted these low-luminosity giant stars, providing some insights into the properties of their planetary systems. Here we present the final data release of the Pan-Pacific Planet Search, a 5-year radial velocity survey using the 3.9m Anglo-Australian Telescope. We present 1293 precise radial velocity measurements for 129 stars, and highlight six potential substellar-mass companions which require additional observations to confirm. Correcting for the substantial incompleteness in the sample, we estimate the occurrence rate of giant planets orbiting low-luminosity giant stars to be approximately 7.8$^{+9.1}_{-3.3}$\%. This result is consistent with the frequency of such planets found to orbit main-sequence A-type stars, from which the PPPS stars have evolved.

astro-ph.EP↗

arXiv2019

Learning Preferences and Demands in Visual Recommendation

Visual information is an important factor in recommender systems, in which users' selections consist of two components: \emph{preferences} and \emph{demands}. Some studies has been done for modeling users' preferences in visual recommendation. However, conventional methods models items in a common visual feature space, which may fail in capturing \emph{styles} of items. We propose a DeepStyle method for learning style features of items. DeepStyle eliminates the categorical information of items, which is dominant in the original visual feature space, based on a Convolutional Neural Networks (CNN) architecture. For modeling users' demands on different categories of items, the problem can be formulated as recommendation with contextual and sequential information. To solve this problem, we propose a Context-Aware Gated Recurrent Unit (CA-GRU) method, which can capture sequential and contextual information simultaneously. Furthermore, the aggregation of prediction on preferences and demands, i.e., prediction generated by DeepStyle and CA-GRU, can model users' selection behaviors more completely. Experiments conducted on real-world datasets illustrates the effectiveness of our proposed methods in visual recommendation.

cs.IR↗

arXiv2019

Exact and Locally Implicit Source Term Solvers for Multifluid-Maxwell Systems

Recently, a family of models that couple multifluid systems to the full Maxwell equations draw a lot of attention in laboratory, space, and astrophysical plasma modeling. These models are more complete descriptions of the plasma than reduced models like magnetohydrodynamic (MHD) since they naturally retain non-ideal effects like electron inertia, Hall term, pressure anisotropy/nongyrotropy, etc. One obstacle to broader application of these model is that an explicit treatment of their source terms leads to the need to resolve rapid kinetic processes like plasma oscillation and electron cyclotron motion, even when they are not important. In this paper, we suggest two ways to address this issue. First, we derive the analytic forms solutions to the source update equations, which can be implemented as a practical, but less generic solver. We then develop a time-centered, locally implicit algorithm to update the source terms, allowing stepping over the fast kinetic time-scales. For a plasma with $S$ species, the locally implict algorithm involves inverting a local $3S+3$ matrix only, thus is very efficient. The performance can be further elevated by using the direct update formulas to skip null calculations. Benchmarks illustrated the exact energy-conservation of the locally implicit solver, as well as its efficiency and robustness for both small-scale, idealized problems and large-scale, complex systems. The locally implicit algorithm can be also easily extended to include other local sources, like collisions and ionization, which are difficult to solve analytically.

physics.comp-ph↗