arXiv subjects

Yong Liu

Publications and source records attributed to Yong Liu.

At least 937 records · Page 52Linked to original sources

arXiv2019

Recent Advances in Diversified Recommendation

With the rapid development of recommender systems, accuracy is no longer the only golden criterion for evaluating whether the recommendation results are satisfying or not. In recent years, diversity has gained tremendous attention in recommender systems research, which has been recognized to be an important factor for improving user satisfaction. On the one hand, diversified recommendation helps increase the chance of answering ephemeral user needs. On the other hand, diversifying recommendation results can help the business improve product visibility and explore potential user interests. In this paper, we are going to review the recent advances in diversified recommendation. Specifically, we first review the various definitions of diversity and generate a taxonomy to shed light on how diversity have been modeled or measured in recommender systems. After that, we summarize the major optimization approaches to diversified recommendation from a taxonomic view. Last but not the least, we project into the future and point out trending research directions on this topic.

cs.IR↗

arXiv2019

Phonon and electron transport in Janus monolayers based on InSe

We systematically investigated the phonon and electron transport properties of monolayer InSe and its Janus derivatives including monolayer In2SSe and In2SeTe by first-principles calculations. The breaking of mirror symmetry produce a distinguishable A1 peak in the Raman spectra of monolayer In2SSe and In2SeTe. The room-temperature thermal conductivity (\k{appa}) of monolayer InSe, In2SSe and In2SeTe is 44.6, 46.9, and 29.9 W/(m K), respectively. There is a competition effect between atomic mass, phonon group velocity and phonon lifetime. The \k{appa} can be further effectively modulated by sample size for the purpose of thermoelectric applications. Meanwhile, monolayer In2SeTe exhibits a direct band and higher electron mobility than that of monolayer InSe, due to the smaller electron effective mass caused by tensile strain on the Se side. These results indicate that 2D Janus group-III chalcogenides can provide a platform to design the new electronic, optoelectronic and thermoelectric devices.

cond-mat.mtrl-sci↗

arXiv2019

Reversible Video Data Hiding Using Zero QDCT Coefficient-Pairs

H.264/Advanced Video Coding (AVC) is one of the most commonly used video compression standard currently. In this paper, we propose a Reversible Data Hiding (RDH) method based on H.264/AVC videos. In the proposed method, the macroblocks with intra-frame $4\times 4$ prediction modes in intra frames are first selected as embeddable blocks. Then, the last zero Quantized Discrete Cosine Transform (QDCT) coefficients in all $4\times 4$ blocks of the embeddable macroblocks are paired. In the following, a modification mapping rule based on making full use of modification directions are given. Finally, each zero coefficient-pair is changed by combining the given mapping rule with the to-be-embedded information bits. Since most of last QDCT coefficients in all $4\times 4$ blocks are zero and they are located in high frequency area. Therefore, the proposed method can obtain high embedding capacity and low distortion.

cs.MM↗

arXiv2019

An Evaluation of Transfer Learning for Classifying Sales Engagement Emails at Large Scale

This paper conducts an empirical investigation to evaluate transfer learning for classifying sales engagement emails arising from digital sales engagement platforms. Given the complexity of content and context of sales engagement, lack of standardized large corpora and benchmarks, limited labeled examples and heterogenous context of intent, this real-world use case poses both a challenge and an opportunity for adopting a transfer learning approach. We propose an evaluation framework to assess a high performance transfer learning (HPTL) approach in three key areas in addition to commonly used accuracy metrics: 1) effective embeddings and pretrained language model usage, 2) minimum labeled samples requirement and 3) transfer learning implementation strategies. We use in-house sales engagement email samples as the experiment dataset, which includes over 3000 emails labeled as positive, objection, unsubscribe, or not-sure. We discuss our findings on evaluating BERT, ELMo, Flair and GloVe embeddings with both feature-based and fine-tuning approaches and their scalability on a GPU cluster with increasingly larger labeled samples. Our results show that fine-tuning of the BERT model outperforms with as few as 300 labeled samples, but underperforms with fewer than 300 labeled samples, relative to all the feature-based approaches using different embeddings.

cs.CL↗

arXiv2019

Dielectric Resonator Method For Determining Gap Symmetry Of Superconductors Through Anisotropic Nonlinear Meissner Effect

We present a new measurement method which can be used to image gap nodal structure of superconductors whose pairing symmetry is under debate. This technique utilizes a high quality factor microwave resonance involving the sample of interest. While supporting a circularly symmetric standing wave current pattern, the sample is perturbed by a scanned laser beam, creating a photoresponse that was previously shown to reveal the superconducting gap anisotropy. Simulation and the measurement of the photoresponse of an unpatterned Nb film show less than 8\% anisotropy, as expected for a superconductor with nearly isotropic energy gap along with expected systematic uncertainty. On the other hand, measurement of a YBa$_2$Cu$_3$O$_{7-δ}$ thin film shows a clear 4-fold symmetric image with $\mathtt{\sim}12.5$\% anisotropy, indicating the well-known 4-fold symmetric $d_{x^2-y^2}$ gap nodal structure in the $ab$-plane. The deduced gap nodal structure can be further cross-checked by low temperature surface impedance data, which is simultaneously measured. The important advantage of the presented method over the previous spiral resonator method is that it does not require a complicated lithographic patterning process which limits one from testing various kinds of materials due to photoresponse arising from patterning defects. This advantage of the presented technique, and the ability to measure unpatterned samples such as planar thin films and single crystals, enables one to survey the pairing symmetry of a wide variety of unconventional superconductors.

cond-mat.supr-con↗

arXiv2019

Electronic, magnetic, and optical properties of Mn-doped GaSb: a first-principles study

Half-metallic ferromagnets can produce fully spin-polarized conduction electrons and can be applied to fabricate spintronic devices. Thus, in this study, the electronic structure, magnetic properties, and optical properties of GaSb, which has exhibited half-metallicity, doped with Mn, a 3d transition metal, are calculated using the generalized gradient approximation and Heyd-Scuseria-Ernzerhof (HSE) functional. Ga$_{1-x}$Mn$_x$Sb ($x = 0.25, 0.5, 0.75$) materials exhibit ferromagnetic half-metallic properties and a high Curie temperature, indicating that this series can applied in spintronic devices. Meanwhile, they absorb strongly in the infrared band, suggesting that Ga$_{1-x}$Mn$_{x}$Sb also has potential applications in infrared photoelectric devices.

cond-mat.mtrl-sci↗

arXiv2019

Diversity-Promoting Deep Reinforcement Learning for Interactive Recommendation

Interactive recommendation that models the explicit interactions between users and the recommender system has attracted a lot of research attentions in recent years. Most previous interactive recommendation systems only focus on optimizing recommendation accuracy while overlooking other important aspects of recommendation quality, such as the diversity of recommendation results. In this paper, we propose a novel recommendation model, named \underline{D}iversity-promoting \underline{D}eep \underline{R}einforcement \underline{L}earning (D$^2$RL), which encourages the diversity of recommendation results in interaction recommendations. More specifically, we adopt a Determinantal Point Process (DPP) model to generate diverse, while relevant item recommendations. A personalized DPP kernel matrix is maintained for each user, which is constructed from two parts: a fixed similarity matrix capturing item-item similarity, and the relevance of items dynamically learnt through an actor-critic reinforcement learning framework. We performed extensive offline experiments as well as simulated online experiments with real world datasets to demonstrate the effectiveness of the proposed model.

cs.IR↗

arXiv2019

Strong phonon anharmonicity and low thermal conductivity of monolayer tin oxides driven by lone-pair electrons

Motivated by the excellent electronic and optoelectronic properties of two-dimensional (2D) tin oxides, we systematically investigated the thermal conductivity of monolayer SnO and SnO2by the first-principles calculations. The room-temperature thermal conductivity of monolayer SnO and SnO2reaches 9.6 W/(mK) and 98.8 W/(mK), respectively. The size effect is much weaker for monolayer SnO than for monolayer SnO2, due to the coexistence of size dependent and independent component in the thermal conductivity of monolayer SnO. The large difference between the thermal conductivity of 2D tin oxides can be attributed to the small phonon group velocity and strong anharmonicity strength of monolayer SnO. Further electronic structure analysis reveals that the existence of lone-pair Sn-5s electrons is the key factor for the small \k{appa} of monolayer SnO. These results provide a guide for the manipulation of thermal transport in the electronic or thermoelectric devices based on 2D tin oxides.

cond-mat.mtrl-sci↗

arXiv2019

Multi-Scale Quasi-RNN for Next Item Recommendation

How to better utilize sequential information has been extensively studied in the setting of recommender systems. To this end, architectural inductive biases such as Markov-Chains, Recurrent models, Convolutional networks and many others have demonstrated reasonable success on this task. This paper proposes a new neural architecture, multi-scale Quasi-RNN for next item Recommendation (QR-Rec) task. Our model provides the best of both worlds by exploiting multi-scale convolutional features as the compositional gating functions of a recurrent cell. The model is implemented in a multi-scale fashion, i.e., convolutional filters of various widths are implemented to capture different union-level features of input sequences which influence the compositional encoder. The key idea aims to capture the recurrent relations between different kinds of local features, which has never been studied previously in the context of recommendation. Through extensive experiments, we demonstrate that our model achieves state-of-the-art performance on 15 well-established datasets, outperforming strong competitors such as FPMC, Fossil and Caser absolutely by 0.57%-7.16% and relatively by 1.44%-17.65% in terms of MAP, Recall@10 and NDCG@10.

cs.IR↗

arXiv2019

Efficient Cross-Validation for Semi-Supervised Learning

Manifold regularization, such as laplacian regularized least squares (LapRLS) and laplacian support vector machine (LapSVM), has been widely used in semi-supervised learning, and its performance greatly depends on the choice of some hyper-parameters. Cross-validation (CV) is the most popular approach for selecting the optimal hyper-parameters, but it has high complexity due to multiple times of learner training. In this paper, we provide a method to approximate the CV for manifold regularization based on a notion of robust statistics, called Bouligand influence function (BIF). We first provide a strategy for approximating the CV via the Taylor expansion of BIF. Then, we show how to calculate the BIF for general loss function,and further give the approximate CV criteria for model selection in manifold regularization. The proposed approximate CV for manifold regularization requires training only once, hence can significantly improve the efficiency of traditional CV. Experimental results show that our approximate CV has no statistical discrepancy with the original one, but much smaller time cost.

cs.LG↗

arXiv2019

Crystal growth, microstructure and physical properties of SrMnSb$_2$

We report on the crystal and magnetic structures, magnetic, and transport properties of SrMnSb$_2$ single crystals grown by the self-flux method. Magnetic susceptibility measurements reveal an antiferromagnetic (AFM) transition at $T_{\rm N} = 295(3)$ K. Above $T_{\rm N}$, the susceptibility slightly increases and forms a broad peak at $T \sim 420$ K, which is a typical feature of two-dimensional magnetic systems. Neutron diffraction measurements on single crystals confirm the previously reported C-type AFM structure below $T_{\rm N}$. Both de Haas-van Alphen (dHvA) and Shubnikov-de Haas (SdH) effects are observed in SrMnSb$_2$ single crystals. Analysis of the oscillatory component by a Fourier transform shows that the prominent frequencies obtained by the two different techniques are practically the same within error regardless of sample size or saturated magnetic moment. Transmission electron microscopy (TEM) reveals the existence of stacking faults in the crystals, which result from a horizontal shift of Sb atomic layers suggesting possible ordering of Sb vacancies in the crystals. Increase of temperature in susceptibility measurements leads to the formation of a strong peak at $T \sim {570}$ K that upon cooling under magnetic field the susceptibility shows a ferromagnetic transition at $T_{\rm C} \sim 580$ K. Neutron powder diffraction on crushed single-crystals does not support an FM phase above $T_{\rm N}$. Furthermore, X-ray magnetic circular dichroism (XMCD) measurements of a single crystal at the $L_{2,3}$ edge of Mn shows a signal due to induced canting of AFM moments by the applied magnetic field. All evidence strongly suggests that a chemical transformation at the surface of single crystals occurs above 500 K concurrently producing a minute amount of ferromagnetic impurity phase.

cond-mat.mtrl-sci↗

arXiv2019

Very Long Term Field of View Prediction for 360-degree Video Streaming

360-degree videos have gained increasing popularity in recent years with the developments and advances in Virtual Reality (VR) and Augmented Reality (AR) technologies. In such applications, a user only watches a video scene within a field of view (FoV) centered in a certain direction. Predicting the future FoV in a long time horizon (more than seconds ahead) can help save bandwidth resources in on-demand video streaming while minimizing video freezing in networks with significant bandwidth variations. In this work, we treat the FoV prediction as a sequence learning problem, and propose to predict the target user's future FoV not only based on the user's own past FoV center trajectory but also other users' future FoV locations. We propose multiple prediction models based on two different FoV representations: one using FoV center trajectories and another using equirectangular heatmaps that represent the FoV center distributions. Extensive evaluations with two public datasets demonstrate that the proposed models can significantly outperform benchmark models, and other users' FoVs are very helpful for improving long-term predictions.

cs.CV↗

arXiv2019

Max-Diversity Distributed Learning: Theory and Algorithms

We study the risk performance of distributed learning for the regularization empirical risk minimization with fast convergence rate, substantially improving the error analysis of the existing divide-and-conquer based distributed learning. An interesting theoretical finding is that the larger the diversity of each local estimate is, the tighter the risk bound is. This theoretical analysis motivates us to devise an effective maxdiversity distributed learning algorithm (MDD). Experimental results show that MDD can outperform the existing divide-andconquer methods but with a bit more time. Theoretical analysis and empirical results demonstrate that our proposed MDD is sound and effective.

cs.LG↗

arXiv2019

The direct and indirect optical absorptions of cubic BAs and BSb

Recently, boron arsenide (BAs) has been measured high thermal conductivity in the experiments, great encouraging for the low-power photoelectric devices. Therefore, in the present work, we have systematically investigated the direct and indirect optical absorptions of BAs and BSb and the doping effect of congeners by using first-principles calculations. We obtain the absorption onset corresponding to the value of indirect bandgap by considering the phonon-assisted second-order optical absorptions. And the redshift of absorption onset, enhancement and smoothness of optical absorptions spectra are also captured in the temperature-dependent calculations. In order to introduce one-order absorptions into the visible range, the doping effect of congeners on optical absorptions is studied without the assists of phonon. It is found that the decrease of local direct bandgap after doping derives from either the small bandgap in the prototypical III-V semiconductors or CBM locating at R$_c$ point. Thus, doping of congeners can improve the direct optical absorptions in visible range.

cond-mat.mtrl-sci↗

arXiv2019

Hexagonal MASnI$_3$ exhibiting strong absorption of ultraviolet photons

MASnI$_3$, an organometallic halide, has great potential in the field of lead-free perovskite solar cells. Ultraviolet photons have been shown to generate deep trapping electronic defects in mesoporous TiO$_2$-based perovskite, affecting its performance and stability. In this study, the structure, electronic properties, and optical properties of the cubic, tetragonal, and hexagonal phases of MASnI$_3$ were studied using first-principles calculations. The results indicate that the hexagonal phase of MASnI$_3$ possesses a larger indirect band gap and larger carrier effective mass along the \emph{c}-axis compared with the cubic and tetragonal phases. These findings were attributed to the enhanced electronic coupling and localization in the hexagonal phase. Moreover, the hexagonal phase exhibited high absorption of ultraviolet photons and high transmission of visible photons, particularly along the \emph{c}-axis. These characteristics demonstrate the potential of hexagonal MASnI$_3$ for application in multijunction perovskite tandem solar cells or as coatings in mesoporous TiO$_2$-based perovskite solar cells to enhance ultraviolet stability and photon utilization.

cond-mat.mtrl-sci↗

arXiv2018

Effects of layer stacking and strain on electronic transport in 2D tin monoxide

Tin monoxide is an interesting two-dimensional material because of the rare oxide semiconductor with bipolar conductivity. However, the lower room temperature mobility limits the applications of SnO in the future. Thus, we systematically investigate the effects of the different layer structures and strains on the electron-phonon coupling and phonon-limited mobility of SnO. The A2u phonon mode in the high frequency region is the main contributor coupling with electron for the different layer structures. And the orbital hybridization of Sn atoms existing only in bilayer structure changes the conduction band edge and decreases the electron-phonon coupling conspicuously, thus the electronic transport performance of bilayer is superior to others. In addition, the compressive strain of -1.0% in monolayer structure makes CBM consist of two valleys at Gamma point and along M-Gamma line, also leads to the intervalley electronic scattering assisted by Eg-1 mode. However, the electron-phonon coupling regional transferring from high frequency (A2u) to low frequency (Eg-1) results in the little significant change of mobility.

cond-mat.mtrl-sci↗

arXiv2018

The Doping effect of Chalcogen on the Two-Dimensional Ferromagnetic Material Chromium Tribromide

Recently the discovery of magnetic order in two-dimensional monolayer chromium trihalides opens the new research field in two-dimensional materials. We use first-principles calculations to systematically examine the doping effect of chalcogen on CrBr3. In the case of S-doping, four stable configurations, Cr2Br5S, Cr2Br4S2-A, Cr2Br4S2-B and Cr2Br3S3-A, are predicted to be ferromagnetic semiconductors. It is found that the new bands appearing in the original bandgap are made up of S-p and Cr-d-egorbits, lead to the obvious reduce of bandgap and the enhanced optical absorption in the visible range. Due to the decrease of valence electron after chalcogen doping, the magnetic moment also decreases with the increase of S atoms, and the character of ferromagnetic semiconductor is always hold in a wide range of strain. The results shown that monolayer CrBr3with chalcogen doping supply a effectual way to control the magnetism and extend the optoelectronic applications.

cond-mat.mtrl-sci↗

arXiv2018

Approach for Semi-automatic Construction of Anti-infective Drug Ontology Based on Entity Linking

Ontology can be used for the interpretation of natural language. To construct an anti-infective drug ontology, one needs to design and deploy a methodological step to carry out the entity discovery and linking. Medical synonym resources have been an important part of medical natural language processing (NLP). However, there are problems such as low precision and low recall rate. In this study, an NLP approach is adopted to generate candidate entities. Open ontology is analyzed to extract semantic relations. Six-word vector features and word-level features are selected to perform the entity linking. The extraction results of synonyms with a single feature and different combinations of features are studied. Experiments show that our selected features have achieved a precision rate of 86.77%, a recall rate of 89.03% and an F1 score of 87.89%. This paper finally presents the structure of the proposed ontology and its relevant statistical data.

cs.CL↗