About

Professor Yi Zhang's research interests are in artificial intelligence, informational search and recommendation, natural language processing, machine learning, data mining, and computational economics. She has received various awards, including the ACM SIGIR Best Paper Award, National Science Foundation Faculty Career Award, Air Force Young Investigator Award, Google Research Award, Microsoft Research Award, and IBM Research Fellowship. She has served as program chair, area chair, and PC member for various top-tier conferences and was an associate editor for ACM Transaction on Information Systems. Dr. Zhang earned her Ph.D. and M.S. from the School of Computer Science at Carnegie Mellon University and her B.S. from Tsinghua University. She is also a co-founder of Rul.ai, a low/no code omni-channel conversational AI platform, and has been a consultant or technical adviser for several large companies including Alibaba, HP, Toyota, among others.

Research topics

• Computer Science
• Natural Language Processing
• Artificial Intelligence
• Programming language
• Physics
• Mathematics
• Engineering
• Optics
• Materials science
• Information Retrieval
• Parallel computing
• Biophysics
• Cartography
• Chemistry
• Geography
• Statistics
• Archaeology
• Database
• Virology
• Optoelectronics
• Chromatography
• Nanotechnology
• Theoretical computer science

Selected publications

• Polymer Architectures Built From Anthryl‐Containing Polystyrene and Their Topological Conversions

  Journal of Polymer Science · 2026-04-10

  article

  ABSTRACT This study elaborates on the preparation of anthryl‐containing polymers and their effective topological conversions conducted by dimerization of anthryl and cleavage of the anthryl dimer. A multifunctional polystyrene (PS) macroinitiator capped with anthryl groups at both chain ends is prepared and used in atom transfer radical polymerization (ATRP) of styrene to afford a triblock PS. Twin‐tail tadpole‐shaped and centipede‐shaped polymers are obtained by photoirradiation of triblock PS at 365 nm due to the dimerization of anthryl. The tadpole‐shaped and centipede‐shaped polymers are able to convert back to linear topology via the cleavage of anthryl dimer by photoirradiation at 280 nm or heating at 150°C. A multicyclic polymer with a repeating linear‐ring structure is afforded by atom transfer radical coupling (ATRC) reaction of twin‐tail tadpole‐shaped PS, and its conversion to a multi‐block linear topology is also realized.

  PublisherDOI

• Study on the effects of exercise intervention combined with virtual reality technology on emotions and brain networks in secondary school students with depressive symptoms

  Frontiers in Psychiatry · 2026-01-30 · 1 citations

  articleOpen access

  Background This study aimed to investigate the functional connectivity characteristics of brain networks in secondary school students with depressive symptoms and to analyze the effects of exercise combined with virtual reality intervention on improving brain networks and emotional states, providing a neurobiological basis for early identification and precise intervention. Methods This study recruited 98 middle school students aged 13 to 18 as research subjects, including 50 in the subclinical depression (ScD) group and 48 in the healthy control (HC) group. The experimental intervention employed a 2×3 two-way mixed design analysis of variance (Two-way ANOVA). All exercise intervention groups underwent 15 minutes of moderate-intensity (50%-80% HRmax) power cycling training. The exercise intervention combined with virtual reality technology group completed their training in an immersive natural landscape environment. Resting-state EEG signals were recorded before and after the intervention, and emotional state changes were assessed using the Positive and Negative Affect Schedule (PANAS). The cerebral cortex was segmented into 78 regions based on the Schaefer template. Phase-locked value ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline" id="im1"> <mml:mrow> <mml:mi>P</mml:mi> <mml:mi>L</mml:mi> <mml:mi>V</mml:mi> <mml:mo>=</mml:mo> <mml:mfrac> <mml:mn>1</mml:mn> <mml:mi>T</mml:mi> </mml:mfrac> <mml:mo>|</mml:mo> <mml:mstyle displaystyle="true"> <mml:msubsup> <mml:mo>∑</mml:mo> <mml:mrow> <mml:mi>t</mml:mi> <mml:mo>=</mml:mo> <mml:mn>1</mml:mn> </mml:mrow> <mml:mi>T</mml:mi> </mml:msubsup> <mml:mrow> <mml:msup> <mml:mi>e</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo stretchy="false">(</mml:mo> <mml:msub> <mml:mi>ϕ</mml:mi> <mml:mi>i</mml:mi> </mml:msub> <mml:mo stretchy="false">(</mml:mo> <mml:mi>t</mml:mi> <mml:mo stretchy="false">)</mml:mo> <mml:mo>-</mml:mo> <mml:msub> <mml:mi>ϕ</mml:mi> <mml:mi>j</mml:mi> </mml:msub> <mml:mo stretchy="false">(</mml:mo> <mml:mi>t</mml:mi> <mml:mo stretchy="false">)</mml:mo> <mml:mo stretchy="false">)</mml:mo> </mml:mrow> </mml:msup> </mml:mrow> </mml:mstyle> <mml:mo>|</mml:mo> </mml:mrow> </mml:math> ) was used as a functional connectivity metric to quantify brain network synchrony in the theta, alpha, and beta frequency bands. Statistical comparisons were performed using independent samples t-tests and two-way analysis of variance (ANOVA). Results Exercise intervention combined with virtual reality technology significantly improved θ and α band SMN-DMN, DAN-SN connectivity, and DMN/DAN activity ( p &amp;lt; 0.05), outperforming conventional exercise. β band SMN-DMN and CEN-DMN activity increased ( p&amp;lt; 0.05). The exercise intervention combined with virtual reality technology significantly increased positive emotions (t = -22.351, p &amp;lt; 0.05) and reduced negative emotions (t = 27.257, p &amp;lt; 0.001). Conclusion Depressive symptoms in adolescents are associated with multifrequency brain network dysregulation. Combining exercise intervention with virtual reality technology (VR-EI) optimizes key brain network connectivity and activity in the theta and alpha bands through multisensory stimulation. Its mood-enhancing effects surpass those of conventional exercise, offering a promising new strategy for personalized intervention in adolescent depression.

  PublisherDOI

• Dual-adiabatic spin-lock measurements for single-scan dispersion mapping in the human heart at 3T

  Journal of Cardiovascular Magnetic Resonance · 2026-01-01

  articleOpen access
  PublisherDOI

• A SP1-Driven Translational-Proteostatic Checkpoint Modulates the Autophagy-Proteasome Network

  Research Square · 2026-04-08

  preprintOpen access
  PublisherOA PDFDOI

• Non-polar nanocluster confinement engineering realizes high capacitive energy storage in Pb-free high-entropy relaxors

  Nature Communications · 2026-01-12 · 3 citations

  articleOpen access

  Dielectric ceramic capacitors with ultrahigh power density have become indispensable in modern power electronics, yet the persistent challenge of achieving superior energy density with high energy efficiency remains a critical bottleneck for practical applications. Herein, we propose an effective non-polar nanocluster confinement strategy through phase-field simulation-guided design of high-entropy (Bi0.2Na0.2Ba0.2Sr0.2Ca0.2)(Ti1-xSnx)O3 lead-free relaxor ferroelectrics. The incorporation of Sn4+ ions with low ionic polarizability leads to the formation of localized non-ferroelectric perovskite units, which constitute robust non-polar nanoclusters, being further stabilized and rendered immobile against electric fields by the substantial local random fields inherent to the high-entropy configuration. Consequently, these engineered non-polar nanoclusters serve as effective pinning centers to impede the merging and growth of polar nanodomains under electric fields, thereby reconciling the inherent conflict between polarization enhancement and hysteresis reduction. The optimized composition (x = 0.06) exhibits a high recoverable energy density of ~18.5 J·cm-3 together with an ultrahigh energy efficiency of ~92.4% in multilayer ceramic capacitors, representing a competitive combination among lead-free counterparts. This approach not only establishes a viable paradigm for next-generation energy storage dielectrics but also provides fundamental insights for designing functional materials with tailored electrical properties. The authors reconcile the conflict between polarization and hysteresis by confining polarization response of non-polar nanoclusters in high-entropy relaxors, thus improving the energy storage properties of multilayer ceramic capacitors.

  PublisherOA PDFDOI

• Electrolysis of Seawater and Salt-Lake Water for Hydrogen Production: A Review on Technical Challenges, Material Design, and Future Directions

  Energy & Fuels · 2026-02-17 · 1 citations

  article1st author

  Water electrolysis (WE) powered by renewable energy represents a pivotal pathway for large-scale hydrogen production. However, its heavy reliance on scarce, high-purity freshwater increasingly conflicts with global water-stress realities. Thus, the direct use of abundant nonpure water sources, such as seawater and salt-lake water, has emerged as a critical research frontier. This perspective provides a comprehensive, cross-technology analysis of the underlying principles, technical challenges, and recent advances in this field. First, alkaline, proton exchange membrane (PEM), anion exchange membrane (AEM), and solid oxide electrolysis pathways were compared, considering water-quality tolerance, energy efficiency, and durability. Subsequently, the specific chemistry of seawater and salt-lake electrolytes was examined, highlighting chloride-induced anode corrosion, competitive chlorine evolution, and cathodic mineral deposition as dominant failure modes. The state-of-the-art mitigation strategies were systematically summarized: (i) protective layers (MnOx, and Lewis-acidic oxides) that selectively block Cl– while preserving oxygen evolution reaction (OER) kinetics; (ii) oxygen-containing anion (PO43– and SO42–) modification of layered double hydroxides to repel chloride via electrostatic and intercalation effects; (iii) chloride-induced surface reconstruction that unexpectedly activates lattice-oxygen–mediated oxygen evolution reaction pathways; and (iv) system-level designs including highly alkaline electrolytes, permselective chloride-blocking anodes, pH-asymmetric cells, and decoupled redox cycles. Finally, we outline key remaining gaps and future research directions, offering guidance for advancing sustainable hydrogen production from nonpure water sources.

  PublisherDOI

• Three-dimentional multiscale simulation of flame inclination mechanisms in filtration combustion

  SSRN Electronic Journal · 2026-01-01

  preprintOpen access
  PublisherDOI

• Waveform Characteristics of OTFS and OFDM Modulated with Zadoff–Chu Sequences

  2026-01-09

  article1st authorCorresponding

  Integrated sensing and communications (ISAC) has emerged as a promising technology for future wireless networks, particularly for aerial platforms where efficient waveform designs are indispensable for joint communication and sensing. In this paper, we investigate the waveform characteristics of orthogonal time frequency space (OTFS) and orthogonal frequency division multiplexing (OFDM) signals modulated with Zadoff–Chu (ZC) sequences. We construct two-dimensional (2D) ZC sequence codewords for OTFS modulation and a truncated ZC sequence for OFDM, and analyze their delay–Doppler (DD) domain behavior using the ambiguity function. The analytical results show that both waveforms exhibit symmetric resolution properties in the delay and Doppler dimensions, while numerical evaluations demonstrate that ZC-modulated OTFS (ZC-OTFS) achieves a narrower mainlobe, a higher peak-to-sidelobe ratio (PSLR), and more stable sidelobes compared to ZC-modulated OFDM (ZC-OFDM), and simultaneously maintains a low peak-to-average power ratio (PAPR). In addition, the simulation results further confirm the DD symmetry of both waveforms.

  PublisherDOI

• A CDGFN-Based Quantum Multisource Information Fusion With Its Application in Time Series Classification

  IEEE Transactions on Knowledge and Data Engineering · 2026-01-12 · 3 citations

  article

  Time series classification (TSC) is a critical area with broad applications. In the field of evidence theory, quantum evidence theory (QET) offers a promising framework for onedimensional TSC tasks, leveraging the capabilities of quantum basic probability amplitude (QBPA) to capture two-dimensional uncertainty. However, as the first step for the application of QET to TSC, how to construct QBPA still remains an open issue. In this paper, a novel approach to generate QBPA is devised. Specifically, we first apply the discrete Fourier transform (DFT) to the original data, extracting two-dimensional features embedded in the magnitude and phase from the frequency domain based on the front-few multi-frequency components, achieved by setting a threshold frequency index (TFI) to limit the frequencies considered. Next, we introduce the complex dual gaussian fuzzy number (CDGFN) as a carrier for QBPA, effectively representing two-dimensional uncertainty in the data. A CDGFN-based multisource information fusion (CDGFN-MSIF) algorithm for decision-making is proposed to combine information from different frequency components. Finally, the decisionmaking algorithm is validated on multiple time series datasets. Experimental results highlight the superior performance of the proposed approach over other state-of-the-art models, demonstrating its effectiveness and enhanced classification accuracy.

  PublisherDOI

• How Small is Too Small? The Limits of Macrocycle Size in Active Template Cu-Mediated Alkyne-Azide Cycloaddition Reactions

  ChemRxiv · 2026-04-19

  article1st authorCorresponding

  To maximise the effect of the mechanical bond, optimise physicochemical properties or even just to simplify the synthesis of an interlocked target, it is often beneficial for the macrocycle to be as small as possible. However, although our previous study led to the broad adoption of 26 atom bipyridine macrocycles for the active template Cu-mediated alkyne-azide cycloaddition reaction, the lower size limit for the mono- and bi-pyridine macrocycles commonly employed in this approach remains unclear. Here we study the interplay between ring size and half-axle steric demand in determining interlocking efficiency. Importantly, although 24 and 23 atom mono- and bi-pyridine macrocycles respectively mediate mechanical bond formation with less sterically demanding half-axles, 24 atom bipyridine macrocycles appear to be more generally useful.

  PublisherDOI

Recent grants

• III: Small: Collaborative Research: Scrutable and Explainable Information Retrieval with Model Intrinsic and Agnostic Approaches

  NSF · $200k · 2020–2023

• III: Small: Towards Explainable Recommendation Systems

  NSF · $508k · 2019–2023

• CAREER: Future of Search: User, Social Networks and Language

  NSF · $537k · 2010–2016

• III-COR: Proactive Personalized Information Integration and Retrieval

  NSF · $333k · 2007–2011

Frequent coauthors

• Valia Kordoni

  Humboldt-Universität zu Berlin

  29 shared

• Yagang Yao

  Nanjing University

  24 shared

• Xu Sun

  Peking University

  20 shared

• Claire Gu

  University of California, Santa Cruz

  17 shared

• Haoting Niu

  Collaborative Innovation Center of Advanced Microstructures

  16 shared

• Meng Zhang

  Taiyuan University of Technology

  15 shared

• Yongfeng Zhang

  Rutgers, The State University of New Jersey

  14 shared

• Jin Z. Zhang

  University of California, Santa Cruz

  13 shared

Awards & honors

• ACM SIGIR Best Paper Award
• National Science Foundation Faculty Career Award
• Air Force Young Investigator Award
• Google Research Award
• Microsoft Research Award