Researcher profile

Hanqi Zhuang

Hanqi Zhuang contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate

eess.SP Sound Computation and Language cs.CY eess.AS eess.IV Machine Learning Social and Information Networks

Trust snapshot

Quick read

Trust 19 - Baseline

5works

0followers

8topics

4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2026arXiv

Unpacking Generative AI in Education: Computational Modeling of Teacher and Student Perspectives in Social Media Discourse

Generative AI (GAI) technologies are quickly reshaping the educational landscape. As adoption accelerates, understanding how students and educators perceive these tools is essential. This study presents one of the most comprehensive analyses to date of stakeholder discourse dynamics on GAI in education using social media data. Our dataset includes 1,199 Reddit posts and 13,959 corresponding top-level comments. We apply sentiment analysis, topic modeling, and author classification. To support this, we propose and validate a modular framework that leverages prompt-based large language models (LLMs) for analysis of online social discourse, and we evaluate this framework against classical natural language processing (NLP) models. Our GPT-4o pipeline consistently outperforms prior approaches across all tasks. For example, it achieved 90.6% accuracy in sentiment analysis against gold-standard human annotations. Topic extraction uncovered 12 latent topics in the public discourse with varying sentiment and author distributions. Teachers and students convey optimism about GAI's potential for personalized learning and productivity in higher education. However, key differences emerged: students often voice distress over false accusations of cheating by AI detectors, while teachers generally express concern about job security, academic integrity, and institutional pressures to adopt GAI tools. These contrasting perspectives highlight the tension between innovation and oversight in GAI-enabled learning environments. Our findings suggest a need for clearer institutional policies, more transparent GAI integration practices, and support mechanisms for both educators and students. More broadly, this study demonstrates the potential of LLM-based frameworks for modeling stakeholder discourse within online communities.

preprint2022arXiv

A Deep-Learning Usability Expansion Model of Ocean Observations

Today's ocean numerical prediction skills depend on the availability of in-situ and remote ocean observations at the time of the predictions only. Because observations are scarce and discontinuous in time and space, numerical models are often unable to accurately model and predict real ocean dynamics, leading to a lack of fulfillment of a range of services that require reliable predictions at various temporal and spatial scales. The process of constraining free numerical models with observations is known as data assimilation. The primary objective is to minimize the misfit of model states with the observations while respecting the rules of physics. The caveat of this approach is that measurements are used only once, at the time of the prediction. The information contained in the history of the measurements and its role in the determinism of the prediction is, therefore, not accounted for. Consequently, historical measurement cannot be used in real-time forecasting systems. The research presented in this paper provides a novel approach rooted in artificial intelligence to expand the usability of observations made before the time of the prediction. Our approach is based on the re-purpose of an existing deep learning model, called U-Net, designed specifically for image segmentation analysis in the biomedical field. U-Net is used here to create a Transform Model that retains the temporal and spatial evolution of the differences between model and observations to produce a correction in the form of regression weights that evolves spatially and temporally with the model both forward and backward in time, beyond the observation period. Using virtual observations, we show that the usability of the observation can be extended up to a one year prior or post observations.

preprint2021arXiv

Physics-informed Tensor-train ConvLSTM for Volumetric Velocity Forecasting of Loop Current

According to the National Academies, a weekly forecast of velocity, vertical structure, and duration of the Loop Current (LC) and its eddies is critical for understanding the oceanography and ecosystem, and for mitigating outcomes of anthropogenic and natural disasters in the Gulf of Mexico (GoM). However, this forecast is a challenging problem since the LC behaviour is dominated by long-range spatial connections across multiple timescales. In this paper, we extend spatiotemporal predictive learning, showing its effectiveness beyond video prediction, to a 4D model, i.e., a novel Physics-informed Tensor-train ConvLSTM (PITT-ConvLSTM) for temporal sequences of 3D geospatial data forecasting. Specifically, we propose 1) a novel 4D higher-order recurrent neural network with empirical orthogonal function analysis to capture the hidden uncorrelated patterns of each hierarchy, 2) a convolutional tensor-train decomposition to capture higher-order space-time correlations, and 3) to incorporate prior physic knowledge that is provided from domain experts by informing the learning in latent space. The advantage of our proposed method is clear: constrained by physical laws, it simultaneously learns good representations for frame dependencies (both short-term and long-term high-level dependency) and inter-hierarchical relations within each time frame. Experiments on geospatial data collected from the GoM demonstrate that PITT-ConvLSTM outperforms the state-of-the-art methods in forecasting the volumetric velocity of the LC and its eddies for a period of over one week.

preprint2020arXiv

North Atlantic Right Whales Up-call Detection Using Multimodel Deep Learning

A new method for North Atlantic Right Whales (NARW) up-call detection using Multimodel Deep Learning (MMDL) is presented in this paper. In this approach, signals from passive acoustic sensors are first converted to spectrogram and scalogram images, which are time-frequency representations of the signals. These images are in turn used to train an MMDL detec-tor, consisting of Convolutional Neural Networks (CNNs) and Stacked Auto Encoders (SAEs). Our experimental studies revealed that CNNs work better with spectrograms and SAEs with sca-lograms. Therefore in our experimental design, the CNNs are trained by using spectrogram im-ages, and the SAEs are trained by using scalogram images. A fusion mechanism is used to fuse the results from individual neural networks. In this paper, the results obtained from the MMDL detector are compared with those obtained from conventional machine learning algorithms trained with handcraft features. It is shown that the performance of the MMDL detector is sig-nificantly better than those of the representative conventional machine learning methods in terms of up-call detection rate, non-up-call detection rate, and false alarm rate.

preprint2016arXiv

Detection of north atlantic right whale upcalls using local binary patterns in a two-stage strategy

In this paper, we investigate the effectiveness of two-stage classification strategies in detecting north Atlantic right whale upcalls. Time-frequency measurements of data from passive acoustic monitoring devices are evaluated as images. Vocalization spectrograms are preprocessed for noise reduction and tone removal. First stage of the algorithm eliminates non-upcalls by an energy detection algorithm. In the second stage, two sets of features are extracted from the remaining signals using contour-based and texture based methods. The former is based on extraction of time-frequency features from upcall contours, and the latter employs a Local Binary Pattern operator to extract distinguishing texture features of the upcalls. Subsequently evaluation phase is carried out by using several classifiers to assess the effectiveness of both the contour-based and texture-based features for upcall detection. Experimental results with the data set provided by the Cornell University Bioacoustics Research Program reveal that classifiers show accuracy improvements of 3% to 4% when using LBP features over time-frequency features. Classifiers such as the Linear Discriminant Analysis, Support Vector Machine, and TreeBagger achieve high upcall detection rates with LBP features.