All Seminars
| Title: Attention-enhanced Deep Learning Models for Data Cleaning and Integration |
|---|
| Defense: Computer Science |
| Speaker: Jing Zhang, Emory University |
| Contact: Joyce Ho, joyce.c.ho@emory.edu |
| Date: 2023-03-28 at 12:00PM |
| Venue: Modern Language 219 |
 Download Flyer Add to Calendar |
| Abstract: Data cleaning and integration is an essential process for ensuring the accuracy and consistency of data used in analytics and decision-making. Schema matching and entity matching tasks are crucial aspects of this process to merge data from various sources into a single, unified view. Schema matching seeks to identify and resolve semantic differences between two or more database schemas whereas entity matching seeks to detect the same real-world entities in different data sources. Given recent deep learning trends, pre-trained transformers have been proposed to automate both the schema matching and entity matching processes. However, existing models only utilize the special token representation (e.g., [CLS]) to predict matches and ignore rich and nuanced contextual information in the description, thereby yielding suboptimal matching performance. To improve performance, we propose the use of the attention mechanism to (1) learn the schema matches between source and target schemas using the attribute name and description, (2) leverage the individual token representations to fully capture the information present in the descriptions of the entities, and (3) jointly utilize the attribute descriptions and entity descriptions to perform both schema and entity matching. |
| Title: Enhancing Document Understanding through the Incorporation of Structural Inference |
|---|
| Defense: Computer Science |
| Speaker: Liyan Xu, Emory University |
| Contact: Dr. Jinho Choi, jinho.choi@emory.edu |
| Date: 2023-03-28 at 1:45PM |
| Venue: Carlos Hall Room 211 |
| Download Flyer Add to Calendar |
| Abstract: Towards resolving a variety of Natural Language Processing (NLP) tasks, pretrained language models (PLMs) have been incredibly successful by simply modeling language sequences, backed by their powerful sequence encoding capabilities. However, for document understanding tasks involving multi-sentence or multi-paragraph inputs, the model still needs to overcome the inherent challenge of processing scattered information across the entire document context, such as resolving pronouns or recognizing relations among multiple sentences. To address the motivation of effectively understanding document context beyond sequence modeling, this dissertation presents an in-depth study on the incorporation of structural inference, utilizing intrinsic structures of languages and documents. Four research works are outlined within this dissertation. Particularly, the first work proposes to integrate syntactic dependency structures into the document encoding process, capturing inter-sentence dependencies through designed graph encoding for the task of machine reading comprehension, especially under the multilingual setting. The second work investigates different methods to perform inference on the discourse structure that concerns coreference relations, allowing for higher-order decision making. The third work presents a novel formulation of structural inference to facilitate joint information extraction, fusing multi-facet information of document entities in terms of both coreference and relations. The last work explores the potential of the sequence-to-sequence generation as an approach that performs implicit inference on linearized entity structures, motivated by its unified encoder-decoder architecture and inherent abilities to perform higher-order inference. Overall, this dissertation demonstrates that incorporating designed structural inference upon certain intrinsic structures of languages or documents can effectively enhance document understanding, and highlights that modeling dependencies among different parts of the context can lead to more accurate and robust encoding and decoding process, where auxiliary information can be provided that complements the sequence modeling of PLMs. Zoom Option: https://emory.zoom.us/j/95753738482 |
| Title: Few Shot Learning for Rare Disease Diagnosis |
|---|
| Seminar: Computer Science |
| Speaker: Emily Alsentzer, MIT & Harvard |
| Contact: Vaidy Sunderam, VSS@emory.edu |
| Date: 2022-12-09 at 1:00PM |
| Venue: https://emory.zoom.us/j/95719302738 |
| Add to Calendar |
| Abstract: Rare diseases affect 300-400 million people worldwide, yet each disease affects no more than 50 per 100,000 individuals. Many patients with rare genetic conditions remain undiagnosed due to clinicians' lack of experience with the individual diseases and the considerable heterogeneity of clinical presentations. Machine-assisted diagnosis offers the opportunity to shorten the diagnostic delays for rare disease patients. Recent advances in deep learning have considerably improved the accuracy of medical diagnosis. However, much of the success thus far is contingent on the availability of large, annotated datasets. Machine-assisted rare disease diagnosis necessitates that approaches learn from limited data and extrapolate beyond training distribution to novel genetic conditions. In this talk, I will present our work towards developing few-shot methods that can overcome the data limitations of deep learning to diagnose patients with rare genetic conditions. To infuse external knowledge into models, we first develop novel graph neural network methods for subgraph representation learning that encode how subgraphs (e.g., a set of patient phenotypes) relate to a larger knowledge graph. We leverage these advances to develop SHEPHERD, a geometric deep learning method that reasons over biomedical knowledge to diagnose patients with rare–even novel–genetic conditions. SHEPHERD operates at multiple facets throughout the rare disease diagnosis process: performing causal gene discovery, retrieving "patients-like-me", and providing interpretable characterizations of novel disease presentations. Our work demonstrates the potential for deep learning methods to accelerate the diagnosis of rare disease patients and has implications for the use of deep learning on medical datasets with very few labels. |
| Title: Language Guided Localization and Navigation |
|---|
| Seminar: Computer Science |
| Speaker: Meera Hahn, |
| Contact: Jinho Choi, jinho.choi@emory.edu |
| Date: 2022-12-02 at 1:00PM |
| Venue: MSC W201 |
| Download Flyer Add to Calendar |
| Abstract: Embodied tasks that require active perception are key to improving language grounding models and creating holistic social agents. In this talk we explore two multi-modal embodied perception tasks which require localization or navigation of an agent in an unknown 3D space with limited information about the environment. First we present the Where Are You? (WAY) dataset which contains over 6k dialogs of two humans performing a localization task. On top of this dataset, we propose the task of Localization from Embodied Dialog (LED). The LED task involves taking a natural language dialog of two agents -- an observer and a locator -- and predicting the location of the observer agent. The second task we examine is the Vision Language Navigation (VLN) task, in which an agent navigates via natural language instructions. For both tasks, we address the objective of improving model accuracy and demonstrate that this can be done using passive data, which can introduce more semantically rich and diverse information during training, in comparison to additional interaction data. We additionally introduce a novel analysis pipeline for both tasks to diagnose and reveal limitations and failure modes of these types of common multi-modal models. Biography: Meera Hahn is a Research Scientist at Google Research working on multi-modal modeling of vision and natural language for applications in artificial intelligence. Her long-term research goal is to develop multi-modal systems capable of supporting robotic or AR assistants that can seamlessly interact with humans. She recently completed her PhD in Computer Science at the Georgia Institute of Technology under Dr. James M. Rehg. Her research at Georgia Tech focused on training embodied agents (in simulation) to perform complex semantic grounding tasks. http://cs.emory.edu/home/ |
| Title: Interpretable and Interactive Representation Learning on Geometric Data |
|---|
| Defense: Computer Science |
| Speaker: Yuyang Guo, Emory University |
| Contact: Liang Zhou, |
| Date: 2022-12-01 at 11:00AM |
| Venue: https://emory.zoom.us/j/5693008550 |
| Download Flyer Add to Calendar |
| Abstract: Abstract: In recent years, representation learning on geometrics data, such as image and graph-structured data, are experiencing rapid developments and achieving significant progress thanks to the rapid development of Deep Neural Networks (DNNs), including Convolutional Neural Networks (CNNs) and Graph Neural Networks (GNNs). However, DNNs typically offer very limited transparency, imposing significant challenges in observing and understanding when and why the models make successful/unsuccessful predictions. While we are witnessing the fast growth of research in local explanation techniques in recent years, the majority of the focus is rather handling how to generate the explanations, rather than understanding whether the explanations are accurate/reasonable, what if the explanations are inaccurate/unreasonable, and how to adjust the model to generate more accurate/reasonable explanations. To explore and answer the above questions, this dissertation aims to explore a new line of research called Explanation-Guided Learning (EGL) that intervenes the deep learning models' behavior through XAI techniques to jointly improve DNNs in terms of both their explainability and generalizability. Particularly, we propose to explore the EGL on geometric data, including image and graph-structured data, which are currently under-explored in the research community due to the complexity and inherent challenges in geometric data explanation. To achieve the above goals, we start by exploring the interpretability methods for geometric data on understanding the concepts learned by the deep neural networks (DNNs) with bio-inspired approaches and propose methods to explain the predictions of Graph Neural Networks (GNNs) on healthcare applications. Next, we design an interactive and general explanation supervision framework GNES for graph neural networks to enable the learning to explain pipeline, such that more reasonable and steerable explanations could be provided. Finally, we propose two generic frameworks, namely GRADIA and RES, for robust visual explanation-guided learning by developing novel explanation model objectives that can handle the noisy human annotation labels as the supervision signal with a theoretical justification of the benefit to model generalizability. This research spans multiple disciplines and promises to make general contributions in various domains such as deep learning, explainable AI, healthcare, computational neuroscience, and human-computer interaction by putting forth novel frameworks that can be applied to various real-world problems where both interpretability and task performance are crucial. |
| Title: Expressive computation: integrating programming and physical making |
|---|
| Seminar: Computer Science |
| Speaker: Jennifer Jacobs, University of California, Santa Barbara |
| Contact: Vaidy Sunderam, vss@emory.edu |
| Date: 2022-11-18 at 1:00PM |
| Venue: MSC W201 |
| Download Flyer Add to Calendar |
| Abstract: Abstract: Creators in many different fields use their hands. Artists and craftspeople manipulate physical materials, manufacturers manually controlmachine tools, and designers sketch ideas. Computers are increasingly displacing many manual practices in favor of procedural description and automated production. Despite this trend, computational and manual forms of creation are not mutually exclusive. In this talk, I argue that by developing methods to integrate computational and physical making, we can dramatically expand the expressive potential of computers and broaden participation in computational production. To support this argument, I will present research across three categories: 1) Integrating physical and manual creation with computer programming through domain-specific programming environments. 2) Broadening professional computational making through computational fabrication technologies. 3) Broadening entry points into computer science learning by blending programming with art, craft, and design. Collectively, my research demonstrates how developing computational workflows, representations, and interfaces for manual and physical making can enable manual creators to leverage existing knowledge and skills. Furthermore, I’ll discuss how collaborating with practitioners from art, craft, and manufacturing science can diversify approaches to knowledge production in systems engineering and open new research opportunities in computer science. Bio: Jennifer Jacobs is Assistant Professor at the University of California Santa Barbara in Media Arts and Technology and Computer Science (by courtesy). At UCSB, she directs the Expressive Computation Lab, which investigates ways to support expressive computer-aided design, art, craft, and manufacturing by developing new computational tools, abstractions, and systems that integrate emerging forms of computational creation and digital fabrication with traditional materials, manual control, and non-linear design practices. Prior to joining UCSB, Jennifer received her Ph.D. from the Massachusetts Institute of Technology and was a Postdoctoral Fellow at the Brown Institute of Media Innovation within the Department of Computer Science at Stanford University. She also received an M.F.A. and B.F.A from Hunter College and the University of Oregon respectively. Her research has been presented at leading human-computer interaction research venues and journals including UIST, DIS, SIGGRAPH, and, most prominently, at the flagship ACM Conference on Human Factors in Computing Systems (CHI). Zoom Option: https://emory.zoom.us/j/95719302738 |
| Title: Improving Interactive Search with User Feedback |
|---|
| Seminar: Computer Science |
| Speaker: Jianghong Zhou, Emory University |
| Contact: Eugene Agichtein, eugene.agichtein@emory.edu |
| Date: 2022-11-17 at 11:00AM |
| Venue: https://emory.zoom.us/j/95514127542?pwd=YWhuVFNZYTc3bUYzMy9wMWw1c3ZaQT09 |
| Download Flyer Add to Calendar |
| Abstract: Capturing users’ feedback can improve the interactive search. In search tasks, users typically generate feedback while browsing the search results. That feedback may include clicking the items, reading important text content, query reformulation, and other interactions. They can reveal users’ latent intents and additional information needs, providing essential extra information to improve users’ search experience. Unlike traditional search, the interactive search is enriched by more interactions and comprises three significant steps: Users browse the initial retrieval contents and generate feedback. The feedback is received and analyzed by the search system. The search system presents new search results based on users’ feedback. However, the complexity of human interactions challenges these three crucial steps when building an efficient interactive search engine. The first challenge is obtaining informative and valuable feedback from the users. This thesis introduces a new approach that can diversify the initial search results, allow users to explore multiple aspects of their original queries, and generate instructive feedback. The approach is the first to use Simpson’s Diversity Index and Binary quadratic optimization in search diversification problems. Compared to the previous research, this method is more efficient and fast. Another critical challenge is reducing the biased noise in the received feedback. In this thesis, we propose a novel de-biased method to decrease the feedback’s high bias caused by users’ observations. The approach uses a new observation mechanism to simulate the users’ observation process and train a neural network model to detect the observation bias. This new model outperforms the previous click models in both click simulation and document ranking. The last challenge is effectively extracting different interaction information and using them to improve the search. In this thesis, we focus on document-level and sentence-level interactions. We propose two different approaches with reinforcement learning frameworks. These methodologies introduce new techniques to reformulate the query and rank the items. Both methods significantly improve the search performance in the interactive search process. Together these techniques provide imperative solutions to the challenges in the three critical steps of the interactive search systems and enable the users to obtain a better search experience. |
| Title: Towards the Robustness of Deep Learning Systems Against Adversarial Examples in Sequential Data |
|---|
| Defense: Computer Science |
| Speaker: Wenjie Wang, Emory University |
| Contact: Li Xiong, lxiong@emory.edu |
| Date: 2022-11-17 at 3:00PM |
| Venue: zoom.us/j/9828106847 |
| Download Flyer Add to Calendar |
| Abstract: Recent studies have shown that adversarial examples can be generated by applying small perturbations to the inputs such that the well-trained deep neural networks (DNNs) will misclassify. With the increasing number of safety and security-sensitive applications of deep learning models, the robustness of deep learning models to adversarial inputs has become a crucial topic. Research on the adversarial examples in computer vision (CV) domains has been well studied. However, the intrinsic difference between image and sequential data has placed great challenges for directly applying adversarial techniques in CV to other application domains such as speech, health informatics, and natural language processing (NLP). To solve these gaps and challenges, My dissertation research combines multiple studies to improve the robustness of deep learning systems against adversarial examples in sequential inputs. First, We take the NLP and health informatics domains as examples, focusing on understanding the characteristics of these two domains individually and designing empirical adversarial defense methods, which are 1) RADAR, an adversarial detection for EHR data, and 2) MATCH, detecting adversarial examples leveraging the consistency between multiple modalities. Following the empirical defense methods, our next step is exploring certified robustness for sequential inputs which is provable and theory-backed. To this end, 1) We study the randomized smoothing on the word embedding space to provide certification to NLP models. 2) We propose WordDP, certified robustness to word substitution attacks in the NLP domain, leveraging the connection of differential privacy and certified robustness. 3) We studied the certified robustness methods to Wasserstein adversarial examples on univariant time-series data. |
| Title: The Applications of Alternating Minimization Algorithms on Deep Learning Models |
|---|
| Seminar: Computer Science |
| Speaker: Junxiang Wang, Emory University |
| Contact: Liang Zhao, |
| Date: 2022-11-15 at 11:30AM |
| Venue: https://emory.zoom.us/j/5693008550 |
| Download Flyer Add to Calendar |
| Abstract: Gradient Descent(GD) and its variants are the most popular optimizers for training deep learning models. However, they suffer from many challenges such as gradient vanishing and poor conditioning, which prevent their more widespread use. To address these intrinsic drawbacks, alternating minimization methods have attracted attention from researchers as a potential way to train deep learning models. Their idea is to decompose a neural network into a series of linear and nonlinear equality constraints, which generate multiple subproblems and they can be minimized alternately. Their empirical evaluations demonstrate good scalability and high accuracy. They also avoid gradient vanishing problems and allow for non-differentiable activation functions, as well as allowing for complex non-smooth regularization and the constraints that are increasingly important for neural network architectures. This dissertation aims to develop alternating minimization methods to train the Multi-Layer Perceptron(MLP) model. This includes deep learning Alternating Direction Method of Multipliers(dlADMM), monotonous Deep Learning Alternating Minimization(mDLAM), and parallel deep learning Alternating Direction Method of Multipliers(pdADMM). The extended pdADMM-G algorithm and the pdADMM-G-Q algorithms are developed to train the Graph-Augmented Multi-Layer Perceptron(GA-MLP) model. For the dlADMM algorithm, parameters in each layer are updated in a backward and forward fashion. The time complexity is reduced from cubic to quadratic in(latent) feature dimensions for subproblems by iterative quadratic approximations and backtracking. Finally, we provide the convergence guarantee of the dlADMM algorithm under mild conditions. For the mDLAM algorithm, our innovative inequality-constrained formulation infinitely approximates the original problem with non-convex equality constraints, enabling our convergence proof of the proposed mDLAM algorithm regardless of the choice of hyperparameters. Our mDLAM algorithm is shown to achieve a fast linear convergence by the Nesterov acceleration technique. For the pdADMM algorithm, we achieve model parallelism by breaking layer dependency: parameters in each layer of neural networks can be updated independently in parallel. The convergence of the proposed pdADMM to a stationary point is theoretically proven under mild conditions. The convergence rate of the pdADMM is proven to be $o(1/k)$, where $k$ is the number of iterations. For the pdADMM-G algorithm and the pdADMM-G-Q algorithm, in order to achieve model parallelism, we extend the proposed pdADMM algorithm to train the GA-MLP model, named the pdADMM-G algorithm. The extended pdADMM-G-Q algorithm reduces communication costs by introducing the quantization technique. Theoretical convergence to a (quantized) stationary point of two proposed algorithms is provided with a sublinear convergence rate $o(1/k)$, where $k$ is the number of iterations. |
| Title: Towards Designing Inclusive Social Virtual Reality Spaces to Combat New Forms of Online Harassment |
|---|
| Seminar: Computer Science |
| Speaker: Guo Freeman, Clemson University |
| Contact: Vaidy Sunderam, VSS@Emory.edu |
| Date: 2022-11-11 at 1:00PM |
| Venue: MSC W201 |
| Download Flyer Add to Calendar |
| Abstract: Abstract: Social Virtual Reality refers to 3D virtual spaces where multiple users can interact with one another through VR head-mounted displays. In recent years, the growing popularity of commercial social VR platforms such as AltspaceVR, VR Chat, RecRoom, and Meta Horizon Worlds is dramatically transforming how people meet, interact, play, and collaborate online and has led to the emerging metaverse paradigm. These platforms have drawn aspects from traditional multiplayer online games and 3D virtual worlds where users engage in various immersive experiences, interactive activities, and choices through avatar-based online representations. However, social VR also demonstrates specific nuances, including full/partial body tracked avatars, synchronous voice conversations, and simulated touching and grabbing features. These novel characteristics have led to greater instances of harassment and potentially more destructive consequences compared to traditional 3D virtual worlds/online gaming or single-user VR. In this talk, Dr. Guo Freeman will introduce her recent research on new forms of online harassment in social VR and how embodied harassment is becoming an emerging but understudied form of harassment in novel online social spaces. She will explain her ongoing work on leveraging innovative technologies, such as AI-based moderation, for proactively combating harassment in social VR. She will also highlight potential future directions for designing safer, inclusive, and more supportive social VR spaces to empower diverse communities, especially marginalized users such as women, ethnic minorities, and LGBTQ individuals. Bio: Dr. Guo Freeman is an Assistant Professor of Human-Centered Computing in the School of Computing at Clemson University. Her research situates at the unique intersection of social computing, social VR, and entertainment computing. She brings a combination of profound theoretical foundation, nuanced empirical perspectives, and participatory technology design and prototype to investigate how interactive technologies such as multiplayer online games, esports, live streaming, social VR, social media, and AI shape interpersonal relationships and group behavior. Her research is also uniquely driven by her focus on marginalized technology users due to their gender, race, sexuality, age, and disability, including women, LGBTQ individuals, ethnic minorities, minors, and persons with disabilities. At Clemson, she leads the Gaming and Mediated Experience Lab (CUGAME). She has authored over 80 peer-reviewed publications and won multiple Best Paper Honorable Mentions (top 5%) at CHI, CSCW, and iConference. She has secured $20.4 million in external grant funding from the National Science Foundation (NSF), Air Force Office of Scientific Research (AFOSR), and US Army, with $1.77 million dedicated to her effort. She is a member of the ACM CHI PLAY Steering Committee and has served on over 18 Program Committees for prestigious international HCI venues such as CHI, CSCW, and CHIPLAY. She especially dedicates to broadening women’s and minorities’ participation in computing and was a Grace Hopper Women in Computing Faculty Mentor. |