SFU Data Science Research Group

Home
Admissions
Trustworthy Data Science and AI Seminars (current)
Courses

Seminar Series on Trustworthy Data Science and AI

Brought to you by SFU Data Science Research Group

Upcoming

Past

Coordinators

Jian Pei

Jian Pei

Jiannan Wang

Jiannan Wang

Tianzheng Wang

Tianzheng Wang

© Data Science Research Group, Simon Fraser University

Making an Impact? A Tale of Two Projects

Speaker:	Kevin Leyton-Brown, University of British Columbia; Canada CIFAR AI Chair, Amii
Time, Location:	2:30-4:00pm (Pacific Time), Oct 16, 2020. Online. [Recording] [Slides] [Poster]
Abstract:	How can AI researchers leverage their specialized knowledge to make a social impact? The notion is beguiling but the reality is complicated. This talk contrasts two strategies that are often employed--loosely described as “write a paper” and “be an entrepreneur”--gained via two, very different projects in electronic market design. The first project focused on developing new theoretical ideas for incentivizing local food pantries to honestly report demand to a centralized food bank. The second project was more practical; it aimed to design an electronic market for agricultural commodities in Uganda that could operate over low-end SMS phones. After discussing technical innovations, lessons learned, and lingering disappointments from both projects, the talk will conclude with some overall thoughts about strategies researchers might employ in pursuit of successful AI for Social Impact projects and how these can be taught in our courses.
Speaker Bio:	Kevin Leyton-Brown is a professor of Computer Science at the University of British Columbia and holds a Canada CIFAR AI Chair at the Alberta Machine Intelligence Institute; he is also an associate member of the Vancouver School of Economics. He holds a PhD and M.Sc. from Stanford University (2003; 2001) and a B.Sc. from McMaster University (1998). He studies artificial intelligence, mostly at the intersection of machine learning and (1) the design and operation of electronic markets and (2) the design of heuristic algorithms. He has co-written two books, "Multiagent Systems" and "Essentials of Game Theory," and over 100 peer-refereed technical articles; his work has received over 16,000 citations and an h-index of 49. He is an Fellow of the Association for the Advancement of Artificial Intelligence (AAAI; elected 2018) and a Distinguished Member of the ACM (elected 2018). He was a member of a team that won the 2018 INFORMS Franz Edelman Award for Achievement in Advanced Analytics, Operations Research and Management Science, described as "the leading O.R. and analytics award in the industry." This award recognizes a "completed, practical application that had significant, verifiable and quantifiable impact on the performance of [a] client organization," in his case the Federal Communications Commission. Leyton-Brown also received UBC's 2015 Charles A. McDowell Award for Excellence in Research, a 2014 NSERC E.W.R. Steacie Memorial Fellowship—previously given to a computer scientist only 10 times since its establishment in 1965—and a 2013 Outstanding Young Computer Science Researcher Prize from the Canadian Association of Computer Science. He and his coauthors have received paper awards from JAIR, ACM-EC, AAMAS and LION, and numerous medals for the portfolio-based SAT solver SATzilla at international SAT solver competitions (2003–15). He is Program Co-Chair for AAAI 2021, one of the top two international conferences on artificial intelligence. He has co-taught two Coursera courses on "Game Theory" to over 850,000 students (and counting!), and has received awards for his teaching at UBC—notably, a 2013/14 Killam Teaching Prize. He is past chair of the ACM Special Interest Group on Electronic Commerce, which runs the annual Economics & Computation conference. He has served as an associate editor for the Artificial Intelligence Journal (AIJ), the Journal of Artificial Intelligence Research (JAIR), ACM Transactions on Economics and Computation (ACM-TEAC), and AI Access; and was program chair for the ACM Conference on Electronic Commerce (ACM-EC) in 2012. He has held a wide range of visiting professor/researcher positions, at: Technion Israel Institute of Technology (April 2018); Microsoft Research New York City (Jan–Feb 2018); Microsoft Research New England (Mar–Jun 2016); Harvard's EconCS group (Mar–Jun 2016); the Simons Institute for the Theory of Computing at UC Berkeley (Aug–Sep, Nov 2016; Sep-Dec 2015); the Institute for Advanced Studies at Hebrew University of Jerusalem, Israel (Mar–Jun 2011); Makerere University in Kampala, Uganda (Sep 2010–Jan 2011). He currently advises Auctionomics, AI21, and Qudos. He is a co-founder of Kudu.ug and Meta-Algorithmic Technologies. He was scientific advisor to UBC spinoff Zite until it was acquired by CNN in 2011. His past consulting has included work for Zynga, Trading Dynamics, Ariba, and Cariocas.
Host:	Jian Pei

Getting Rid of Data

Speaker:	Tova Milo, Tel Aviv University; ACM Fellow, Member of Academia Europaea
Time, Location:	10:00am (Pacific time), Nov 12, 2020 Online. [Recording] [Slides] [Poster]
Abstract:	We are experiencing an amazing data-centered revolution. Incredible amounts of data are collected, integrated and analyzed, leading to key breakthroughs in science and society. This well of knowledge, however, is at a great risk if we do not dispense with some of the data flood. First, the amount of generated data grows exponentially and already at 2025 is expected to exceed the available storage by an order of magnitude. Second, even disregarding storage constraints, uncontrolled data retention risks privacy and security, as recognized, e.g., by the recent EU Data Protection reform. Data disposal policies must be developed to benefit and protect organizations and individuals. Retaining the knowledge hidden in the data while respecting storage, processing and regulatory constraints is a great challenge. The difficulty stems from the distinct, intricate requirements entailed by each type of constraint, the scale and velocity of data and the constantly evolving needs. While multiple data sketching, summarization and deletion techniques were developed to address specific aspects of the problem, we are still very far from a comprehensive solution. Every organization has to battle the same tough challenges, with ad hoc solutions that are application specific and rarely sharable. In this talk I will discuss the logical, algorithmic, and methodological foundations required for the systematic disposal of large-scale data, for constraints enforcement and for the development of applications over the retained information. I will overview relevant related work, highlighting new research challenges and potential reuse of existing techniques, as well as the research performed in this direction in the Tel Aviv Databases group.
Speaker Bio:	Tova Milo is the Dean of the Faculty of Exact Science at Tel Aviv University. She received her Ph.D. degree in Computer Science from the Hebrew University, Jerusalem, in 1992. After graduating she worked at the INRIA research institute in Paris and at University of Toronto and returned to Israel in 1995, joining the School of Computer Science at Tel Aviv university, where now she is a full Professor and holds the Chair of Information Management. She served as the Head of the Computer Science Department from 2011-2014. Her research focuses on large-scale data management applications such as data integration, semi-structured information, Data-centered Business Processes and Crowd-sourcing, studying both theoretical and practical aspects. Tova served as the Program Chair of multiple international conferences, including PODS, VLDB, ICDT, XSym, and WebDB, and as the chair of the PODS Executive Committee. She served as a member of the VLDB Endowment and the PODS and ICDT executive boards and as an editor of TODS, IEEE Data Eng. Bull, and the Logical Methods in Computer Science Journal. Tova has received grants from the Israel Science Foundation, the US-Israel Binational Science Foundation, the Israeli and French Ministry of Science and the European Union. She is an ACM Fellow, a member of Academia Europaea, a recipient of the 2010 ACM PODS Alberto O. Mendelzon Test-of-Time Award, 2017 VLDB Women in Database Research award and Weizmann award for Exact Sciences Research. The prestigious EU ERC Advanced Investigators grant.
Host:	Jian Pei

Humans in Online Job Markets

Speaker:	Sihem Amer-Yahia, Computer Scientist; Director of French National Centre for Scientific Research, Editor-in-Chief of the VLDB Journal for Europe and Africa
Time, Location:	1:00pm (Pacific time), Nov 18, 2020 Online. [Recording] [Slides] [Poster]
Abstract:	Online labor markets are increasingly becoming a destination for work. These marketplaces include freelancing platforms such as Qapa and MisterTemp' in France, and TaskRabbit and Fiverr in the USA. On those platforms, workers can find temporary jobs in the physical world such as moving furniture, or in the form of virtual micro-gigs such as helping with designing a website. I will how human factors affect algorithm design on virtual marketplaces, and a study of algorithmic fairness on those platforms. The talk will end with open questions raised a year ago at a Shonan workshop on the Future of Work.
Speaker Bio:	Sihem Amer-Yahia is a Silver Medal CNRS Research Director and Co-Head of the Lab of Informatics of Grenoble. She works on exploratory data analysis and fairness in job marketplaces. Before joining CNRS, she was Principal Scientist at QCRI, Senior Scientist at Yahoo! Research and Member of Technical Staff at at&t Labs. Sihem currently leads the SIGMOD and VLDB Diversity&Inclusion initiative.
Host:	Jian Pei

Graph Representation Learning for Drug Discovery

Speaker:	Jian Tang, Mila-Quebec AI Institute; HEC Montreal, CIFAR AI Research Chair
Time, Location:	4:30pm (Pacific time), Dec 16, 2020 Online. [Recording] [Slides] [Poster]
Abstract:	Drug discovery is a very long and expensive process, taking on average more than 10 years and costing $2.5B to develop a new drug. Artificial intelligence has the potential to significantly accelerate the process of drug discovery by extracting evidence from a huge amount of biomedical data and hence revolutionizes the entire pharmaceutical industry. In particular, graph representation learning techniques---a fast growing topic in the machine learning and data mining community focusing on deep learning for graph-structured data---has seen great opportunities for drug discovery as many data in the domain are graph-structured such as molecules and biomedical knowledge graphs. In this talk, I will introduce our recent progress on graph representation learning for drug discovery including: (1) molecule properties prediction; (2) de novo molecular design and optimization; and (3) retrosynthesis prediction.
Speaker Bio:	Jian Tang is currently an assistant professor at Mila-Quebec AI Institute and HEC Montreal (business school of University of Montreal), and also a Canada CIFAR AI Research Chair. His main research interests are graph representation learning, graph neural networks, deep generative models, knowledge graphs and drug discovery. He obtained his PhD from School of EECS, Peking University in 2014, was an associate researcher at Microsoft Research Asia in 2014-2016, and was a joint postdoc fellow at University of Michigan and Carnegie Mellon University. During his PhD, he was awarded with the best paper in one of the top three machine learning conferences—ICML2014; in 2016, he was nominated for the best paper award in the top data mining conference World Wide Web (WWW); in 2020, he is awarded with Amazon and Tencent Faculty Research Award. He is one of the most representative researchers in the growing field of graph representation learning and has published a set of representative works in this field such as LINE, LargeVis, and RotatE. His work LINE on node representation learning has been widely recognized and is the most cited paper at the WWW conference between 2015 and 2019. He is the program committee and area chair of many prestige conferences in machine learning and data mining including NeurIPS, ICML, ICLR, KDD, WWW, AAAI, IJCAI, etc.
Host:	Jian Pei

Quality-driven Analytics on Scientific Data

Speaker:	Angela Bonifati, Full Professor; Lyon 1 University
Time, Location:	11:30am (Pacific time), Apr 14, 2021 Online. [Recording] [Slides] [Poster]
Abstract:	One of the key principles of data analytics in data science pipelines is that the quality of the results is dependent on the quality of the input data analyzed. Data collection and acquisition in scientific domains, such as for instance in healthcare and life sciences domains, tend to often introduce errors into the data, such as violations of business rules, typos, missing values and replicated entries. Moreover, data collected for the patients' signals might exhibit peculiar features that need to be taken into account within the analytical and inference processes. I will present our latest results on enhancing the quality of querying and inference processes on scientific data and beyond. Among the others, we operate on real-life data of patients from several hospitals in the EU and provide the domain experts with useful data management and learning techniques that can help them with their diagnoses and analyses. First, inconsistency-aware annotations can enhance the data input to analytical processes. These annotations are further exploited during query processing in order to enhance the output of queries with inconsistency degrees. Second, feature-based similarities among time series corresponding to patients’ signals help to better identify groups of patients and to assess their risks. Third, logic-based declarative privacy-preserving data integration allows to migrate clinical data from one hospital to another while ensuring privacy guarantees. In all cases, our research aims at providing the caregivers with a better understanding of their clinical data thanks to the improved outcomes of the performed analytics. In the talk, I will overview the research performed in my group on this topic as well as the related work and the future directions of investigation.
Speaker Bio:	Angela Bonifati is a Professor of Computer Science at Lyon 1 University and affiliated with the CNRS Liris research lab since 2015. In 2019 and 2020, she has been on leave at INRIA. Prior to that, she was working as a Professor at Lille 1 University (2011-2015) and as a researcher at CNR, Italy since 2003. She received her Ph.D. from Politecnico di Milano in 2002 and right after she was a postdoctoral researcher at INRIA Rocquencourt in Paris for one year. Her current research interests are on the interplay of relational and graph-oriented data paradigms, particularly on data integration and Big Data curation for life sciences, query processing and learning for structured and unstructured data models. She is the Principal Investigator of the ANR research project QualiHealth: Enhancing the Quality of Healthcare Data (2018-2023) and was the recipient of the IDEX Palse Impulsion Individual grant in 2016. She is involved in several grants at Lyon 1 University, including a EU H2020 grant on data quality and data integration in a clinical data lake. She has also co-authored several publications in first-rate venues of the data management field along with two books (edited by Springer in 2011 and Morgan&Claypool in 2018) and an invited paper in ACM Sigmod Record 2018. She is the Program Chair of ACM Sigmod 2022 and was the Program Chair of EDBT 2020. She is Associate Editor of PVLDB in 2020-2022 and Associate Editor of IEEE ICDE 2021, 2018 and 2011. She is Associate Editor of the VLDB Journal, ACM TODS, Distributed and Parallel Databases and Frontiers in Big Data. She is the President of the EDBT Executive Committee and a member of the ICDT council. She holds many visiting scholar positions in foreign universities in both Europe and North America, the latest of which at the University of Waterloo (Canada) in 2019. She is currently Adjunct Professor in the latter university.
Host:	Jian Pei

Towards Democratizing Data Science

Speaker:	Carsten Binnig, Full Professor, TU Darmstadt; Adjunct Associate Professor, Brown University
Time, Location:	11:30am (Pacific time), Mar 24, 2021 Online. [Recording] [Slides] [Poster]
Abstract:	Technology has been the key enabler of the current Big Data movement. Without open-source tools like TensorFlow and Spark, as well as the advent of cheap, abundant computing and storage in the cloud, the trend toward datafication of almost every field in research and industry could never have happened. However, the current Big Data tool set is ill-suited for an efficient knowledge discovery by domain experts with only limited IT skills and thus represents a major bottleneck in our data-drivensociety. In this talk, I will present an overview of my current research efforts to revisit the Big Data stack from the user interface to the underlying hardware for making Big Data tools more efficient and easier to use for domain experts and thus enablethe democratization of data science.
Speaker Bio:	Carsten Binnig is a Full Professor in the Computer Science department at TU Darmstadt and an Adjunct Associate Professor in the Computer Science department at Brown University. Carsten received his PhD at the University of Heidelberg in 2008.Afterwards, he spent time as a postdoctoral researcher in the Systems Group at ETH Zurich and at SAP working on in-memory databases. Currently, his research focus is on the design of data management systems for modern hardware as well as modern workloads suchas interactive data exploration and machine learning. His work has been awarded with a Google Faculty Award, as well as multiple best paper and best demo awards for his research.
Host:	Jian Pei

The Rise of Intelligent Data Assistants

Speaker:	Georgia Koutrika, Research Director, Athena Research Center
Time, Location:	9:30am (Pacific time), April 8, 2021 Online. [Recording] [Poster]
Abstract:	For many, data is considered the 21st century’s most valuable commodity growing at an exponential rate – but is it for everyone? Analysts exploring data sets for insight, scientists looking for patterns, and consumers looking for information are just a few examples of user groups that need to access and dig into data. However, existing data exploration tools are falling behind in bridging the chasm between data and users, making data exploration intended only for the few. In this talk, we will discuss about what it takes to bridge this chasm and the new generation of intelligent data exploration tools that are emerging at the intersection of data management, natural language processing, machine learning and visualization. The talk will end with a summary of open questions on intelligent data assistants.
Speaker Bio:	Georgia Koutrika is a Research Director at Athena Research Center in Greece. She has more than 15 years of experience in multiple roles at HP Labs, IBM Almaden, and Stanford. She has received a PhD and a diploma in Computer Science from the Department of Informatics and Telecommunications, University of Athens, Greece. Her work focuses on data exploration, recommendations, and data analytics, and has been incorporated in commercial products, described in 14 granted patents and 26 patent applications in the US and worldwide, and published in more than 90 papers in top-tier conferences and journals. Georgia is an ACM Senior Member, IEEE Senior Member, and ACM Distinguished Speaker. She is the editor of ACM SIGMOD Blog. Furthermore, her academic activities include: Editor-in-chief for VLDB Journal, PC co-chair for VLDB 2023, associate editor for TKDE, SIGMOD2021 and VLDB2022, ICDE2021 sponsorship chair.
Host:	Jian Pei

From Fact-Checking to Journalistic Data Integration

Speaker:	Ioana Manolescu, Senior researcher, Inria; Part-time professor, Ecole Polytechnique
Time, Location:	11:30am (Pacific time), April 21, 2021 Online. [Recording] [Slides] [Poster]
Abstract:	Professional journalism is of utmost importance nowadays. It is a main feature distinguishing dictatorships from democracies, and a mirror sorely needed by society to look upon itself and understand its functioning. In turn, understanding is necessary for making informed decisions, such as political choices. With the world turning increasingly digital, journalists need to analyze very large amounts of data, while having no control over the structure, organization, and format of the data. Since 2013, my team has been working to understand data journalism and computational fact-checking use cases, to identify and develop tools adapted for this challenging setting. My talk will present highlights from several years of research in this area. First, I will describe efforts to improve the accessibility to reference data sources which serve as trusted background for verifying (fact-checking) statistical claims. Second, I will present ConnectionLens, a system for integrating very heterogeneous data sources as graphs, leveraging Information Extraction and Entity Disambiguation. Information needs on ConnectionLens graphs are answered through graph-based keyword search; I will discuss the characteristics of ConnectionLens keyword search which make it harder in our setting, and our solution to these issues. Finally, I will present an application we currently develop, in collaboration with Stéphane Horel, an investigative journalist from Le Monde.
Speaker Bio:	Ioana Manolescu is a senior researcher at Inria Saclay and a part-time professor at Ecole Polytechnique, France. She is the lead of the CEDAR INRIA team focusing on rich data analytics at cloud scale. She is also the scientific director of LabIA, a program ran by the French government whereas AI problems raised by branches of the local and national French public administration are tackled by French research teams. She is a member of the PVLDB Endowment Board of Trustees, and has been Associate Editor for PVLDB, president of the ACM SIGMOD PhD Award Committee, chair of the IEEE ICDE conference, and a program chair of EDBT, SSDBM, ICWE among others. She has co-authored more than 150 articles in international journals and conferences and co-authored books on "Web Data Management" and on "Cloud-based RDF Data Management". Her main research interests algebraic and storage optimizations for semistructured data, in particular Semantic Web graphs, novel data models and languages for complex data management, data models and algorithms for fact-checking and data journalism, a topic where she is collaborating with journalists from Le Monde. She is also a recipient of the ANR AI Chair titled "SourcesSay: Intelligent Analysis and Interconnexion of Heterogeneous Data in Digital Arenas" (2020-2023).
Host:	Jian Pei

Toward Understanding End-to-End Learning in the Context of Data: Machine Learning Dancing over Semirings and Codd's Table

Speaker:	Ce Zhang, Assistant Professor, ETH Zurich
Time, Location:	04:00pm (Pacific time), June 16, 2021 Online. [Recording] [Slides] [Poster]
Abstract:	Recent advances in machine learning (ML) systems have made it incredibly easier to train ML models given a training set. However, our understanding of the behavior of the model training process has not been improving at the same pace. Consequently, a number of key questions remain: How can we systematically assign importance or value to training data with respect to the utility of the trained models, may it be accuracy, fairness, or robustness? How does noise in the training data, either injected by noisy data acquisition processes or adversarial parties, have an impact on the trained models? How can we find the right data that can be cleaned and labeled to improve the utility of the trained models? Just when we start to understand these important questions for ML models in isolation recently, we now have to face the reality that most real-world ML applications are way more complex than a single ML model. In this talk, I will revisit these questions for an end-to-end ML pipeline, which consists of a noise model for data and a feature extraction pipeline, followed by the training of an ML model. In the first part of this talk, I will introduce some recent theoretical results in an abstract way: How to calculate the Shapley value of a training example for ML models trained over feature extractors, modeled as a polynomial in the provenance semiring? How to compute the entropy and expectation of ML models trained over data uncertainty, modeled as a Codd Table? As we will see, even these problems are #P-hard for general ML models, though, surprisingly, we can obtain PTIME algorithms for a simpler proxy model (namely a K-nearest neighbor classifier), for a large family of polynomials, input noise distributions, and utilities. I will then put these theoretical results into practice. Given a set of heuristics and a proxy model to approximate a realworld end-to-end ML pipeline into these abstract problems, I will provide a principled framework for three applications: (1) certifiable defence of backdoor attacks, (2) targeted data cleaning for ML, and (3) data valuation and debugging for end-to-end ML pipelines. I will describe both our positive empirical results but also those cases that our current approach failed at.
Speaker Bio:	Ce is an Assistant Professor in Computer Science at ETH Zurich. He believes that by making data—along with the processing of data—easily accessible to non-expert users, we have the potential to make the world a better place. His current research focuses on understanding and building next-generation machine learning systems and platforms. Before joining ETH, Ce was advised by Christopher Ré, finished his PhD roundtripping between the University of Wisconsin-Madison and Stanford University, and spent another year as a postdoctoral researcher at Stanford. He contributed to the research efforts that won the SIGMOD Best Paper Award and SIGMOD Research Highlight Award, and was featured in special issues, including in the Science magazine, the Communications of the ACM, “Best of VLDB”, and the Nature magazine. His work has also been reported by media such as the Atlantic, WIRED, and Quanta Magazine. More information about his group can be found at: ds3lab.org.
Host:	Jian Pei

HoloClean and Kamino: Structured Learning for Data Cleaning and Private Data Generation

Speaker:	Ihab Ilyas, Professor, University of Waterloo
Time, Location:	04:00pm (Pacific time), June 30, 2021 Online. [Poster]
Abstract:	Data scientists spend a big chunk of their time preparing, cleaning, and transforming raw data before getting the chance to feed this data to their well-crafted models. This massive labor-intensive exercises to clean data remain the main impediment to automatic end-to-end AI pipeline for data science. Real world scenarios are further complicated by data privacy and the need to protect sensitive user information. Data privatization efforts often compromise the utility of the data; for example, existing differentially private data synthesis methods usually fail in preserving one of the most fundamental data properties: the underlying correlations and dependencies among tuples and attributes (i.e., the structure of the data), compromising the utility of the output to downstream tasks. In this talk I describe our work in Kamino and HoloClean: systems for data cleaning and private data synthesis. HoloClean builds two main probabilistic models: a data generation model (describing how data was intended to look like); and a realization model (describing how errors might be introduced to the intended clean data). The framework uses few-shot learning, data augmentation, and self supervision to learn the parameters of these models, and use them to predict both error and their possible repairs. Kamino is a new system that learns the HoloClean models privately and use them to synthesize "useful" private data instances. Kamino shows that structured-preserving private data generation can be a powerful tool to help data scientists work with samples of sensitive data sets.
Speaker Bio:	Ihab Ilyas is a professor in the Cheriton School of Computer Science and the NSERC-Thomson Reuters Research Chair on data quality at the University of Waterloo. His main research focuses on the areas of big data and database systems, with special interest in data quality and integration, managing uncertain data, machine learning for data curation, and information extraction. Ihab is a co-founder of Tamr, a startup focusing on large-scale data integration, and he is also the co-founder of inductiv (acquired by Apple), a Waterloobased startup on using AI for structured data cleaning. He is a recipient of the Ontario Early Researcher Award, a Cheriton Faculty Fellowship, an NSERC Discovery Accelerator Award, and a Google Faculty Award, and he is an ACM Fellow.
Host:	Jian Pei

Toward Grounded Dialogue

Speaker:	Jian-Yun Nie, Professor, University of Montreal
Time, Location:	04:00pm (Pacific time), July 21, 2021 Online. [Recording] [Slides] [Poster]
Abstract:	Dialogue in natural language is a natural way for humans to interact with systems. Many methods have been developed for various types of dialogue, from chitchat to task-oriented dialogue in specific domains. In open-domain dialogue, most methods focused on generating or retrieving a natural response to a user’s utterance, with limited attention paid to the content of the response. In practice, to be useful, a system-generated response should provide the user with useful information, in addition to be natural. In this talk, we will describe several studies on grounded dialogue in which a response is generated or retrieved by considering the information provided in a document (document-grounded), the domain knowledge (knowledge-grounded), or the dialogue goal (goal-oriented). A special case is grounded question-answering, which aims at finding the right answer grounded in a document. These studies suggest that it is time to move toward dialogue systems capable of generating better grounded responses.
Speaker Bio:	Jian-Yun Nie is a professor at the Department of Computer Science and Operations Research, University of Montreal, and head of the RALI lab on Applied Research on Computational Linguistics. His research is in the areas of information retrieval and natural language processing. He has worked on various information retrieval models, web search, crosslanguage information retrieval, recommendation systems and deep learning models for query suggestion and dialogue. Jian-Yun Nie has published over 200 papers in the main journals and conferences in IR and NLP. He is on editorial board of 5 international journals (e.g. Journal of information retrieval), has served as chair, PC chair or area chair of the major conferences in these areas (SIGIR, CIKM, ACL, WWW, …). He received several best paper awards, including a Best paper award and a Test-of-Time honorable mention award from SIGIR.
Host:	Jian Pei

Nothing is for granted: Making wise decisions using real-time intelligence

Speaker:	Anastasia Ailamaki, Professor, the École Polytechnique Fédérale de Lausanne (EPFL)
Time, Location:	09:00am (Pacific time), August 12, 2021 Online. [Recording] [Slides] [Poster]
Abstract:	In today's ever-growing demand for fast, data-driven decisions, heterogeneity severely undermines performance and fragments efforts for building unified data exploration tools. The variety in data formats and workloads forces data pipelines to be manually split across a variety of task-specialized systems and combined through expensive ETL and orchestration processes, or to adapt both the data and the workloads to match the requirements of a single-system, sacrificing expressiveness and structural information. Furthermore, the ever-increasing hardware heterogeneity causes task-based specialization of the tools to specific hardware such as CPUs or GPUs, forcing a trade-off: designing optimized hardware often means wasting accelerator-level parallelism (ALP) opportunities or tolerating slow and unnecessary communication between devices. In general, data processing is adapted to the pre-determined data processing system architecture, losing valuable information in the translation. Real-time intelligence means to make all decisions during execution, when all relevant information is available for optimal utilisation of resources, while it also learns and extracts information about the query requests, instead of depending on pre-determined workload expectations. I will show how designing top-down the system architecture to allow a data- and workload-driven just-in-time specialization enables fast query execution over unprepared, potentially dirty data without time consuming preparation, as well as efficient orchestration and utilization of heterogeneous hardware devices.
Speaker Bio:	Anastasia Ailamaki is a Professor of Computer and Communication Sciences at the École Polytechnique Fédérale de Lausanne (EPFL) in Switzerland and the co-founder of RAW Labs SA, a Swiss company developing real-time analytics infrastructures for heterogeneous big data from multiple sources. She earned a Ph.D. in Computer Science from the University of Wisconsin-Madison in 2000. She received the 2019 ACM SIGMOD Edgar F. Codd Innovations and the 2020 VLDB Women in Database Research Award. She is also the recipient of an ERC Consolidator Award (2013), the Finmeccanica endowed chair from the Computer Science Department at Carnegie Mellon (2007), a European Young Investigator Award from the European Science Foundation (2007), an Alfred P. Sloan Research Fellowship (2005), an NSF CAREER award (2002), and ten best-paper awards in database, storage, and computer architecture conferences. She is an ACM fellow, an IEEE fellow, the Laureate for the 2018 Nemitsas Prize in Computer Science, and an elected member of the Swiss, the Belgian, the Greek, and the Cypriot National Research Councils. She is a member of the Academia Europaea and of the World Economic Forum Expert Network.
Host:	Jian Pei

Rethink e-Commerce Search

Speaker:	Haixun Wang, VP Engineering, Algorithms & Distinguished Scientist @ Instacart
Time, Location:	04:00pm (Pacific time), January 26, 2022 Online. [Recording] [Slides] [Poster]
Abstract:	The quality of the search experience on an e-commerce site plays a critical role in customer conversion and the growth of the e-commerce business. In this talk, I will discuss the current status and challenges of product search. In particular, I will highlight the significant amount of effort it takes to create a high-quality product search engine using classical information retrieval methods. Then, I will discuss how recent advances in NLP and deep learning, especially the advent of large pre-trained language models, may change the status quo. While embedding-based retrieval has the potential to improve classical information retrieval methods, creating a machine learning-based, end-to-end system for general-purpose, web search is still extremely difficult. Nevertheless, I will argue that product search for e-commerce may prove to be an area where deep learning can create the first disruption to classical information retrieval systems.
Speaker Bio:	Haixun Wang is currently an IEEE fellow, editor in chief of IEEE Data Engineering Bulletin, and a VP of Engineering and Distinguished Scientist at Instacart. Before Instacart, he was a VP of Engineering and Distinguished Scientist at WeWork, a Director of Natural Language Processing at Amazon, and he led the NLP team working on Query and Document Understanding at Facebook. From 2013 to 2015, he was with Google Research working on natural language processing. From 2009 to 2013, he led research in semantic search, graph data processing systems, and distributed query processing at Microsoft Research Asia. He had been a research staff member at IBM T. J. Watson Research Center from 2000 to 2009. He received the Ph.D. degree in Computer Science from the University of California, Los Angeles in 2000. He has published more than 150 research papers in referred international journals and conference proceedings.
Host:	Jian Pei

Towards an Integral System for Large-scale Graph Processing

Speaker:	Jingren Zhou, Senior Vice President @ Alibaba and Ant Group
Time, Location:	04:00pm (Pacific time), Feb 23, 2022 Online. [Poster]
Abstract:	An increasing number of business applications naturally model data as graphs to capture complex relationships and dynamic interactions among entities. Processing such big graphs typically involves a variety of graph computations over billions of vertices and edges. In this talk, I will first give an overview of the landscape of big graph processing and then present the design and implementation of GraphScope, a unified and open-source engine for big graph processing from Alibaba and Ant Group. GraphScope provides a unified programming interface to a wide range of graph computations such as graph traversal, pattern matching, iterative algorithms, and graph neural networks with a high-level programming language and supports parallel and distributed execution of sophisticated graph analysis on a cluster of machines. In addition, it provides an seamless integration of a highly optimized graph engine in a general-purpose data-parallel computing system. Finally, I will outline some challenges and future research directions for complex graph computations.
Speaker Bio:	Jingren Zhou is Senior Vice President at Alibaba and Ant Group. He is responsible for driving big data infrastructure and many key data-driven businesses at Alibaba and Ant Group. He has led work to develop advanced techniques for personalized search, product recommendation, and advertisement at Alibaba’s e-commerce platform and Alipay, a leading online payment and digital lifestyle platform. He also manages data analytics and intelligence research lab at Alibaba DAMO Academy. His research interests include cloud-computing, databases, and large-scale machine learning. He received his PhD in Computer Science from Columbia University. He is a Fellow of IEEE.
Host:	Jian Pei

Towards a Trusted Database System for Supporting Verifiability and Digital Trust

Speaker:	Beng Chin OOI, Lee Kong Chian Centennial Professor @ National University of Singapore
Time, Location:	04:00pm (Pacific time), March 30, 2022 Online. Registration [Poster]
Abstract:	The advancement of technologies has caused many organizations to re-strategize and adopt digital transformation. Along with it, transactions and interactions between organizations have been digitized. However, digitalization is creating the dependency of society on increasingly complex, interconnected systems and networks, which increases our exposure to various threats such as loss of data, loss of control over the software systems and networks, mistrust due to ill-intended contents and actions, misuse of technology and so on. To bootstrap trust and support secure data sharing, cross-institution transactions and business in the digital world, various technologies are needed to empower cross-institutional collaboration with increasing trust through secure communication, transfer and exchange of digital documents and assets, verifiability, and auditability. Trusted database technology is foundational to supporting digital trust. It is required to not only ensure privacy, integrity of data, but also of the computation on the data. The design space for trusted database systems is vast, and there is no one-size-fits-all solution. I shall discuss design and implementation issues and requirements of a flexible trusted database system and our works in this area.
Speaker Bio:	Beng Chin is Lee Kong Chian Centennial Professor and NGS faculty member at the National University of Singapore (NUS), an adjunct Chair Professor at Zhejiang University, China, and the director of NUS AI Innovation and Commercialization Centre at Suzhou, China. He obtained his BSc (1st Class Honors) and PhD from Monash University, Australia, in 1985 and 1989 respectively. He is a fellow of the ACM , IEEE, and Singapore National Academy of Science (SNAS). Beng Chin serves as a non-executive and independent director of ComfortDelgro, a transportation company, and AlDigi Holdings, a Fintech company. Beng Chin's research interests include database systems, distributed and blockchain systems, and large scale analytics, in the aspects of system architectures, performance issues, security, accuracy and correctness. He works closely with the industry (eg. NUHS, Jurong Health, Tan Tok Seng Hospital and Singapore General Hospital on healthcare analytics and prediabetes prevention), and exploits IT for efficiency in various application domains, including healthcare, finance and smart city. Beng Chin was the recipient of ACM SIGMOD 2009 Contributions award, a co-winner of the 2011 Singapore President's Science Award, the recipient of 2012 IEEE Computer Society Kanai award, 2013 NUS Outstanding Researcher Award, 2014 IEEE TCDE CSEE Impact Award, 2016 China Computer Federation (CCF) Overseas Outstanding Contributions Award, and 2020 ACM SIGMOD EF Codd Innovation Award. He was a recipient of VLDB'14 and VLDB’19 Best Paper awards, and SIGMOD 2020 Research Highlight award. Beng Chin has served as a PC member for international conferences such as ACM SIGMOD, VLDB, IEEE ICDE, WWW, and SIGKDD, and as Vice PC Chair for ICDE'00,04,06, PC co-Chair for SSD'93 and DASFAA'05, PC Chair for ACM SIGMOD'07, Core DB PC chair for VLDB'08, and PC co-Chair for IEEE ICDE'12, IEEE Big Data'15, BOSS'18, IEEE ICDE'18, Industry track of VLDB'19, BCDL'19 and ACM SoCC'20. He has served as the President of VLDB Endowment (2014-2017), and editor-in-chief of IEEE Transactions on Knowledge Engineering (2008-2012), Elsevier Journal of Big Data Research (2014-2016), and ACM/IMS Transactions on Data Science (2018-2021).
Host:	Jian Pei

Synthesizing Relational Data with Differential Privacy

Speaker:	Xiaokui Xiao, Associate Professor @ National University of Singapore
Time, Location:	04:00pm (Pacific time), April 13, 2022 Online. [Recording] [Slides] [Poster]
Abstract:	Providing access to sensitive data while preserving privacy is an important problem in the era of big data. A canonical solution to this problem is to replace the sensitive data with synthetic data that follow a similar distribution but do not reveal private information. In this talk, I will introduce our research efforts towards synthesizing relational data under differential privacy, which is a rigorous privacy framework that has gained mainstream adoption in both academia and industry. In particular, I will first present a method for synthesizing a single relational table, and then describe a more advanced technique for tackling the case of multiple tables with foreign key constraints. I will conclude the talk with directions for future work.
Speaker Bio:	Xiaokui Xiao is an associate professor at the School of Computing, National University of Singapore. His research focuses on data management, especially on data privacy and algorithms for large data. He received the best research paper award in VLDB 2021, and is a distinguished member of ACM. He obtained a PhD in Computer Science from the Chinese University of Hong Kong in 2008.
Host:	Jian Pei

A Systematic View of Data Science

Speaker:	M. Tamer Özsu, University Professor @ University of Waterloo
Time, Location:	04:00pm (Pacific time), April 27, 2022 Online. [Recording] [Slides] [Poster]
Abstract:	There is a data-driven revolution underway in science and society, disrupting every form of enterprise. We are collecting and storing data more rapidly than ever before. There is an increasing recognition that data science can assist in leveraging this data and the insights obtained from it into products, systems, and policies. This has resulted in the formation within academia of data science research centres, institutes and even academic units and the establishment of major initiatives within every major industrial organization. However, our understanding of data science is vague and highly varied and, in many cases, are squeezed to fit the available openings within an institution. There is a need to approach this field systematically to define its scope and its boundaries. The objective of this talk is to provide such a consistent and systematic study of the scoping of data science.
Speaker Bio:	M. Tamer Özsu is a University Professor at Cheriton School of Computer Science at University of Waterloo. Previously, he was the Director of the Cheriton School and Associate Dean (Research) of the Faculty of Mathematics. His research is on data engineering aspects of data science focusing on distributed data management and the management of non-conventional data. He is a Fellow of the Royal Society of Canada, American Association for the Advancement of Science, Association for Computing Machinery and Institute of Electrical and Electronics Engineers, an elected member of Science Academy, Turkey and a member of Sigma Xi. Dr. Özsu is the recipient of the IEEE Innovation in Societal Infrastructure Award (2022), CS-Can/Info-Can Lifetime Achievement Award (2018), ACM SIGMOD Test-of-Time Award (2015), the ACM SIGMOD Contributions Award (2006), and The Ohio State University College of Engineering Distinguished Alumnus Award (2008). He is the Founding Editor-in-Chief of ACM Books (2014-2020) and the Founding Series Editor of Synthesis Lectures on Data Management (2009-2014). He serves on the editorial boards of three journals and one book series. He co-edited with Ling Liu the Encyclopedia of Data Management.
Host:	Jian Pei

Next-Generation Intelligent Assistants for AR/VR Devices

Speaker:	Xin Luna Dong, Head Scientist @ Facebook AR/VR Assistant
Time, Location:	04:00pm (Pacific time), April 20, 2022 Online. [Recording] [Slides] [Poster]
Abstract:	An intelligent assistant is an agent that knows you and the world, can receive your requests or predict your needs, and provide you the right services at the right time with your permission. As smart devices such as Alexa, Google Home, Ray-ban Stories get popular, Intelligent Assistants are gradually playing an important role in people's lives. The Emergence of AR/VR devices bring more opportunities and call for the next generation of Intelligent Assistants. In this talk, we discuss the many challenges and opportunities we face to grow intelligent assistants from voice-only to multi-modal, from context-agnostic to context-aware, and from listening to the users' requests to predicting the user's needs. We hope these new challenges will open doors to new research areas.
Speaker Bio:	Xin Luna Dong is the Head Scientist at Facebook AR/VR Assistant. Prior to joining Facebook, she was a Senior Principal Scientist at Amazon, leading the efforts of constructing Amazon Product Knowledge Graph, and before that one of the major contributors to the Google Knowledge Vault project, and has led the Knowledge-based Trust project, which is called the “Google Truth Machine” by Washington’s Post. She has co-authored books "Machine Knowledge: Creation and Curation of Comprehensive Knowledge Bases" and “Big Data Integration”, was awarded ACM Distinguished Member, and VLDB Early Career Research Contribution Award for “Advancing the state of the art of knowledge fusion”. She serves in the VLDB endowment and PVLDB advisory committee, and is a PC co-chair for KDD'2022 ADS track, WSDM 2022, VLDB 2021, and Sigmod 2018.
Host:	Jian Pei

Knowledge Graphs 2022: Achievements, Challenges and Opportunities

Speaker:	Gerhard Weikum, Scientific Director @ Max Planck Institute for Informatics
Time, Location:	09:00am (Pacific time), May 25, 2022 Online. [Recording] [Slides] [Poster]
Abstract:	Machines with comprehensive knowledge of the world's entities and their relationships has been a long-standing vision and challenge of AI. Over the last two decades, huge knowledge bases, also known as knowledge graphs, have been automatically constructed from web data and text sources, and have become a key asset for search engines and other use cases. Machine knowledge can be harnessed to semantically interpret text in news, social media and web tables, contributing to question answering, natural language processing and data analytics. This talk reviews these advances and discusses lessons learned (see http://dx.doi.org/10.1561/1900000064 for a comprehensive survey). Moreover, the talk identifies open challenges and new research opportunities.
Speaker Bio:	Gerhard Weikum is a Scientific Director at the Max Planck Institute for Informatics in Saarbruecken, Germany, and an Adjunct Professor at Saarland University. He co-authored a comprehensive textbook on transactional systems, received the VLDB Test-of-Time Award 2002 for his work on automatic database tuning, and is one of the creators of the YAGO knowledge base which was recognized by the WWW Test-of-Time Award in 2018. Weikum is an ACM Fellow and elected member of various academies. He received the ACM SIGMOD Contributions Award in 2011, a Google Focused Research Award in 2011, an ERC Synergy Grant in 2014, the ACM SIGMOD Edgar F. Codd Innovations Award in 2016, and the Konrad Zuse Medal in 2021.
Host:	Jian Pei

Deep Data Integration

Speaker:	Wang-Chiew Tan, Research Scientist Manager @ Meta
Time, Location:	04:00pm (Pacific time), September 19, 2022 Online. [Recording] [Slides] [Poster]
Abstract:	We are witnessing the widespread adoption of deep learning techniques as avant-garde solutions to different computational problems in recent years. In data integration, the use of deep learning techniques has helped establish several state-of-the-art results in long standing problems, including information extraction, entity matching, data cleaning, and table understanding. In this talk, I will reflect on the strengths of deep learning and how that has helped move forward the needle in data integration. I will also discuss a few challenges associated with solutions based on deep learning techniques and describe some opportunities for the future work.
Speaker Bio:	Wang-Chiew is a research scientist manager at Meta AI. Before she was the Head of Research at Megagon Labs, where she led the research efforts on building advanced technologies to enhance search by experience. This included research on data integration, information extraction, text mining and summarization. Prior to joining Megagon Labs, she was a Professor of Computer Science at the University of California, Santa Cruz. She also spent two years at IBM Research - Almaden.
Host:	Tianzheng Wang

A Theory of Human k-shot Learning and Consciousness

Speaker:	Ming Li, Canada Research Chair in Bioinformatics and University Professor, @ University of Waterloo
Time, Location:	04:00pm (Pacific time), December 06, 2022 Online. [Recording] [Poster] This event is freely available to the public.
Abstract:	From a well-accepted law of thermodynamics, a theory of human-like k-shot learning is derived. The theory is implemented as a hierarchical VAE and VAE-based universal compressor model. Classical wisdom is that all learning paradigms compress data to smaller models at the training stage. Our theory indicates that nontrivial compression also happens in the inference stage. This is particularly important for few shot learning when training data is scarce. We prove that all other possible models for human or animal k-shot learning approximate our theory. The model achieves human level performance and significantly better performance than baseline models including deep learning models. This bridges the gap between our common-sense few-sample human learning and large-data machine learning, and can be easily applied in practice. (Joint work with Zhiying Jiang)
Speaker Bio:	Ming Li is a Canada Research Chair in Bioinformatics and a University Professor at the University of Waterloo. He is a fellow of Royal Society of Canada, ACM, and IEEE. He is a recipient of Canada's E.W.R. Steacie Fellowship Award in 1996, the 2001 Killam Fellowship and the 2010's Killam Prize. Together with Paul Vitanyi, they co-authored the book "An introduction to Kolmogorov complexity and its applications."
Host:	Tianzheng Wang