Invited Industrial Talks

1. Title: Applications of Graph Mining in Public Security

Speaker: Dr. Kaibo Xu, Mininglamp Technology

Abstract: Graph mining is becoming a pervasive technology in activities such as using structural analysis to find criminal gang, looking for patterns to discover illegal activities, and analyzing trajectory sequences to predict security risks. In this talk, we will present a collection of practical efforts made by Mininglamp Technology on the use of graph mining technologies in the sector of public security. After introducing a few related industrial cases, we will demonstrate our solutions to some key problems and shed some lights on applications of graph mining technologies in other possible industries.

Bio: Dr. Kaibo Xu received his Bachelor degree (1998) in Computer Science from Beijing University of Chemical Technology and his Master (2005) and PhD (2010) in Computer Science from the University of the West of Scotland. He worked as a Teaching Assistant (1998-2004), Lecturer (2004-2009), and Associate Professor (2009-2017) at Beijing Union University. He has supervised more than 20 master and doctoral students who are successful in their academic and industrial careers. As the principal investigator, he has received 7 governmental funds and 5 industrial funds with a total amount of 5M in the Chinese yuan. Dr. Kaibo Xu has also consulted extensively and been involved in many industrial projects. He worked as the Chief-Information-Officer (CIO) of Yunbai Clothing Retail Group, China (2016-2019). Currently, he is serving as Vice president and Principal Scientist of Mininglamp Technology. His research interests include graph mining, knowledge graph and knowledge reasoning.

2. Title: AI for Transportation

Speaker: Dr. Jieping Ye, Didi Chuxing

Abstract: Didi Chuxing is the world’s leading mobile transportation platform that offers a full range of app-based transportation options for 550 million users. Every day, DiDi’s platform receives over 100TB new data, processes more than 40 billion routing requests, and acquires over 15 billion location points. In this talk, Dr. Jieping Ye will show how AI technologies have been applied to analyze such big transportation data to improve the travel experience for millions of users.

Bio: Dr. Jieping Ye is head of Didi AI Labs and a VP of Didi Chuxing. He is also a professor of University of Michigan, Ann Arbor. His research interests include big data, machine learning, and data mining with applications in transportation and biomedicine. He has served as a Senior Program Committee/Area Chair/Program Committee Vice Chair of many conferences including NIPS, ICML, KDD, IJCAI, ICDM, and SDM. He has served as an Associate Editor of Data Mining and Knowledge Discovery, IEEE Transactions on Knowledge and Data Engineering, and IEEE Transactions on Pattern Analysis and Machine Intelligence. He won the NSF CAREER Award in 2010. His papers have been selected for the outstanding student paper at ICML in 2004, the KDD best research paper runner up in 2013, and the KDD best student paper award in 2014.

3. Title: Industrializing AI with Baidu AutoDL

Speaker: Dr. Haoyi Xiong, Baidu

Abstract: Tremendous new business opportunities and market revenues have been brought by the incorporation of deep learning (DL) techniques in the modern data-intensive intelligent information systems. Designing, implementing, and deploying DL models nowadays becomes the key practices to adopt AI in industry. With her mission to simplify the world with advanced technologies, Baidu offers AutoDL a comprehensive set of automated deep learning techniques (e.g., developing frameworks with advanced algorithms through cloud services) that lowers the technical burden of enabling AI in almost full spectrum of industrial scenarios. In this talk, we briefly cover our research efforts in four key areas of  AutoDL, including AutoDL Design, AutoDL Transfer, AutoDL Federation, and AutoDL Edge, where we would like to share our experiences of using Baidu AutoDL to automatically design the architectures for deep neural networks, automatically source the pre-trained models and finetune the knowledge transfer knowledge from them, automatically discover partners’ on the blockchain and collaborate with them for federated learning, and automatically train and deploy DL models on the resource-constrained edge devices. During this talk, you are expected to understand the advanced technologies used by Baidu AutoDL and take the hands-on experience home.

Bio: Dr. Haoyi Xiong is currently a Tech Lead and Senior Staff Research Scientist of AI at Beijing Big Data Lab, Baidu Inc. Prior to his industrial research endeavor, he was a tenure-track Assistant Professor at Department of Computer Science, Missouri University of Science and Technology, Rolla Missouri, United States. Before joining Missouri S&T, he was a Research Associate at Department of Systems and Information Engineering, University of Virginia. He received his PhD in Computer Science from Institut Mines-Telecom and University of Paris 6, France. His research interests lay on the intersection between pervasive computing systems and AI. He published intensively in top computer science conferences and journals, such as ICLR, UbiComp, RTSS, IJCAI, AAAI, PerCom, ICDM, and IEEE/ACM Transactions. He has supervised several PhD students and Postdocs, some of them have already secured faculty positions in leading United State universities and researcher positions in top AI companies.

4. Title: Research on Practical Data Management & Business Intelligence Solutions for Online-to-Offline Fusion

Speaker: Dr. Jie Cao, Nanjing University of Finance and Economics, Nanjing, China

Abstract: Online-to-Offline (O2O) Fusion is a new-emerging paradigm that will promote the further transformation of e-business. It is also meaningful for both online enterprise and retail enterprise to break through the bottlenecks of traditional business models that they have suffered. Currently, O2O research mainly focuses on discussing paradigm or strategies itself. Nevertheless, how to release the intelligence of O2O fusion remains a pretty much open issue. Although the penetration of big data computing technologies can provide significant supports to O2O intelligent business decision, there are still a multitude of challenges that are worth exploring, including models are not perfect enough to support mining heterogeneous data, fragmented knowledge is difficult to be effectively fused, and the storage and computing infrastructure is not clear enough for business big data. In light of this, we organize a multidisciplinary team from the fields of data mining, database, behavior analysis, and business intelligence to explore several crucial issues within O2O big data fusion and its applications. In particular, we will first build a big data mining framework for O2O intelligence, meeting the challenges rising from data decentralization and heterogeneity. We then focus on designing novel mining and fusion methods for big trajectory data and fragmented knowledge. Last but not least, we will design and develop a real platform for O2O big data mining and construct several demonstration applications.

Bio: Jie Cao received the Ph.D. degree in manufacturing automation from Southeast University, Nanjing, China, in 2002. His current research interests include cloud computing, business intelligence, and data mining. He has published prolifically in refereed journals and conference proceedings including KDD, AAAI, and TKDE etc.. Dr. Cao has been selected in the Program for New Century Excellent Talents in University and awarded with Young and Mid-Aged Expert with Outstanding Contribution in Jiangsu province.

5. Title: Democratizing AI: Back-fitting end-to-end machine learning at LinkedIn at scale!

Speaker: Dr. Zi Li

Abstract: For the past decade, LinkedIn has embraced AI across our product lines—from anti-abuse anomaly detection to career recommendations to feed curation, we use AI in a variety of ways that improve the experience for members and customers. However, this approach doesn’t always scale efficiently. Each AI stack was built by separate teams, with little sharing between them. Additionally, custom workflows add complexity when onboarding new engineers, new features, and new modeling technologies. In recent years, we became acutely aware that these systems make it difficult for non-AI engineers to build, train, and run their own models. In August 2017, we began a new program at LinkedIn called “Productive Machine Learning” (“Pro-ML” for short). The goal of Pro-ML is to double the effectiveness of machine learning engineers while simultaneously opening the tools for AI and modeling to engineers from across the LinkedIn stack. In this post, we’ll talk about how LinkedIn scales AI and ML systems so that more engineers can take advantage of these techniques.

Bio: Zi Li is manager of machine learning at LinkedIn China. He has joint LinkedIn for over 6 years. His team is responsible for building ML algorithms and systems to empower various LinkedIn products in China. His team mainly focuses on recommendation systems, data standardization and Bayesian smoothing based inference. Before joining LinkedIn China, he worked at LinkedIn US and focused on job recommendation. Zi holds a PhD in Computer Science from Iowa State University and BS in Computer Science from Beijing Institute of Technology.

6. Title: Federated Learning: Connecting the Dots, Empowering AI Landing

Speaker: Dr. Xiaolin Li

Abstract: As the data privacy and regulation becomes critical, e.g., CDPR (China Data Protection Regulations) and GDPR (General Data Protection Regulation), it calls for a paradigm shift in big data and AI applications. In the meantime, the current Artificial Intelligence deeply depends on the diverse and broad datasets. This talk will introduce our federated learning framework (iBond) to connect data islands, protect data privacy, and empower AI landing in finance, business, health, and education applications.

Bio: Dr. Xiaolin Andy Li is Head of Tongdun AI Institute and VP of Tongdun Technology. AI Institute is missioned to pioneer in fundamental research and empower a broad spectrum of AI applications. AI Institute is composed of divisions of deep learning (machine learning, reinforcement learning, federated learning), computer vision, natural language processing, intelligent interaction (intelligent speech), and AI platforms and operating systems. He is also a Professor and University Term Professor in the Department of Electrical & Computer Engineering and Department of Computer & Information Science & Engineering at the University of Florida. As the founding director, he founded National Science Foundation Center for Big Learning, the first national center on deep learning in USA, along with UF as the lead, CMU, U. Oregon, and UMKC, supported by NSF and over 30 industry members. He has published over 150 peer-reviewed papers in journals and conference proceedings, 5 books, and 4 patents. He received a PhD degree in Computer Engineering from Rutgers University. He is a recipient of the National Science Foundation CAREER Award and several best paper awards (IEEE SECON 2016, IEEE ICMLA 2016, ACM CAC 2013 and IEEE UbiSafe 2007).