W01: 1st International Workshop on Continual Semi-Supervised LearningAug 19 10:00 – 16:00 Montreal Time (UTC-4)
Aug 20 10:00 – 16:00 Montreal Time (UTC-4)
Fabio Cuzzolin, Irina Rish, Kevin Cannons, Vincenzo Lomonaco, Mohamad Asiful Hossain, Salman Khan, Ajmal Shahbaz
Whereas continual learning has recently attracted much attention in the machine learning community, the focus has been mainly on preventing the model updated in the light of new data from ‘catastrophically forgetting’ its initial knowledge and abilities. This, however, is in stark contrast with common real-world situations in which an initial model is trained using limited data, only to be later deployed without any additional supervision. In these scenarios the goal is for the model to be incrementally updated using the new (unlabelled) data, in order to adapt to a target domain continually shifting over time. The aim of this workshop is to formalise this new continual semi-supervised learning (CSSL) paradigm, and to introduce it to the machine learning community in order to mobilise effort in this direction. We present two new benchmark datasets for this problem and propose a number of challenges to the research community.
W02: Long-Tailed Distribution Learning (LTDL)Aug 20 20:00 – 2:00 (Next day) Montreal Time (UTC-4)
Xiu-Shen Wei, Peng Wang,Yu-Feng Li, Rui Xia, Jian Yang
As a commonly seen and natural data distribution in various real-world applications, the long-tailed distribution adversely affects the performance of learning-based methods in Artificial Intelligence (AI) at various levels, including the input, intermediate or mid-level stages of the processing or the objectives to be optimized in a multi-task setting, etc. Our Long-Tailed Distribution Learning (LTDL) workshop focuses on the interesting and challenging long-tailed distribution problems in the broad Artificial Intelligence area, which consists of a paper submission session and an invited talk session. Specifically, in the paper submission session, we will peer-review paper submissions involving the LTDL related topics as listed in https://ltdl-ijcai21.github.io/. Moreover, we will invite several domain-specific experts in Computer Vision (CV), Pattern Recognition (PR), Data Mining (DM), and Natural Language Processing (NLP) for sharing their insights and research progress on the topic of LTDL.
W03: Weakly Supervised Representation Learning WorkshopAug 21 20:00 – 2:00 (Next day) Montreal Time (UTC-4)
Bo Han, Tongliang Liu, Quanming Yao, Mingming Gong, Chen Gong, Gang Niu, Ivor W. Tsang, Masashi Sugiyama
Modern machine learning is migrating to the era of complex models (e.g., deep neural networks), which emphasizes the data representation highly. This learning paradigm is known as representation learning. Specifically, via deep neural networks, learned representations often result in much better performance than can be obtained with hand-designed representations. It is noted that representation learning normally requires a plethora of well-annotated data. Giant companies have enough money to collect well-annotated data. Nonetheless, for startups or non-profit organizations, such data is barely acquirable due to the cost of labeling data or the intrinsic scarcity in the given domain. These practical issues motivate us to research and pay attention to weakly supervised representation learning (WSRL), since WSRL does not require such a huge amount of annotated data. We define WSRL as the collection of representation learning problem settings and algorithms that share the same goals as supervised representation learning but can only access to less supervised information than supervised representation learning. In this workshop, we discuss both theoretical and applied aspects of WSRL.
W04: Artificial Intelligence for EducationAug 20 10:00 – 18:30 Montreal Time (UTC-4)
Zitao Liu, Richard Tong, Jiliang Tang, Xiangen Hu, Hang Li
Recent years have witnessed growing efforts from AI research community devoted to advancing our education and promising results have been obtained in solving various critical problems in education. However, developing and applying AI technologies to educational practice is fraught with its unique challenges, including, but not limited to, extreme data sparsity, lack of labeled data, and privacy issues. In this workshop, we will focus on introducing research progress on applying AI to education and discussing recent advances of handling challenges encountered in AI educational practice.
W05: 3rd Workshop on Financial Technology and Natural Language ProcessingAug 21 10:00 – 16:00 Montreal Time (UTC-4)
Aug 21 20:00 – 2:00 (Next day) Montreal Time (UTC-4)
Hsin-His Chen, Hiroya Takamura, Hen-Hsen Huang, Chung-Chi Chen
FinTech is an emerging and popular topic in both financial and engineering domains, and the technology of textual information analysis is highly related to the scope of IJCAI. The third workshop on Financial Technology and Natural Language Processing (FinNLP 2021) provides a forum where international participants share knowledge on applying NLP to the FinTech domain. Topics of interest include but not limited to text-based market provisioning, NLP-based investment management, crowdfunding analysis with textual data, text-oriented customer preference analysis, insurance application with textual information, NLP-based know your customer (KYC) approaches, and applications or systems for FinTech with NLP methods. Besides the regular paper submission, two shared-tasks, FinSim-2 and FinSBD-3, are co-located. FinSim-2 focuses on the evaluation of semantic representations by assessing the quality of the automatic classification of a given list of carefully selected terms from the Financial domain against a domain ontology. FinSBD-3 aims to extract sentences, lists, and list items from financial documents. The Proceedings of FinNLP-2021 will be published at ACL Anthology.
W06: Reinforcement Learning for Intelligent Transportation Systems WorkshopAug 21 10:00 – 16:00 Montreal Time (UTC-4)
Aug 21 20:00 – 2:00 (Next day) Montreal Time (UTC-4)
Zhiwei (Tony) Qin, Liam Paull, Rui Song, Hongtu Zhu
RL for ITS is an emerging interdisciplinary area that is undergoing rapid development and still require further understanding on a number of issues, e.g., dynamic environments, run-time performance, and safe deployment. The main goal of this workshop is to bring together researchers and practitioners from both the RL and transportation communities to answer the following questions: • What are the recent advances in the state-of-the-art of RL for ITS? • What are the new application areas in transportation that can be benefited by RL? • What are the successful stories in real-world applications of RL to ITS? • What are the challenges and critical issues that may have significant impact on a successful application? Answering these questions is critical to unleash the potential of the core intelligence in ITS beyond classical machine learning. We hope the participants will leave the workshop with a better idea about the current status of RL for ITS, about what is working well and what is not, thus providing guidance on directions of future research in this exciting area.
W07: IJCAI-21 Workshop on Applied Semantics Extraction and Analytics (ASEA)Aug 21 10:00 – 16:00 Montreal Time (UTC-4)
Kamalakar Karlapalem, Craig Knoblock, Mausam, Sameena Shah
Semantics has a vital role in an AI system. In building AI systems for real-world applications, semantics are often extracted from the data and managed through knowledge bases. The problems are determining what these semantics are, how to extract the relevant semantics, which representation format to use, and applying these semantics to run an AI system. Although there is research in extracting semantics from databases, tables, documents, images, videos, and even multi-modal data, the work has often been conducted in independent research silos without emphasizing the overarching focus on semantics and AI systems. This workshop aims to bring together researchers working on various streams related to semantics to study and develop the broader research area of semantics extraction and analytics and their applications. The workshop’s program will include invited talks, spotlight paper presentations, and lightning poster presentations to showcase research opportunities, novel solutions and systems, success stories, and future directions.
W08: 9th Workshop “What can FCA do for Artificial Intelligence”Aug 21 7:00 – 13:00 Montreal Time (UTC-4)
Sergei O. Kuznetsov, Amedeo Napoli, Sebastian Rudolph
This Ninth edition of the FCA4AI workshop (http://www.fca4ai.hse.ru/2021/) will be co-located with the IJCAI 2021 Conference to be held in Montréal, 21st-26th August, 2021. The objective of the FCA4AI workshop is to foster the interactions between AI and Formal Concept Analysis (FCA). Actually FCA is a mathematically well-founded theory aimed at classification and knowledge discovery that can be used for many purposes in AI. The preceding editions of the workshop have shown that many interactions and bridges exist between AI and FCA researchers. Accordingly, the objective of the 9th FCA4AI workshop is to investigate important issues, e.g. (i) how can FCA support various AI activities such as knowledge discovery, knowledge engineering, machine learning, data mining, information retrieval, recommendation…, (ii) how can FCA be extended in order to help AI researchers to solve new and complex problems in their domains, e.g. hybrid knowledge discovery, and how FCA can play a role in current trends such as explainable AI, assessing fairness of algorithms in decision making, and study of causality, among others.
W09: AI for Sports Analytics (AISA)Aug 17 8:00 – 14:00 Montreal Time (UTC-4)
Shayegan Omidshafiei, Karl Tuyls, Jesse Davis, Jan Van Haaren, Ian Graham, Jackson Broshear, Daniel Hennes, Jerome Connor, Zhe Wang, Adria Recasens, Romuald Elie, Praneet Dutta
The past decade has seen a tremendous growth of interest in sports analytics, not only from an economic and commercial perspective, but also from a purely scientific one. In fact, there has been a growing number of papers published at top artificial intelligence (AI) and machine learning venues that revolve around problems in sports, as well as multiple scientific events organized on the topic. Similar to other downstream domains that have benefited from advances of AI and machine learning, this growth in interest is due to important technological improvements in data collection and processing capabilities, progress in statistical learning and in particular deep learning, increased computational resources, and ever-growing economic activities associated with sports and culture (e.g., emergent consultancy ventures revolving around sports data collection and statistics). The purpose of this workshop is to bring together AI researchers, sports data scientists, and other stakeholders who are interested in the intersection of AI and sports. In particular, we are interested in attracting work that focuses on the overlap of multiple areas of AI such as machine learning, game theory, and computer vision and their application to the sports analytics domain. We also plan to have a strong list of invited speakers, participant interaction, and panel discussions.
W10: 7th International Workshop on Mining Actionable Insights from Social Networks, Special Edition on Responsible Social Media Mining (MAISoN’21)Aug 21 10:00 – 16:00 Montreal Time (UTC-4)
Aug 21 20:00 – 2:00 (Next day) Montreal Time (UTC-4)
Eduardo Hargreaves, Huan Liu, Cristina Miguel, Zeinab Noorian
The wide adoption of social media presents an interesting opportunity for performing data mining and knowledge discovery in a real-world context. However, the observations and research findings made with social media data can only provide unprecedented insights into human behavior if social data provides a representative description of human activity. Recent empirical evidence on US 2016 election or 2016 Brexit prediction outcomes has warned against biases and inaccuracies occurring not only at the source of the data and during the processing but also under the influence of those that design, build, maintain or use the analytical models, embedding and amplifying their biases. The recognition of different sorts of biases in the data analytics pipeline and their consequences open new research venues related to bot detection, neutralizing content pollution, noise removal from social media data, needs for innovative evaluation methods as well as ethical boundaries around the use of social data. The goal of this workshop is to contribute a practical perspective to the body of research that aims to quantify biases, to devise better methods for prototypical processing pipeline for social data, to evaluate such methods in context and to develop actionable frameworks on different dimensions of responsible social analytics aiming to mitigate the potential issues associated with many sorts of biases.
W11: International Workshop on Artificial Intelligence for Social Good (AI4SG)Aug 21 10:00 – 16:00 Montreal Time (UTC-4)
Aug 21 20:00 – 2:00 (Next day) Montreal Time (UTC-4)
Amulya Yadav, Haipeng Chen, Sasha Luccioni, Thanh Hong Nguyen
This workshop will explore how AI research can contribute to solving challenging problems faced by current-day societies. For example, what role can AI research play in promoting health, sustainable development and infrastructure security? How can AI initiatives be used to achieve consensus among a set of negotiating self-interested entities (e.g., finding resolutions to trade talks between countries)? To address such questions, this workshop will bring together researchers and practitioners across different strands of AI research and a wide range of important real-world application domains. The objective is to share the current state of research and practice, explore directions for future work, and create opportunities for collaboration. In addition, the workshop will place a special emphasis on highlighting AI approaches for tackling the COVID-19 pandemic.
W12: 7th Linguistic and Cognitive Approaches to Dialog Agents (LaCATODA 2021)Aug 20 20:00 – 2:00 (Next day) Montreal Time (UTC-4)
Rafal Rzepka, Jordi Vallverdú, Andre Wlodarczyk, Michal Ptaszynski, Pawel Dybala
Creating human-like dialog systems have been for decades one of the main goals of AI, but we are still far from achieving this goal for many reasons. LaCATODA series brings together researchers and entrepreneurs working on various facets of understanding, inviting not only computer scientists and engineers but also researchers working on cognition and language to present their work (including work in progress and position papers). Except traditional dialog processing we are interested in common sense knowledge and reasoning, emotions, moral aspects (like detecting hate speech or biases), etc. We are looking forward to discussing social, psychological and philosophical aspects of dialog systems, their problems and ideas for solving these problems. For that reason, also papers analyzing existing systems and describing their shortcomings are welcome.
W13: Generalization in PlanningAug 19 10:00 – 16:00 Montreal Time (UTC-4)
Javier Segovia-Aguas, Siddharth Srivastava, Guillem Francès, Blai Bonet
Humans are good at solving sequential decision-making problems, generalizing from few examples, and learning skills that can be transferred to the solution of unseen problems. These problems remain long standing open problems for Artificial Intelligence (AI). Over the last two decades, there has been remarkable progress in the performance of automated planning systems to solve decision making problems by including novel search techniques and heuristics. However, real-world scalability and skill/plan generalization for complex, long-horizon tasks still remains an open challenge for current AI algorithms. This workshop aims to build synergies across different AI communities in order to address all aspects of generalization of solutions for sequential decision making including, but not limited to, representation of problems and solution concepts that enable efficient generalization and transfer of relevant knowledge, and algorithms for learning or synthesizing such generalized knowledge and solutions. We welcome contributions focusing on different formulations/representations for generalization, empirically validated methods, and theoretical analyses and foundations for generalization.
W14: AIMA4Edu: AI-based Multimodal Analytics for Understanding Human Learning in Real-World Educational ContextsAug 21 10:00 – 16:00 Montreal Time (UTC-4)
Richard Tong, Edgar Kalns, Yiqiang Chen, Feiyue Wang
Human learning is a complex interactive and iterative process that takes place at a very fine-grained level. However, our ability to understand and support this fascinating latent learning process is often limited by what we can perceive and how we can measure. Recent advances in sensing technology and accompanying techniques for processing multimodal data give us a new opportunity to look at this classical problem with a new pair of lenses. We are particularly interested in those data gathered from the real-world educational activities versus those from the controlled lab environment. In this 3rd annual convening of the AIMA4EDU workshop, we expand topic areas to include nascent methodological areas and work beginning to apply multimodal data and AI to support learners. We will continue to explore cross-field experience from healthcare, HCI, neural science, etc. A multimodal dataset collected by Squirrel AI Learning is published on the workshop’s website, and we also encourage attendants to share more data to the community for research purposes only.
W15: AI and Product DesignAug 19 10:00 – 16:00 Montreal Time (UTC-4)
Oliver Niggemann, Frank Mantwill, Cesare Fantuzzi, Alexander Feldman, Johan de Kleer
Designs have to fulfill several, often contradicting requirements: First, a product must take technical considerations such as production constraints, costs, physical properties and material characteristics and availability into consideration. Second, aspects such as creativity, novelty and user acceptance play a crucial role. The resolution of contradictions is accompanied by a high degree of uncertainty, which results from different unknown dependencies and probabilities. Artificial Intelligence (AI) has the potential to contribute to these challenges. Current advances in AI such as deep learning or symbolic-neuronal architectures have the potential to support the design engineer in the entire product development process. The utilization of AI methods can not only help relieving design engineers from tedious work in routine design, it can also lay the foundation for innovative or creative design, as well as significantly shorten development times. Furthermore, existing information and historical data from previous design projects represent a possible potential for future smart assistance systems.
W16: Artificial Intelligence for Autonomous DrivingAug 20 10:50 – 15:00 Montreal Time (UTC-4)
Aug 20 16:00 – 21:40 Montreal Time (UTC-4)
Xinshuo Weng, Ye Yuan, Daniel Omeiza, Shangxuan Wu, Kris Kitani
Autonomous driving provides a rich source of high-impact research problems for the broad artificial intelligence community across different fields such as computer vision, machine learning, robotics, language and speech, civil engineering, human-computer interaction, environmental science, and neuroscience. Further, full self-driving capability (“Level 5”) is far from solved and extremely complex, beyond the capability of any one institution or company, necessitating larger-scale communication and collaboration between researchers in different fields. Beyond the research community, autonomous driving research also has significant social impacts such as reducing road accidents; giving independence to those unable to drive; reducing emissions to protect the environment and alleviate climate change; reducing traffic congestion, and improving the efficiency of urban transportation. The goal of this workshop is to embrace interdisciplinary knowledge in different fields of AI, from both academia and industry, to discuss how different fields can contribute to self-driving technology altogether and increase its social impact. This will be the 1st IJCAI workshop in this series.
W17: Robust and Reliable Autonomy in the Wild (R2AW)Aug 19 10:00 – 16:00 Montreal Time (UTC-4)
Nick Hawes, Ece Kamar, Bruno Lacerda, Sandhya Saisubramanian, Shlomo Zilberstein
The rapid development and deployment of autonomous systems in the open world has highlighted the need for advances in the theory and practice of robust and reliable AI. Autonomous cars, drone delivery services, health care robots, and emergency response systems are examples in which system decisions directly impact human lives. This workshop aims to bring together researchers from academia and industry to discuss the challenges involved in deployed autonomous systems, particularly systems that operate in the presence of minor perturbations or noise in the environment, and under model imprecision and uncertainty. The workshop will offer a forum for researchers to discuss progress and knowledge gaps in robust and reliable decision making, drawing on a wide range of methods, including automated planning, reinforcement learning, decision-theory, formal verification, multi-agent systems, game theory, robotics, AI ethics, and human-centered AI. We welcome the participation of researchers from different disciplines representing different perspectives on the topic.
W18: 3rd International Workshop on Deep Learning for Human Activity RecognitionAug 19 20:00 – 2:00 (Next day) Montreal Time (UTC-4)
Zhenghua Chen, Wu Min, Jianfei Yang, Xiaoli Li
Human activity recognition (HAR) can be used for a number of applications, such as health-care services and smart home applications. Recently, deep learning has achieved great success in many challenging research areas, such as image recognition and natural language processing. The key merit of deep learning is to automatically learn representative features from massive data. This technology can be a good candidate for human activity recognition. Some initial attempts can be found in the literature. However, many challenging research problems in terms of accuracy, device heterogeneous, environment changes, etc. remain unsolved. This workshop intends to prompt state-of-the-art approaches on deep learning for human activity recognition.
W19: Is Neuro-Symbolic SOTA still a myth for Natural Language Inference?Aug 21 10:00 – 16:00 Montreal Time (UTC-4)
Aug 21 20:00 – 2:00 (Next day) Montreal Time (UTC-4)
Somak Aditya, Maria D. Chang, Swarat Chaudhuri, Monojit Choudhury, Sebastijan Dumančić
The Natural Language Inferencing (NLI) task has been central to track progress towards understanding natural language. Current state-of-the-art Deep Learning-based NLI methods achieve very high accuracy, but have been shown to generalize poorly, often in simpler examples — indicating lack of reasoning capabilities. On the other hand, neuro-symbolic methods can do robust reasoning, but extensions of such methods are hard for tasks with a large range of linguistic variability. Slightly independently, the deep learning community has made progress towards Automated Theorem Proving and Program Synthesis, which are also relevant to NLI. Hence, to solve this “paradox of choice”, we revisit the potential and practicality of Neuro-symbolic methods for the NLI task. We want to take specific reasoning dimensions, informed by Logic, Knowledge Representation and Reasoning, and Linguistics literature; and track the progress of both Neural and Neuro-Symbolic efforts. We also aim to bring together researchers from Deep Learning, Natural Language Processing, Symbolic Logic, and Program Synthesis in the same platform.
W20: Artificial Intelligence for Anomalies and Novelties (AI4AN)Aug 20 20:00 – 2:00 (Next day) Montreal Time (UTC-4)
Aug 21 10:00 – 16:00 Montreal Time (UTC-4)
Guansong Pang, Jundong Li, Anton van den Hengel, Longbing Cao, Thomas G. Dietterich
Anomalies are referred to as observations or events that are rare or significantly different from the majority of observations we have in hand, while novelties are observations from novel classes that were unseen during learning. Recognition, detection and accommodation to anomalies and novelties are active research areas in multiple communities, including data mining, machine learning, and computer vision. The successful early detection of anomalies and novelties is of great significance across many domains. Specialized techniques have been studied in some of these areas for decades, but recent developments are raising a wide variety of new research questions and challenges in deep anomaly detection, fundamental theories of novelty and abnormality, characterization and accommodation of anomalies, etc. This workshop will gather researchers and practitioners from diverse communities and knowledge background to promote the development of fundamental theories, effective algorithms, and novel applications of anomaly and novelty detection, characterization, and adaptation.
W21: AIofAI: 1st Workshop on Adverse Impacts and Collateral Effects of Artificial Intelligence TechnologiesAug 19 10:00 – 17:30 Montreal Time (UTC-4)
Esma Aïmeur, Nicolas Diaz Ferreyra, Hicham Hage
The role of Artificial Intelligence (AI) in people’s everyday life has grown exponentially over the last decade. Currently, individuals rely heavily on intelligent software applications across different domains including healthcare, logistics, defense, and governance. Particularly, AI systems facilitate decision-making processes across these domains through the automatic analysis and classification of large data sets and the subsequent identification of relevant patterns. To a large extent, such an approach has contributed to the sustainable development of modern societies and remains a powerful instrument for social and economic growth. However, recent events related to the massive spread of misinformation and deepfakes, along with large privacy and security breaches, have raised concerns among AI practitioners and researchers about the negative and detrimental impacts of these technologies. Hence, there is an urgent call for guidelines, methods, and techniques to assess and mitigate the potentially adverse impacts and side effects of AI applications. This workshop explores how and up to which extent AI technologies can serve deceptive and malicious purposes either intentionally or not. Furthermore, it seeks to elaborate on countermeasures and mitigation actions to prevent potential negative effects and collateral damages of AI systems.
W22: International Workshop on Federated and Transfer Learning for Data Sparsity and ConfidentialityAug 20 20:00 – 2:00 (Next day) Montreal Time (UTC-4)
Lixin Fan, Heng Huang, Sinno Jialin Pan, Han Yu, Wei-Wei Tu, Yang Liu
Privacy and security are becoming a key concern in our digital age. Increasingly strict data privacy regulations such as the European Union’s General Data Protection Regulation (GDPR) bring new legislative challenges to the big data and artificial intelligence (AI) community. Many operations in the big data domain, such as merging user data from various sources for building an AI model, will be considered illegal under the new regulatory framework if performed without explicit user authorization. To explore how the AI research community can adapt to this new regulatory reality, we organize this one-day workshop. The workshop will focus on technical issues including but not limit to data collection, integration, training, sparsity and modelling, both in the centralized and distributed setting. It intends to provide a forum to discuss the open problems and share the most recent and ground-breaking work on the study and application of privacy-preserving machine learning, while dealing with sparse data.
W23: 9th AI4KM invites the 1st AIES 2021 (Artificial Intelligence for Energy and Sustainability)Aug 19 10:00 – 16:00 Montreal Time (UTC-4)
Aug 20 10:00 – 16:00 Montreal Time (UTC-4)
Gülgün Kayakutlu, Eunika Mercier-Laurent, Mieczyslaw Owoc, Georgios Saharidis, Jean-François Rizand
This workshop has the specific objective of facing environmental challenges using AI approaches and techniques to support and improve the management of sustainable energy systems within smart cities, smart facilities, smart buildings, smart transportation and smart houses. Decision support systems, planning and optimization in the sustainability and energy management fields require understanding of related systems, modelling, applying appropriate AI techniques including machine learning algorithms and handling temporary aspects of big data. Energy generation, transmission, distribution, sales and consumption need distinct focus with the energy mix optimization, distributed approaches, renewable energies with storage. Any short-term, mid-term and long-term forecasting, optimization models, trend foresights and prescriptions based on scenarios are studied in the energy world and the smart systems for sustainability. The objective of this multidisciplinary session is to bring both researchers and practitioners together to discuss methodological, technical and organizational aspects of AI used for the traditional and renewable energy lifecycles. Knowledge sharing in constructing the AI models, Business Operations, Support of short, medium and long-term decisions and innovations in the fields of sustainability and energy will improve the interactions of human life and economies. Ultimate goal of the workshop is bring a contribution to achieve some of 17 UN sustainability goals.
W24: 1st International Workshop on Adaptive Cyber Defense (ACD 2021)Aug 19 10:00 – 13:00 Montreal Time (UTC-4)
Aug 20 10:00 – 13:00 Montreal Time (UTC-4)
Damian Marriott, Kimberly Ferguson-Walter, Sunny Fugate, Marco Carvalho
Adaptive Cyber Defense Workshop – The cyber domain cannot currently be reliably and effectively defended without extensive reliance on human experts. While AI and ML solutions for detecting malicious activity are on the rise, skilled cyber defenders are in short supply and often cannot respond fast enough to mitigate the detected cyber threats. With the growing adoption of AI and ML techniques to both cyber and non-cyber settings, there is an increasing need to bridge the critical gap between AI and Cyber researchers and practitioners. We must accelerate our efforts to create autonomous and semi-autonomous cyber defenses that can learn to recognize and respond to cyber attacks or discover and mitigate weaknesses in cooperation with other cyber operation systems and human experts. Furthermore, these defenses must be adaptive, and able to evolve over time to take into account changes in attacker behavior, benign changes in the systems, and expected drift in user behavior.
W25: Toward Intellectual Property Protection on Deep Learning as a ServicesAug 21 10:00 – 16:00 Montreal Time (UTC-4)
Chee Seng Chan, Lixin Fan, Qiang Yang
Machine learning techniques, especially deep learning (DL) techniques, have made significant technological break-throughs in recent years and are widely applied in many fields, such as image classification, object detection, voice recognition, natural language processing, self-driving cars, smart healthcare, etc. Trained DL models are of high value and must be considered intellectual property of the legitimate owner, i.e. the party that created it. The value of DL models lies in the effort and resources allocated in the process of training data collection, cleansing, pre-processing, organizing, storing, and in certain cases even manual labelling, which is often time-consuming and expensive. Therefore, there is an urgent need to protect deep learning (DL) models from being illegally copied, redistributed, or abused without respecting the intellectual properties of legitimate owners. This workshop is intended to be posi-tioned at the frontier of IPR protection research and showcase the most excellent and advanced work underway at academic and private research organizations as well as government labs.
W26: ’D*** it… I’m an engineer, not a doctor!’ : AI for Spacecraft LongevityAug 21 10:00 – 16:00 Montreal Time (UTC-4)
Nikola Simidjievski, Dragi Kocev, Luke Lucas, Ljupčo Todorovski
Spacecraft operation challenges concern remote spacecraft, operating under limited computational constraints in harsh environments, equipped with decaying batteries and components (e.g., processors and memory) lagging decades behind the current technology. They relate to tasks such as autonomous spacecraft route planning, optimal spacecraft operations (forecasting and decision making), anomaly detection and diagnosis, mission planning as well as analysis of the spacecraft configuration flow. More importantly, they are associated with high-cost of failure, where repairs are impossible, diagnosis limited and emergency reaction-times long and unpredictable. There is great potential for state-of-the-art AI to address these operation challenges, particularly related to spacecraft longevity and endurance. The goal of this workshop is to bring attention to these challenges and offer a platform for discussing efforts and solutions from different communities across the different AI disciplines. More importantly, it will highlight the overarching goals of analysing large volumes of spacecraft telemetry data as well as designing, developing and deploying methods that operate on limited-capacity hardware in high-radiation environments.
W27: The Workshop on Competitive and Cooperative Social Interactions: From HRI to Machine Learning and Back AgainAug 21 10:00 – 16:00 Montreal Time (UTC-4)
Aug 21 20:00 – 2:00 (Next day) Montreal Time (UTC-4)
Pablo Barros, Francisco Cruz, German Parisi, Doreen Jirak
In this workshop, we put together the top-tier researchers on machine learning, in particular social signal processing, and human-robot interaction to discuss, understand and propose possible solutions for our overall topic “Competition and Cooperation: How to integrate machine learning and human-robot interaction in meaningful applications?”. Our intention is to leverage an enriching discussion of the existing problems on competitive and cooperative interactions on the multidisciplinary view of our contributors, and not to present individual solutions for each of the related topics. Our invited speakers, all experts in their respective areas, will discuss the open problems involving social agents’ interactions within their specialty. We also invite the submissions of position papers that present a critical view on this topic to complement and widen our discussions.
W28: 2nd International Workshop on Deceptive AIAug 19 3:00 – 7:00 Montreal Time (UTC-4)
Aug 19 9:00 – 13:00 Montreal Time (UTC-4)
Peta Masters, Stefan Sarkadi, Ben Wright, Peter McBurney, Iyad Rahwan, Liz Sonenburg
There is no dominant theory of deception. The literature on deception treats different aspects and components of deception separately, offering contradictory evidence and conflicting opinions. Emerging AI techniques offer a unique opportunity to expand our understanding of the topic from a computational perspective. However, the design, modelling and engineering of deceptive machines is not trivial from either conceptual, engineering, scientific, or ethical perspectives. DeceptAI brings together people from academia, industry and policy-making to discuss not only the practical and theoretical challenges but also the benefits, risks and potential threats of developing deceptive AI systems. This workshop takes a multidisciplinary (Computer Science, Psychology, Sociology, Philosophy & Ethics, Military Studies, Law etc.) approach to the debate, aiming to shed light on aspects of deceptive AI ranging from behaviour, reasoning and cognition through ethical and societal impact to engineering and deployment.
W29: Semantic Data Mining (SEDAMI)Aug 20 10:00 – 16:00 Montreal Time (UTC-4)
Martin Atzmueller, Grzegorz J. Nalepa, Szymon Bobek, Nada Lavrac
The general goal of data mining is to uncover novel, interesting, and ultimately understandable patterns, cf. (Fayyad et al. 1996), i.e., relating to valuable, useful and implicit knowledge. Recent advances of data mining and machine learning apparently bring new challenges in its practical use, including interpretability, introduction and preservation of knowledge, as well as the provisioning of explanations. Using semantic information such as domain/background knowledge is a promising emerging direction for addressing these problems, where the knowledge is typically represented in a knowledge repository, such as an ontology, or a knowledge base. With this workshop we aim to get an insight into the current status of research in semantic data mining. The main aspect of semantic data mining, which we focus on in this workshop, is the explicit integration of semantic information and domain knowledge into data mining and machine learning, where the algorithms for data mining/modeling or post-processing make use of the formalized knowledge to improve the results. We encourage contributions on methods, techniques and applications that are both domain-specific but also transversal to different application domains. In particular, we solicit contributions that aim to focus on semantic data mining for providing and/or enhancing interpretability, the introduction and preservation of knowledge, as well as the provisioning of explanations.
W30: 34th International Workshop on Qualitative ReasoningAug 19 10:00 – 16:00 Montreal Time (UTC-4)
Núria Agell, Anthony Cohn, Johan de Kleer, Zoe Falomir, Kenneth D. Forbus, Lledó Museros, Diedrich Wolter
The Qualitative Reasoning (QR) community develops qualitative representations and reasoning algorithms to understand the world from incomplete, imprecise, or uncertain data. Qualitative models span natural systems (e.g., physics, biology, ecology, geology), social systems (e.g., economics, cultural decision-making), cognitive systems (e.g., conceptual learning, spatial reasoning, intelligent tutors, robotics), and more. QR research includes: • Developing new formalisms and algorithms for qualitative reasoning. • Building and evaluating qualitative models in novel domains for prediction, diagnosis, explanation, and other reasoning tasks. • Characterizing how humans learn and reason qualitatively about the continuous world with incomplete knowledge. • Developing novel representations to describe time, space, change, uncertainty, causality, and continuity. QR2021 provides a forum to share progress toward these goals. Given the lack of travel costs, we hope that this makes the workshop more accessible to those who are interested in QR but might not have been able to participate previously.
W31: 2nd Workshop on Artificial Intelligence for Function, Disability, and HealthAug 20 10:00 – 16:00 Montreal Time (UTC-4)
Aug 21 10:00 – 16:00 Montreal Time (UTC-4)
Denis Newman-Griffis, Bart Desmet, Ayah Zirikly, Suzanne Tamang, Hongfang Liu
The Second Workshop on Artificial Intelligence for Function, Disability, and Health (AI4Function 2021) invites the submission of abstracts and original research papers focused on applying informatics methods, artificial intelligence (AI) or data mining techniques in the area of whole-person care, disability, and functional status information. AI4Function is a venue for researchers cutting across data science and AI methods to discuss new ways to collect and utilize functional status information within healthcare delivery, public health, and social well-being. The first AI4Function workshop brought together researchers from fields as diverse as natural language processing, in-home sensing, psychometrics, and health policy, and we aim to continue this multidisciplinary exchange in the second iteration of the workshop. For the full Call for Papers and more information on the workshop, see https://ai4function.github.io/2021.
W32: MRC-HCCS – Human-Centric and Contextual Systems, 13th International Workshop Modelling | Reasoning | ContextAug 19 10:00 – 16:00 Montreal Time (UTC-4)
Aug 20 10:00 – 16:00 Montreal Time (UTC-4)
Jörg Cassens, Rebekah Wegener, Anders Kofod-Petersen
Context is a central topic in Artificial Intelligence, essential for understanding causation and crucial for sub fields like explainable AI and ambient intelligence. It is critical for human-centred and ethical, sustainable approaches. Context is fundamental for personalisation and the development of situated & contextual AI. Context is also central in Human-Computer Interaction, essential for usability of and user experience with interac-tive systems and crucial in specialised fields, such as affective computing and social robotics. Context is inherently an interdisciplinary topic that has clear relations to linguistics, semiotics, cognitive science, mathematics, and philosophy as well as other areas such as sociology and anthropology. MRC is a bridge between diverse communities and serves as a means for integrating models and findings from dif-ferent areas. We invite contributions from other fields of study in order to further trans- and interdisciplinary ap-proaches and further the integration of discipline specific knowledge into AI research.
W33: 4th Workshop on AI for Aging, Rehabilitation and Intelligent Assisted Living (ARIAL)Aug 21 10:00 – 16:00 Montreal Time (UTC-4)
Shehroz Khan, Alex Mihailidis, Amir Ahmad
According to a WHO report, the number of people in the world aged 60 or over is projected to grow to 2.1 billion by the year 2050. Aging can come with various complexities and challenges, e.g., decline in physical and cognitive health. These changes affect a person’s everyday life, resulting in decreased social participation, lack of physical activity, and vulnerability to injury and disability, that can be exacerbated by the occurrence of various acute health events, such strokes, or long term illnesses. Leveraging AI and novel machine learning models is essential to make advancements in the field of aging and technology. Building AI models on health data will facilitate independent assisted living, promote healthy and active lifestyle, and manage rehabilitation routines effectively. With this workshop, we will bring together interdisciplinary researchers from different sub-fields of AI, in general, machine learning and deep learning to identify and approach the ARIAL-related problems.
W34: Deep Learning, Case-Based Reasoning, and AutoML: Present and Future SynergiesAug 20 10:00 – 16:00 Montreal Time (UTC-4)
David W. Aha, David Crandall, David Leake
Deep learning (DL) research has made dramatic progress in recent years, achieving high performance on supervised learning tasks for numerous problem domains. Simultaneously, there remain well-known challenges such as the need for large amounts of labeled training data, solving synthesis problems with structured solutions (e.g., designs, plans, or schedules), and explainability. Case-based reasoning is a knowledge-based methodology for reasoning from prior episodes, with complementary capabilities—such as to solve problems with small data sets or those requiring structured solutions, and to generate concrete explanations—and limitations. AutoML concerns processes for automatically generating end-to-end-machine learning pipelines, which could employ DL techniques and be built from prior cases (of successful pipeline components). This workshop will bring together researchers interested in DL, CBR, and AutoML to identify new opportunities and beneficial strategies for integrating these approaches to address current challenges.
W35: 2021 International Workshop on Safety & Security of Deep LearningAug 19 10:00 – 16:00 Montreal Time (UTC-4)
Aug 19 20:00 – 2:00 (Next day) Montreal Time (UTC-4)
Yuanfang Guo, Bo Li, Xianglong Liu, Jiantao Zhou
Deep learning, which is the current representative technology in artificial intelligence, has demonstrated tremendous success in various tasks, such as computer vision, natural language processing, data mining, etc. Unfortunately, deep learning models have also encountered critical safety and security threats in recent years. Due to its implicit vulnerability, deep learning models can be easily affected by the adversarial perturbations and perform abnormally, which may yield serious consequences in certain applications such as autonomous driving, etc. Meanwhile, deep learning models may also be utilized by the malicious attackers to generate forged multimedia content, such as fake images/videos, to deceive people, which may induce trust issues among different people and organizations. In this workshop, we aim to bring more attentions from the researchers in the fields of adversarial attack & defense, forensics, robust deep learning, explainable deep learning, etc., to discuss the recent progresses and future directions for tackling the various safety and security issues of deep learning models.
W36: Workshop on Artificial Intelligence Safety (AISafety)Aug 19 8:00 – 14:00 Montreal Time (UTC-4)
Aug 20 8:00 – 14:00 Montreal Time (UTC-4)
Huáscar Espinoza, José Hernández-Orallo, Xiaowei Huang, Cynthia Chen, Gabriel Perdroza, Mauricio Castillo-Effen, John McDermid, Seán Ó hÉigeartaigh, Richard Mallah
In the last decade, there has been a growing concern on risks of AI. Safety is becoming increasingly relevant as humans are progressively ruled out from the decision/control loop of intelligent, and learning-enabled machines. In particular, the technical foundations and assumptions on which traditional safety engineering principles are based, are inadequate for systems in which AI algorithms, in particular Machine Learning (ML) algorithms, are interacting with the physical world at increasingly higher levels of autonomy. We must also consider the connection between the safety challenges posed by present-day AI systems, and more forward-looking research focused on more capable future AI systems, up to and including Artificial General Intelligence (AGI). The AISafety workshop seeks a new perspective of system engineering where multiple disciplines such as AI and safety engineering are viewed as a larger whole, while considering ethical and legal issues, in order to build trustable intelligent autonomy.
W37: Continual and Multimodal Learning for Internet of ThingsAug 21 10:00 – 16:00 Montreal Time (UTC-4)
Tong Yu, Susu Xu, Handong Zhao, Ruiyi Zhang, Shijia Pan
Internet of Things (IoT) provides streaming, large-amount, and multimodal sensing data over time. The statistical properties of these data are often significantly different by sensing modalities and temporal traits, which are hardly captured by conventional learning methods. Continual and multimodal learning allows integration, adaptation and generalization of the knowledge learnt from previous experiential data collected with heterogeneity to new situations. Therefore, continual and multimodal learning is an important step to improve the estimation, utilization, and security of real-world data from IoT devices. This workshop aims to explore the intersection and combination of continual machine learning and multimodal modeling with applications in Internet of Things. The workshop welcomes the works addressing these issues in different applications and domains, such as natural language processing, computer vision, human-centric sensing, smart cities, health, etc. We aim at bringing together researchers from different areas to establish a multidisciplinary community and share the latest research.
W38: Data Science Meets Optimization (DSO)Aug 19 10:00 – 16:00 Montreal Time (UTC-4)
Aug 20 10:00 – 16:00 Montreal Time (UTC-4)
Patrick De Causmaecker, Tias Guns, Michele Lombardi, Yingqian Zhang
This workshop on the close relationship and interplay between data science and optimization continues on the DSO@IJCAI2020, DSO@IJCAI2019, and the DSO@IJCAI-ECAI workshop in 2018. Invited are studies on how techniques from combinatorial optimization and mathematical programming can be enforced by learning from historical data and on how such advanced techniques can contribute to machine learning and data mining. The DSO workshop is closely related to the DSO working group of The Association of European Operational Research Societies (EURO) with yearly streams and workshops at major conferences such as EURO2021 in Athens, IFORS 2021 Virtual, EURO 2018 in Valencia, EURO 2019 in Dublin, IFORS 2017 in Quebec, CPAIOR 2017 in Padua, CEC 2017 in San Sebastian.