Machines are amazing learners. Humans are amazing learners. AI is best when powered by both. AI solutions that employ both machine learning and knowledge-engineered rules learn continuously from data whilst at the same time are informed by wisdom and commonsense expressed in rules. Forrester Research Vice President & Principal Analyst, Mike Gualtieri, will convey the key trends in enterprise AI and rules/decision management; and discuss how leading enterprises can use them in combination to build truly learning AI-infused applications at scale.
Existential rules, also known as Datalog+, are an expressive knowledge representation and reasoning language, which has been mainly investigated in the context of ontological query answering. This talk will first review the landscape of decidable classes of existential rules with respect to two fundamental problems, namely chase termination (does a given set of rules ensure that the chase terminates for any factbase?) and FO-rewritability (does a given set of rules ensure that any conjunctive query can be rewritten as a first-order query?). Regarding the chase, we will specifically focus on four well-known variants: the oblivious chase, the semi-oblivious (or skolem) chase, the restricted chase and the core chase. We will then study the relationships between chase termination and FO-rewritability, which have been little investigated so far. This study leads us to another fundamental problem, namely boundedness (does a given set of rules ensure that the chase terminates for any factbase within a predefined depth?). The boundedness problem was deeply investigated in the context of datalog. It is known that boundedness and FO-rewritability are equivalent properties for datalog rules. Such an equivalence does not hold for general existential rules. We will provide a characterization of boundedness in terms of chase termination and FO-rewritability for the oblivious and semi-oblivious chase variants. Interesting questions remain open. This talk will rely on results from the literature and joint work published at ICDT 2019 and IJCAI 2019.
Marie-Laure Mugnier is a professor at the University of Montpellier and the scientific leader of a research team in knowledge representation and reasoning. Her current research mainly focuses on rule-based formalisms to do reasoning on data, specifically within the Ontology-Based Data Access paradigm. She regularly publishes papers in the main international venues in artificial intelligence. Recently, she was program co-chair of the 8th International Conference on Web Reasoning and Rule Systems (RR 2014), general co-chair of the 30th workshop on Description Logics (DL 2017) and co-organised workshops associated with IJCAI 2015 & 2016 and The Web 2019.
Business processes are the operational backbone of modern organizations. Their continuous management and improvement is key to the achievement of business objectives. Accordingly, a common task for analysts and managers is to discover, assess, and exploit process improvement opportunities. Current approaches to discover process improvement opportunities are expert-driven. In these approaches, data are used to assess opportunities derived from experience and intuition rather than to discover them in the first place. Moreover, as the assessment of opportunities is manual, analysts can only explore a fraction of the overall space of improvement opportunities.
Recent advances in data mining, machine learning, and artificial intelligence are making it possible to move from manual to semi-automated or automated approaches to business process management. This talk will cover three such trends: (1) process mining, (2) predictive process monitoring, and (3) robotic process mining.
Process mining is a body of methods for analyzing data generated by the execution of business processes in order to extract insights about weaknesses and improvement opportunities. These methods allow us to understand how a given business process is actually executed, if and how its execution deviates with respect to expected or normative pathways, and what factors contribute to poor process performance or undesirable outcomes.
Meanwhile, predictive process monitoring techniques allow us to monitor ongoing executions of a business process in order to predict future states and undesirable outcomes at runtime. These predictions can be used to trigger interventions at runtime in order to maximize a given reward function, for example by generating alarms or making recommendations.
Robotic process mining, on the other hand, seeks to automatically identify repetitive routines (particularly clerical routines) that are fully deterministic and can therefore be automated by means of so-called Robotic Process Automation (RPA) scripts, thus relieving workers from tedious and error-prone work.
The talk will introduce the above trends and will position them with respect to a broader class of techniques for AI-driven automated process improvement. The talk will review the state of the art in this emerging field and will outline research opportunities and challenges.
Marlon Dumas is Professor of Information Systems at University of Tartu. He conducts research in the fields of business process management, business data analytics, and information systems engineering. His research has earned seven best paper awards at international conferences in the fields of software engineering and information systems, as well as three best student paper awards (with his PhD students), three test-of-time awards, and two best prototype demonstration awards. During his career, Prof. Dumas has attracted and executed research projects funded by the Australian Research Council, the European Union (FP7 and H2020 programs), DARPA, U.S. Army Research Lab, and several companies including SAP, Microsoft/Skype, and Swedbank. He is currently recipient of an Advanced Grant from the European Research Council (ERC), with the mission of developing algorithms for automated identification and assessment of business process improvement opportunities. His research has led to several major open-source prototypes including YAWL – a workflow management system used in several large organizations worldwide – and Apromore.org – a process mining tool used in several financial companies in Europe and Australia. He is also co-inventor of 10 U.S./EU patents and co-author of the textbook "Fundamentals of Business Process Management" (Springer), which is used over 250 universities worldwide.
Dynamic systems that operate autonomously in nondeterministic (uncertain) environments are becoming a reality. These include intelligent robots, self-driving cars, but also manufacturing systems (Industry 4.0), smart objects and spaces (IoT), advanced business process management systems (BPM), and many others. These systems are currently being revolutionized by advancements in sensing (vision, language understanding) and actuation components (autonomous mobile manipulators, automated storage and retrieval systems). However, in spite of these advances, their core logic is still mainly based on hard-wired rules either designed or possibly obtained through a learning process. On the other hand, we can envision systems that are able to deliberate by themselves about their course of action when un-anticipated circumstances arise, new goals are submitted, new safety conditions are required, and new regulations and conventions are imposed. Crucially, empowering dynamic systems with deliberating capabilities carries significant risks and therefore we must be able to balance such power with trust. For this reason it is of interest to make these systems queryable, analyzable and explainable in human terms, so as to be guarded by human oversight. Recent scientific discoveries in Knowledge Representation and Planning combined with insights from Verification and Synthesis in Formal Methods, Data-Aware Processes in Databases, as well as other areas of AI, chart a novel path for realizing what we may call White-Box Self-Programming Mechanisms, that is, systems with a multifaceted model of the world that can be exploited to deliberate on their course of action and answer queries about their behavior.
Giuseppe De Giacomo is full professor in Computer Science and Engineering at Univ. Roma “La Sapienza". His research activity has concerned theoretical, methodological and practical aspects in different areas of AI and CS, most prominently Knowledge Representation, Reasoning about Actions, Generalized Planning, Autonomous Agents, Service Composition, Business Process Modeling, Data Management and Integration. He is AAAI Fellow, ACM Fellow, and EurAI Fellow. He is Program Chair of ECAI 2020. He is has got an ERC Advanced Grant for the project WhiteMech: White-box Self Programming Mechanisms (2019-2024).