Interview with Beishui Liao

Can you briefly introduce yourself?
I am a full professor of logic and artificial intelligence at Zhejiang University, China. Since 2014 I am a visiting professor of University of Luxembourg, and have well collaborated with Prof. Leon van der Torre in the area of AI and ethics as well as in many research projects.

What are the main research challenges you are working on now? Currently I apply logic as a formal tool to deal with the reasoning of incomplete, inconsistent, and dynamic information. It has been well known that these kinds of information cannot be resolved by the classic logical methods. I develop new theories, methods and algorithms of formal argumentation and apply them to some real life applications, e.g. mining and reasoning with legal texts, and machine ethics, etc.

What are in your opinion the main research challenges in AIs, robotics and reasoning? Recently I am leading a new project in the direction of AIs, robotics and reasoning, which is supported by the Convergence Research Project for Brain Research and Artificial Intelligence of Zhejiang University. We focus on the following challenges from two perspectives. In the perspective of human cognition, on the one hand we need to consider how AIs or robotics perceive what human need, on the other hand we as human have to understand what have been happened inside machines. Game theory is the second perspective of our research project. The human-AI interaction needs us to think about decision making via social networks where human and machines interact each other.

What is are favorite online resources? Usually I do search papers and books via google and the data-base of Zhejiang University. I also read papers regularly from some international conferences in my field, like IJCAI, KR, and COMMA, etc.

What are your aims outside of academia? Outside academia, there are still a lot to be done, such as how well can we connect theories to applications. We can do this by taking people from the industries into research discussions. We should also consider how to use logic to benefit the education. Logic can provide a good training of the critical thinking for the public. We will try to work on this hand in hand with the media.

Interview with Leon van der Torre

Can you briefly introduce yourself?
I was born in Rotterdam on 18 March 1968 and I am a full professor of computer science at the University of Luxembourg and head of the Individual and Collective Reasoning (ICR) group at the Faculty of Science, Technology and Communication. My main research interests concern deontic logic and multi-agent systems, with a special glance on logic for AI. I am the deputy editor in chief of the Journal of Applied Logics – IfCoLoG Journal of Logics and their Applications. My main stance is that there are no applications without in depth theory. Since 2000, I am married to the artist Egberdien van der Peijl with whom I have two sons.

What are the main research challenges you are working on now?
I am mainly addressing basic research challenges in unifying modal preference based semantics and norm based semantics in deontic logic, aligning formal argumentation with deontic logic/normative systems and social network theory. I also works on applications of reasoning in space AI, social robots, and legal AI.

What are in your opinion the main research challenges in AIs, robotics and reasoning?
In my view, Artificial Intelligence lost the connection with its roots. Since Aristotle, logics and mathematics are at the basis of formal reasoning, while in recent years AI systems are seen as “black boxes” able to take decisions but unable to provide logical explanations for their decisions. For this reason, many researchers recently advocated the need of more Explainable AI systems. AI in computer science also needs to be reunited with cognitive science in psychology and mathematics, in order to responsibly manage expectations of AI in society.

What are your favorite online resources?
I am subscribed to several mailing lists (IAAIL, PHILOS-L, Robotics, UAI, LOGIC, agents, Folli, etc.) as well as newsletter such as the ChinAI newsletter, that I read regularly to remain aware of the progresses done by the different scientific communities. For teaching, I usually refer to material available at the MIT OpenCourseWare.

What are your aims outside of academia?
I like to fly, and China is indeed one of my favorite countries in which flying. I also like to travel; concerning China, in particular I love to all train tracks to Russia. For the forthcoming year, I would to learn Chinese as I am planning to spend part of my sabbatical at Zhejiang University.

Interview with Alexander Steen

Can you briefly introduce yourself?
I am a Post-Doc at the University of Luxembourg, more specifically in the Individual and Collective Reasoning (ICR) group of Leon van der Torre at the Faculty of Science, Technology and Communication. I am interested in the foundations and applications of automated reasoning technology, in particular in expressive logical formalisms, and one of the main authors of the Leo-III theorem prover for higher-order logic. I joined the ICR group last year in summer after finishing my PhD in computational logic at Freie Universität Berlin. At Freie Universität, I studied Computer Science and Mathematics after moving to Berlin in 2009 from the most beautiful city at the North Sea and my birthplace, Cuxhaven, in Lower Saxtony, Germany.

What are the main research challenges you are working on now?
Currently, I am working on the automation of non-classical logics for deontic and normative reasoning via higher-order theorem provers and proof assistants. For this I extend previous work on semantical embeddings of various logical formalisms and their inclusion into Leo-III. In addition, I am working on general purpose improvements to both the theory and practice of higher-order automation, e.g. the development of suitable data structures and calculi optimizations. Also, I am interesting in including machine learning technology to reasoning systems.

What are in your opinion the main research challenges in AIs, robotics and reasoning?
To my mind, one of the main challenges in the next decades will be the combination of the classical symbolic reasoning approaches with the more recent developments in learning-based AI technology; not only on the level of joint technological projects, but also on the leval of creating a cohesive AI community that learns from each other’s strengths and capabilities. I feel that currently these communities are too far apart; this needs to be addressed in the future.

What is are favorite online resources?
In addition to the usual mailing lists and search/recommendation engines for papers (ResearchGate, Google Scholar, etc.), I am a frequent reader of the Y combinator hacker news; they are a great resource for a large variety of interesting topics.

What are your aims outside of academia?
There are a lot of things I like to do; this includes hiking, singing, listening to good music, attending concerts, spending time with my family and friends, etc. I guess the aim would be to do all that more consistently.

Interview with Wu Fei

Can you briefly introduce yourself?
I am a full professor of Zhejiang University at the college of computer science. Currently, I am the vice-dean of college of computer science, and the director of Institute of Artificial Intelligence of Zhejiang University. My research interests mainly include Artificial Intelligence, Multimedia Analysis and Retrieval and Machine Learning.

What are the main research challenges you are working on now?
Artificial intelligence is now heading towards how to integrate data-driven learning and knowledge-guided inference to perform better reasoning and decision instead of correlation learning via metric matching. My current research is about how to perform symbolic AI, data-driven learning and reinforcement learning together to support causal reasoning.

What are in your opinion the main research challenges in AIs, robotics and reasoning?
In general, the relevant information (e.g., knowledge instance and exemplar data) w.r.t the input data is sparked from external memory in the manner of memory-augmented learning. Memory-augmented learning is an appropriate method to integrate data-driven learning, knowledge-guided inference and experience exploration.

What is are favorite online resources?
I like to use Internet search engine to find my favorite papers or articles.

What are your aims outside of academia?
I have engaged in some MOOCs. Now I teach a MOOC (Artificial Intelligence: Models and Algorithms) in https://www.icourse163.org/course/ZJU-1003377027. At MOOC, I force myself to explain the complex equations via a concise way.

Interview with Nicolas Guelfi

Can you briefly introduce yourself?
I am full professor at University of Luxembourg. Since 20 years I have contributed to the creation of the university. I founded the Laboratory for Advanced Software Systems at University of Luxembourg which I directed for more than 10 years. I have been member of the ERCIM (European Research Consortium in Mathematics and Informatics) executive committee and founded the ERCIM Working Group on Rapid Integration of Software Engineering Techniques and the ERCIM Working Group on Software Engineering for Resilient Systems. I also have been expert for the courthouse of Luxembourg concerning trials on conformance questions for 10 years. I mainly have contributed to the field of software engineering as a researcher by publishing articles, editing books, acting as program chair or reviewing committee member. My research contributions has mainly focused on: requirements definition and specification using formal, semi-formal or informal methods; and system dependability and resiliency. My current main field of research is on development methods for deep learning. Finally, I am the head a new bachelor in Computer Science at the University of Luxembourg which opened in 2017 and I am planning to open a private academy in artificial intelligence and software engineering.

What are the main research challenges you are working on now?
Software engineers require a large amount of data for building neural network-based software systems. The engineering of these data is often neglected, though, it is a critical and time-consuming activity. In my team we try to develop a novel software engineering approach for dataset augmentation using neural networks. We propose a rigorous process for generating synthetic data to improve the training of neural networks.

What are in your opinion the main research challenges in AIs, robotics and reasoning?
I would say, first, mastering from a software engineering perspective the foundations of AI systems engineering. This means finding the right methods and tools based on sound theories to master the quality of AI based systems. Secondly introduce causality in probabilistic reasoning.

What is are favorite online resources?
IEEE and ACM websites for high quality publications in my research domains ResearchGate and linked in for networking AWS and GCP for online computing resources

What are your aims outside of academia?
Not so much except …. health and family 🙂

Interview with Huaxin Huang

Can you briefly introduce yourself?
My name is Huaxin HUANG, and I am a full professor of the institute of logic and cognition at Zhejiang University. My recent research focuses on the interdisciplinary areas of logic, language, and cognition. And I supervise graduate students in the topics of logic of language and cognitive pragmatics. A few years ago I had presided over and completed a Major Program of the National Social Science Fund of China, a cognitive study based on logical horizon.

What are the main research challenges you are working on now?
From the second half of last year, I am chairing another major project of the National Social Science Fund of China, the logical representation and cognitive computing of metaphors in Chinese. This project will take five years. And it involves the following five aspects: 1. To build a corpus of metaphors in Chinese for natural language processing; 2. To have hybrid inference models for computing metaphors in Chinese; 3. To analyze the generation and understanding of metaphors in Chinese in game theory; 4. To set up the argumentation-based reasoning system and semantic computing for metaphors in Chinese; 5. To achieve the algorithm implement and synthetic simulation for the cognition of metaphors in Chinese. I think that the study on metaphors is not only the key to solve the cognition issues in language, but also the main point to deeply understand the nature of human reasoning. The aim of our research is to solve the semantic ambiguity and its dynamic in the semantic level, and then to build semantic reasoning systems to have more powerful explanation ability by taking multi-agent interaction and pragmatic factors into consider.

What are in your opinion the main research challenges in AIs, robotics and reasoning?
Recently I pay more attention on the progress of existing work on language understanding and logical reasoning in AI. Especially due to the wide applications of deep learning and neural networks, many achievements have been made in the area of natural language understanding in the industrial area. I think the next challenge shall be the explainable issues in these models. And this could inspire the work in linguistics and logicians or other areas in related.

What is are favorite online resources?
The Chinese data bases I often use include CNKI, Wanfang Data, and Duxiu Data, while the English include IEEE, EBSCO, and Elsevier. I also use the CCL corpus of Pecking University and the BCC corpus of Beijing Language and Culture University. To check the corpus for metaphors, I use the VU Amsterdam Metaphor Corpus too. I like to check papers via https://arxiv.org/.

What are your aims outside of academia?
My hobbies outside of academics include reading novels and poetry. I usually take a walk every day, listen to music, and taste tea with friends. I am also very happy to give public lectures on logic and cognitive science.