Although spoken personal assistants and chatbots are now more and more important for users and businesses, we still don’t really know how new methods and growing data could lead us to a human-level naturalness in human-computer conversation. Six decades of failure to pass the Turing test by machines lead us to rethink previous approaches, to lean towards new technologies and knowledge sources, and to combine them with advances in philosophy, linguistics and cognitive science. We stress the fact that the age of information explosion and more powerful learning algorithms gives us a whole new spectrum of possibilities for creating an intelligent machine. Many marvelous ideas of the dawn of Artificial Intelligence research faced problems of exceptions and the impossibility of manual input of all needed knowledge, but today we have vast amounts of data from sensors, images and text so that we can rethink classical AI methods and approaches to dialog systems. The increased use of WWW, Internet of Things or knowledge bases, etc. could allow us to determine standard human behaviors, emotions or even moral reasoning according to the Wisdom of Crowds hypothesis. Collective input data could also help to retrieve knowledge about the physical world we live in. By combining Natural Language Processing methods and/or Machine Learning techniques with cognitive architectures and philosophy of mind, we can discover a new range of intelligent systems that understand us, our environment, and our feelings. In this context, we see a role for NLP/ML and cognitive approaches to play in developing a new generation of user-friendly, more autonomous but still safe systems that, through interaction with the user and the world, can learn how to reason, behave or speak naturally. We are interested in original papers on methods for dialog systems that use common sense knowledge and reasoning, affective computing, cognitive methods, learning from broad sets of data and acquiring knowledge, language itself, user preferences. etc.

The workshop intends to spark an interdisciplinary discussion on joining forces to return AI to its original, broader and deeper goals which are currently represented by AGI – Artificial General Intelligence.

After two successful LaCATODAs associated with Society for the Study of Artificial Intelligence and Simulation of Behaviour (AISB) and International Association for Computing and Philosophy (IACAP) conferences, we plan an 1-day workshop under IJCAI-17 umbrella to take place in Melbourne on (most probably) August 19 or 20, 2017. An invited talk from Dr. Mark Ellison (Australian National University) is planned.

Topics of interest include, but are not limited to:

Affect analysis (and its applications)
Context processing
Dialog processing
Ethical issues of dialog systems
Evaluating conversational skills
Machine Learning for Dialog
Opinion Mining
Probabilistic Approaches
Sentiment Analysis
Social Informatics
Text mining techniques for dialogs
Turing Test
Preference models
Knowledge acquisition for dialog agents
Pragmatics of decision making
Cognitive aspects of decisions and opinions

How does (or could) your work improve Human-Computer Conversation?

The Third Linguistic and Cognitive Approaches to Dialog Agents (LaCATODA 2017) will be held as a workshop at IJCAI 2017, in Melbourne, Australia, August 21, 2017.

NEWS: Paper submission deadline extended till May 22!

Registration: LaCATODA 2017 is an integral part of the Twenty-sixth International Joint Conference on Artificial Intelligence (IJCAI-17), please refer to the main conference site.