Adam Pease
Amit Sheth

Adam Pease

Amit Sheth

Adam Pease, Principal Scientist, Infosys Foothill Research, Palo Alto, California, USA

Title: Conceptual Pragmatics: A Library of Logical Definitions

Abstract:

What is an apple, a jump or the number 2 and how can we hope to have a computer understand these things with any of the same depth or richness that people do? We now have machine learning systems that can mimic, at some level, human sensory subsystems, recognizing objects in pictures, or voices and words in streams of audio. But we also need a cognitive-level representation - one that not only can recognize patterns but also hold information about those patterns that allows for explanation and communication. A person can describe a previously unseen object to another person, who can then recognize it and understand its characteristics before seeing it, and before seeing it a million times. Someone who has never seen a child skip can still be told how to recognize skipping. We can tell another person the context of skipping, as an isolated action or the likely context in which such actions occur. In this talk I describe a unified corpus of logically-expressed and computable meaning about concepts that has application in language and image understanding. It is a library of pragmatics that can be used to express facts independently of whether they are learned over many presentations of visual or auditory data, or related in communication. I also describe its application in image recognition and language understanding.

Biography:

Adam Pease is a Principal Scientist at the Infosys Foothill Research Center in Palo Alto. He has led research in ontology, linguistics, and formal inference, including development of the Suggested Upper Merged Ontology (SUMO), the Controlled English to Logic Translation (CELT) system, and the Sigma knowledge engineering environment. Sharing research under open licenses, in order to achieve the widest possible dissemination and technology transfer, has been a core element of his research program. He is the author of the book “Ontology: A Practical Guide”.

Amit Sheth, Professor and Executive Director of Kno.e.sis, Wright State University, Dayton, Ohio, USA

Title: On Exploiting Multimodal Information for Machine Intelligence and Natural Interactions - With Examples from Health Chatbots

Abstract:

The Holy Grail of machine intelligence is the ability to mimic the human brain. In computing, we have created silos in dealing with each modality (text/language processing, speech processing,image processing, video processing, etc.). However, the human brain’s cognitive and perceptual capability to seamlessly consume (listen and see) and communicate (writing/typing, voice, gesture) multimodal (text, image, video, etc.) information challenges the machine intelligence research. Emerging chatbots for demanding health applications present the requirements for these capabilities. To support the corresponding data analysis and reasoning needs, we have to explore a pedagogical framework consisting of semantic computing, cognitive computing, and perceptual computing (http://bit.ly/w-SCP). In particular, we have been motivated by the brain’s amazing perceptive power that abstracts massive amounts of multimodal data by filtering and processing them into a few concepts (representable by a few bits) to act upon. From the information processing perspective, this requires moving from syntactic and semantic big data processing to actionable information that can be weaved naturally into human activities and experience (http://bit.ly/w-CHE). Exploration of the above research agenda, including powerful use cases, is afforded in a growing number of emerging technologies and their applications - such as chatbots and robotics. In this talk, I will provide these examples and share the early progress we have made towards building health chatbots (http://bit.ly/H-Chatbot) that consume contextually relevant multimodal data and support different forms/modalities of interactions to achieve various alternatives for digital health (http://bit.ly/k-APH). I will also discuss the indispensable role of domain knowledge and personalization using domain and personalized knowledge graphs as part of various reasoning and learning techniques.

Biography:

Amit Sheth is an educator, researcher, and entrepreneur. He is the LexisNexis Ohio Eminent Scholar, and IEEE Fellow, an AAAI Fellow, and the executive director of Kno.e.sis - the Ohio Center of Excellence in Knowledge-enabled Computing. Kno.e.sis. is a multi-disciplinary Ohio Center of Excellence in BioHealth Innovation. Its faculty and researchers are computer scientists, cognitive scientists,biomedical researchers, and clinicians. Sheth is working towards a vision of Computing for Human Experience enabled by the capabilities at the intersection of AI (semantic, cognitive, and perceptual computing), Big and Smart Data (exploiting multimodal Physical-Cyber-Social data), and Augmented Personalized Health. His recent work has involved Web 3.0 technologies and involves enterprise, social sensor/IoT data and applications.