Ontology Summit Synthesis 29 March 2017

Summit theme: Relationships between AI and ontology Tracks focus on 3 specific relationships Track A: Using Automation and ML to Extract Knowledge and Improve Ontologies (Learning →Ontology) Track B: Using background knowledge to improve machine learning results (Ontology→Learning) Track C: Using ontologies for logical reasoning (Ontology→Reasoning)

Track A, B, C What was covered in Session 1 Key points from session Chat Log etc. Material synthesized in Track blog page

Track A: 8 March 2017 Track Co-champion Gary Berg-Cross Estevam Hruschka (Associate Professor at Federal University of Sao Carlos DC-UFSCar & adjunct Professor at Carnegie Mellon University) “Never-Ending Language Learning (NELL)” Valentina Presutti, (Semantic Technology Laboratory of the Institute of Italian National Research Council (CNR)), “Semantic Web machine reading with FRED” Alessandro Oltramari (Research Scientist at Bosch) "From machines that learn to machines that know: the role of ontologies in machine intelligence"

Track B: 15 March 2017 Co-champions: Mike Bennett, Andrea Westerinen Track B Session 1 Content SimonDavidson, Psonify The Investigator's Toolkit – Deriving Immediate, Actionable Insights from Unstructured Data Ken Baclawski (College of Computer and Information Science, Northeastern University) Combining Ontologies and Machine Learning to Improve Decision Making

Track C: 22 March 2017 Co-Champions: Donna Fritzsche and Ram D. Sriram Eugen Kuksa, University of Magdeburg, Germany. “Reasoning with Ontologies in Ontohub” Pascal Hitzler, Wright State University in Dayton, Ohio, U.S.A. “On the Roles of Logical Axiomatizations for Ontologies”  Jans Aasman, Franz Inc. “Cognitive Probability Graphs need an Ontology”

Track A Report-back

Track B Report-back

Track C Report-back

Discussion and Common Themes

Synthesizing All That

Adjourn

ANNEX: Possible things to ruminate on Use Cases Regulatory parsing / action Chatter / malfeasance and intel Company filings Questions Natural language analysis Kidsn of ontology needed What’s generated (Track A) and how deep os the meaning What’s needed to enhance – and how Context, heteronymy handled Humans in the loop Logical formalisms Compared to other SW use cases, do these applications require the same kinds of formalisms? Is DL a limitation or a requirement? Statistics / clustering v deeper taxonomies The role of reasoning