2nd March 2021 - 9am - Jose Manuel Gomez-Perez
Title: On the Role of Knowledge Graphs and Language Models in Machine Understanding of Scientific Documents
Slides are available for download below:
Abstract:
Reading a textbook in a particular discipline and being able to answer the questions at the end of each chapter is one of the grand challenges of artificial intelligence, which requires advances in language, vision, problem-solving, and learning theory. Recent breakthroughs in language modeling, as well as the proliferation of knowledge graphs, have created great expectations in this regard. Such expectations are particularly high in domains like science, where complex phenomena need to be tackled that motivate the need for intelligent systems able to understand scientific documents and to assist scientists in their daily work. In this talk, we will see some of the advantages and disadvantages of knowledge-based and neural approaches to natural language processing and understanding, and propose different ways in which such approaches can be combined. We will also explore the application of such methods in different tasks related to machine understanding of scientific documents and illustrate how the resulting models represent multimodal information internally and reason with it.
Speaker: Jose Manuel Gomez-Perez, Director of Language Technology Research, Expert.AI (Madrid, Spain)
The meeting was held online via Zoom.
2nd March 2021 - 9am - Jose Manuel Gomez-Perez
Zoom