Knowledge graphs (DUS.AD016.118)
Thematic area
Social sciences and humanities, cultural heritage
Project area
Cognizione naturale e artificiale: comunicazione, linguaggio, etica (DUS.AD016)Structure responsible for the research project
Institute of cognitive sciences and technologies (ISTC)
Project manager
VALENTINA PRESUTTI
Phone number: 0644595212
Email: valentina.presutti@istc.cnr.it
Abstract
This project aims at investigating methodologies for creating knowledge graphs as well as for using them as objects for empirical research. Research on knowledge graphs creation include ontology development methodologies as well as knowledge extraction for populating knowledge graphs according to ontological models. Empirical research on knowledge graphs focus on looking for recurrent structures i.e. patterns, that may provide evidence of common practices in knowledge engineering or of "unintentional" formalisation of commonsense facts. These patterns can support the validation of foundational / cognitive theories on the way humans organise their knowledge cognitively, e.g. frame theory. In addition, knowledge graphs can be used in order to test semantic hypotheses as input of deep learning models for commonsense reasoning tasks e.g. natural language generation / understanding.
Goals
To identify recurrent (modeling) structures in knowledge graphs
To test semantic hypotheses over deep learning models for natural language processing
To build commonsense knolwedge graphs
Start date of activity
01/05/2020
Keywords
knowledge graphs, common sense reasoning, frames
Last update: 27/04/2024