| Full text | |
| Author(s): Show less - |
Ligabue, Pedro de Moraes
;
Franco Brandao, Anarosa Alves
;
Peres, Sarajane Marques
;
Cozman, Fabio Gagliardi
;
Pirozelli, Paulo
;
Li, P
;
Yu, K
;
Chawla, N
;
Feldman, R
;
Li, Q
;
Wu, X
Total Authors: 11
|
| Document type: | Journal article |
| Source: | 2022 IEEE INTERNATIONAL CONFERENCE ON KNOWLEDGE GRAPH (ICKG); v. N/A, p. 8-pg., 2022-01-01. |
| Abstract | |
The Blue Amazon represents a large area off the coast of Brazil. The region is important for economic reasons, as it holds the majority of Brazil's oil and gas resources, as well as for its biological, geological and environmental features. Not much information about this region, however, is organized in systematic and computer-readable ways. In this context, we built a knowledge graph for the Blue Amazon based on a set of 496 scientific paper abstracts on the subject. The corpus includes a large number of topics, such as biodiversity, natural resources and climate change. Entities and relationships were automatically extracted from text through the use of OpenIE (a relationship triple extraction method) and were combined using word embeddings. The generated knowledge graph was then evaluated by comparing it to the source documents. Results show that key knowledge was correctly extracted, but also that there is still room for improvement, specially when it comes to weeding out undesirable or context-dependent data. (AU) | |
| FAPESP's process: | 19/07665-4 - Center for Artificial Intelligence |
| Grantee: | Fabio Gagliardi Cozman |
| Support Opportunities: | Research Grants - Research Program in eScience and Data Science - Research Centers in Engineering Program |