1- Artificial Engineering Depatement, Faculty of Computer Engineering, K. N. Toosi University of Technology, Tehran, Iran
2- CIC, Instituto Politécnico Nacional
2- CIC, Instituto Politécnico Nacional
Abstract: (593 Views)
The application of Abstract Meaning Representation (AMR) is widely increasing as a principal form of structured sentence semantics, and it is considered as a turning point for Natural Language Processing (NLP) research. AMRs are rooted and labeled graphs, which capture semantics on sentence level and abstract away from Morpho-Syntactic properties. The nodes of the graph represent meaning concepts, and the edge labels show relationships between them. In this paper, we give a brief review about the existing approaches of generating text from AMR and parsing input text to produce AMR by studying various research from 2013 to 2022. Besides, we explain how the researchers have been used AMR for prevalent NLP tasks. Afterwards, we describe the existing datasets and evaluation metrics, which can be used in this regard. Finally, we discuss some basic features and challenges of AMR.
Type of Article: Review paper |
Subject:
General
Received: 2022/10/2 | Accepted: 2023/04/29 | ePublished ahead of print: 2023/10/31
Received: 2022/10/2 | Accepted: 2023/04/29 | ePublished ahead of print: 2023/10/31
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
