Abstract
This chapter illustrates how suitable neuro-symbolic models for language understanding can enable domain generalizability and robustness in downstream tasks. Different methods for integrating neural language models and knowledge graphs are discussed. The situations in which this combination is most appropriate are characterized, including quantitative evaluation and qualitative error analysis on a variety of commonsense question answering benchmark datasets.
| Original language | English |
|---|---|
| Title of host publication | Neuro-Symbolic Artificial Intelligence |
| Subtitle of host publication | The State of the Art |
| Editors | Pascal Hitzler, Md Kamruzzaman Sarker |
| Publisher | IOS Press BV |
| Chapter | 13 |
| Pages | 294-310 |
| Number of pages | 17 |
| ISBN (Electronic) | 9781643682440 |
| ISBN (Print) | 9781643682440 |
| DOIs | |
| Publication status | Published - 2021 |
Publication series
| Name | Frontiers in Artificial Intelligence and Applications |
|---|---|
| Publisher | IOS |
| Volume | 342 |
| ISSN (Print) | 0922-6389 |
| ISSN (Electronic) | 1879-8314 |
Bibliographical note
Publisher Copyright:© 2022 The authors and IOS Press. All rights reserved.
Fingerprint
Dive into the research topics of 'Generalizable Neuro-Symbolic Systems for Commonsense Question Answering'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver