Abstract
Recognition of causality is important to achieve natural language discourse understanding. Previous approaches rely on shallow linguistic features. In this work, we propose to identify causality in verbnoun pairs by exploiting deeper semantics of nouns and verbs. Particularly, we acquire and employ three novel types of knowledge: (1) semantic classes of nouns with a high and low tendency to encode causality along with information regarding metonymies, (2) data-driven semantic classes of verbal events with the least tendency to encode causality, and (3) tendencies of verb frames to encode causality. Using these knowledge sources, we achieve around 15% improvement in Fscore over a supervised classifier trained using linguistic features.
| Original language | English (US) |
|---|---|
| State | Published - 2014 |
| Event | 15th Annual Meeting of the Special Interest Group on Discourse and Dialogue, SIGDIAL 2014 - Philadelphia, United States Duration: Jun 18 2014 → Jun 20 2014 |
Other
| Other | 15th Annual Meeting of the Special Interest Group on Discourse and Dialogue, SIGDIAL 2014 |
|---|---|
| Country | United States |
| City | Philadelphia |
| Period | 6/18/14 → 6/20/14 |
Fingerprint
ASJC Scopus subject areas
- Computer Graphics and Computer-Aided Design
- Computer Vision and Pattern Recognition
- Modeling and Simulation
- Human-Computer Interaction
Cite this
In-depth exploitation of noun and verb semantics to identify causation in verb-noun pairs. / Riaz, Mehwish; Girju, Roxana.
2014. Paper presented at 15th Annual Meeting of the Special Interest Group on Discourse and Dialogue, SIGDIAL 2014, Philadelphia, United States.Research output: Contribution to conference › Paper
}
TY - CONF
T1 - In-depth exploitation of noun and verb semantics to identify causation in verb-noun pairs
AU - Riaz, Mehwish
AU - Girju, Roxana
PY - 2014
Y1 - 2014
N2 - Recognition of causality is important to achieve natural language discourse understanding. Previous approaches rely on shallow linguistic features. In this work, we propose to identify causality in verbnoun pairs by exploiting deeper semantics of nouns and verbs. Particularly, we acquire and employ three novel types of knowledge: (1) semantic classes of nouns with a high and low tendency to encode causality along with information regarding metonymies, (2) data-driven semantic classes of verbal events with the least tendency to encode causality, and (3) tendencies of verb frames to encode causality. Using these knowledge sources, we achieve around 15% improvement in Fscore over a supervised classifier trained using linguistic features.
AB - Recognition of causality is important to achieve natural language discourse understanding. Previous approaches rely on shallow linguistic features. In this work, we propose to identify causality in verbnoun pairs by exploiting deeper semantics of nouns and verbs. Particularly, we acquire and employ three novel types of knowledge: (1) semantic classes of nouns with a high and low tendency to encode causality along with information regarding metonymies, (2) data-driven semantic classes of verbal events with the least tendency to encode causality, and (3) tendencies of verb frames to encode causality. Using these knowledge sources, we achieve around 15% improvement in Fscore over a supervised classifier trained using linguistic features.
UR - http://www.scopus.com/inward/record.url?scp=84977865367&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84977865367&partnerID=8YFLogxK
M3 - Paper
AN - SCOPUS:84977865367
ER -