Attentive Tree-structured Network for Monotonicity Reasoning

Zeming Chen


Abstract
Many state-of-art neural models designed for monotonicity reasoning perform poorly on downward inference. To address this shortcoming, we developed an attentive tree-structured neural network. It consists of a tree-based long-short-term-memory network (Tree-LSTM) with soft attention. It is designed to model the syntactic parse tree information from the sentence pair of a reasoning task. A self-attentive aggregator is used for aligning the representations of the premise and the hypothesis. We present our model and evaluate it using the Monotonicity Entailment Dataset (MED). We show and attempt to explain that our model outperforms existing models on MED.
Anthology ID:
2021.naloma-1.3
Volume:
Proceedings of the 1st and 2nd Workshops on Natural Logic Meets Machine Learning (NALOMA)
Month:
June
Year:
2021
Address:
Groningen, the Netherlands (online)
Venues:
IWCS | NALOMA
SIG:
SIGSEM
Publisher:
Association for Computational Linguistics
Note:
Pages:
12–21
Language:
URL:
https://aclanthology.org/2021.naloma-1.3
DOI:
Bibkey:
Cite (ACL):
Zeming Chen. 2021. Attentive Tree-structured Network for Monotonicity Reasoning. In Proceedings of the 1st and 2nd Workshops on Natural Logic Meets Machine Learning (NALOMA), pages 12–21, Groningen, the Netherlands (online). Association for Computational Linguistics.
Cite (Informal):
Attentive Tree-structured Network for Monotonicity Reasoning (Chen, NALOMA 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.naloma-1.3.pdf
Data
HELPMEDMultiNLISNLI