Attentive Tree-structured Network for Monotonicity Reasoning

TL;DR

This paper introduces an attentive tree-structured neural network that effectively models syntactic parse trees for improved monotonicity reasoning, especially on downward inferences, outperforming existing models on the MED dataset.

Contribution

The paper proposes a novel Tree-LSTM with soft attention and self-attentive aggregation to enhance monotonicity reasoning in neural models.

Findings

Outperforms existing models on the MED dataset

Effectively models syntactic parse trees for reasoning

Addresses shortcomings in downward inference performance

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.