Multi-Label Transfer Learning for Multi-Relational Semantic Similarity

TL;DR

This paper introduces a multi-label transfer learning method using LSTM to predict multiple semantic relations between texts simultaneously, outperforming traditional approaches and achieving state-of-the-art results on key datasets.

Contribution

The paper presents a novel multi-label transfer learning approach for multi-relational semantic similarity, enabling joint learning of multiple relations with improved performance.

Findings

Outperforms single-task and multi-task methods

Achieves state-of-the-art on most relations of the dataset

Effective multi-relational semantic similarity prediction

Abstract

Multi-relational semantic similarity datasets define the semantic relations between two short texts in multiple ways, e.g., similarity, relatedness, and so on. Yet, all the systems to date designed to capture such relations target one relation at a time. We propose a multi-label transfer learning approach based on LSTM to make predictions for several relations simultaneously and aggregate the losses to update the parameters. This multi-label regression approach jointly learns the information provided by the multiple relations, rather than treating them as separate tasks. Not only does this approach outperform the single-task approach and the traditional multi-task learning approach, but it also achieves state-of-the-art performance on all but one relation of the Human Activity Phrase dataset.