Rethinking Text Attribute Transfer: A Lexical Analysis

TL;DR

This paper introduces a lexical analysis framework called Pivot Analysis to understand how specific words influence text attribute transfer, revealing that many models mainly modify key pivot words rather than sentence structure.

Contribution

The paper proposes a novel lexical analysis method to interpret text attribute transfer, highlighting the pivotal role of certain words and exposing the lexical-level focus of existing models.

Findings

Pivot words are strong features for attribute classification.

Changing pivot words can alter sentence attributes.

Most transfer models perform lexical modifications without changing sentence structure.

Abstract

Text attribute transfer is modifying certain linguistic attributes (e.g. sentiment, style, authorship, etc.) of a sentence and transforming them from one type to another. In this paper, we aim to analyze and interpret what is changed during the transfer process. We start from the observation that in many existing models and datasets, certain words within a sentence play important roles in determining the sentence attribute class. These words are referred to as \textit{the Pivot Words}. Based on these pivot words, we propose a lexical analysis framework, \textit{the Pivot Analysis}, to quantitatively analyze the effects of these words in text attribute classification and transfer. We apply this framework to existing datasets and models and show that: (1) the pivot words are strong features for the classification of sentence attributes; (2) to change the attribute of a sentence, many datasets only requires to change certain pivot words; (3) consequently, many transfer models only perform the lexical-level modification, while leaving higher-level sentence structures unchanged. Our work provides an in-depth understanding of linguistic attribute transfer and further identifies the future requirements and challenges of this task\footnote{Our code can be found at https://github.com/FranxYao/pivot_analysis}.