IMaT: Unsupervised Text Attribute Transfer via Iterative Matching and Translation

By Zhijing Jin, Di Jin, Jonas W. Mueller, Nicholas Matthews, Enrico Santus. 2019

77 citations recorded by the source corpus.

Text attribute transfer aims to automatically rewrite sentences such that they possess certain linguistic attributes, while simultaneously preserving their semantic content. This task remains challenging due to a lack of supervised parallel data. Existing approaches try to explicitly disentangle content and attribute information, but this is difficult and often results in poor content-preservation and ungrammaticality. In contrast, we propose a simpler approach, Iterative Matching and Translation (IMaT), which: (1) constructs a pseudo-parallel corpus by aligning a subset of semantically similar sentences from the source and the target corpora; (2) applies a standard sequence-to-sequence model to learn the attribute transfer; (3) iteratively improves the learned transfer function by refining imperfections in the alignment. In sentiment modification and formality transfer tasks, our method outperforms complex state-of-the-art systems by a large margin. As an auxiliary contribution, we produce a publicly-available test set with human-generated transfer references.

Subfields: Artificial Intelligence, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

Citations in the corpus

A Recipe for Arbitrary Text Style Transfer with Large Language Models (2021, methodology): “Such methods include latent representation manipulation (Liu et al., 2020), prototype-based text editing (Li et al., 2018), and pseudo-parallel corpus construction (Jin et al., 2019).”
Encode, Tag, Realize: High-Precision Text Editing (2019, background): “…virtually every text generation task, including summarization (Tan et al., 2017), image captioning (Xu et al., 2015), text style transfer (Rao and Tetreault, 2018; Nikolov and Hahnloser, 2018; Jin et al., 2019), and grammatical error correction (Chollampatt and Ng, 2018; Grundkiewicz et al., 2019).”
Deep Learning for Text Style Transfer: A Survey (2020, methodology): “Jin et al. (2019) empirically observe that semantically similar sentences in the two mono-style corpora tend to be the attribute-transferred counterparts of each other.”
Style Transformer: Unpaired Text Style Transfer without Disentangled Latent Representation (2019, methodology): “Most of the previous methods (Hu et al., 2017; Shen et al., 2017; Fu et al., 2018; Carlson et al., 2017; Zhang et al., 2018b,a; Prabhumoye et al., 2018; Jin et al., 2019; Melnyk et al., 2017; dos Santos et al., 2018) formulate the style transfer problem into the “encoder-decoder” framework.”
Civil Rephrases Of Toxic Texts With Self-Supervised Transformers (2021, methodology): “2019) and to further conﬁrm the performance of CAE-T5, we hired human annotators on Appen to rate in a blind fashion different models’ civil rephrasings of 100 randomly selected test toxic comments, in terms of attribute transfer (Att), ﬂu-ency (Flu), content preservation (Con) and overall quality…”
Text Style Transfer: A Review and Experiment Evaluation (2020, methodology): “Another interesting approach is to generate pseudo-parallel datasets and apply seq2seq models to perform TST. Jin et al. [27] first constructed a pseudo-parallel corpus by matching text sentences in a source style corpus X with text sentences in target style corpus Y using cosine similarity.”
A Review of Human Evaluation for Style Transfer (2021, background): “The most common QC strategies are to require workers to pass a qual-iﬁcation test (Jin et al., 2019; Li et al., 2016; Ma et al., 2020; Pryzant et al., 2020) to hire the top-ranked workers based on pre-computed scores that reﬂect the number of their past approved tasks (Kr-ishna et al., 2020; Li et…”
PersonaLLM: Investigating the Ability of Large Language Models to Express Personality Traits (2023, background): “…et al., 2016; Tadesse et al., 2018), and personalized dialogue generation (Mairesse and Walker, 2007, 2011; Zhang et al., 2018; Qian et al., 2018) including stylistic transfer of personality traits such as formality and polite-ness (Kabbara and Cheung, 2016; Jin et al., 2019; Madaan et al., 2020).”