An Empirical Study on Multi-Task Learning for Text Style Transfer and Paraphrase Generation

Bujnowski, Paweł; Ryzhova, Kseniia; Choi, Hyungtak; Witkowska, Katarzyna; Piersa, Jarosław; Krumholc, Tymoteusz; Beksa, Katarzyna

doi:10.18653/v1/2020.coling-industry.6

Cited by 3 publications

(4 citation statements)

References 33 publications

(37 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Without complying with any psychological models, Bujnowski et al (2020) addressed a task that could belong to this node in our hierarchy. Neutral sentences were transferred into “cute” ones, i.e., excited, positive, and slangy.…”

Section: Methodsmentioning

confidence: 99%

“…There exist many examples of this stream of research. Shang et al (2019) paraphrased old Chinese poems, Bujnowski et al (2020) and Carlson, Riddell, and Rockmore (2018) switched between the prose attributes of various versions of the Bible (“ Then Samuel gave him an account of everything, keeping nothing back ” “ And Samuel told all things, and did not hold back ”); Xu et al (2012), Jhamtani et al (2017), and He et al (2020) accounted for the features of Shakespearean plays, transferring Early Modern to contemporary English (“ I will bite thee by the ear for that jest ” “ I’ll bite you by the ear for that joke ”) or vice versa (“ Send thy man away ” “ Send your man away ”). A similar goal was addressed by Pang and Gimpel (2019) but with Dickens’ literature, while Krishna et al (2020) performed style transfer with different styles and attributes, transforming tweets into Shakespearean-like texts, Shakespearean texts into Joyce-sounding writings, al Joyce-authored texts into Bible-styled ones, and Bible verses into poems.…”

Section: Intended Stylesmentioning

confidence: 99%

See 1 more Smart Citation

From theories on styles to their transfer in text: Bridging the gap with a hierarchical survey

2022

View full text Add to dashboard Cite

Humans are naturally endowed with the ability to write in a particular style. They can, for instance, rephrase a formal letter in an informal way, convey a literal message with the use of figures of speech or edit a novel by mimicking the style of some well-known authors. Automating this form of creativity constitutes the goal of style transfer. As a natural language generation task, style transfer aims at rewriting existing texts, and specifically, it creates paraphrases that exhibit some desired stylistic attributes. From a practical perspective, it envisions beneficial applications, like chatbots that modulate their communicative style to appear empathetic, or systems that automatically simplify technical articles for a non-expert audience. Several style-aware paraphrasing methods have attempted to tackle style transfer. A handful of surveys give a methodological overview of the field, but they do not support researchers to focus on specific styles. With this paper, we aim at providing a comprehensive discussion of the styles that have received attention in the transfer task. We organize them in a hierarchy, highlighting the challenges for the definition of each of them and pointing out gaps in the current research landscape. The hierarchy comprises two main groups. One encompasses styles that people modulate arbitrarily, along the lines of registers and genres. The other group corresponds to unintentionally expressed styles, due to an author’s personal characteristics. Hence, our review shows how these groups relate to one another and where specific styles, including some that have not yet been explored, belong in the hierarchy. Moreover, we summarize the methods employed for different stylistic families, hinting researchers towards those that would be the most fitting for future research.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Intended Stylesmentioning

confidence: 99%

From theories on styles to their transfer in text: Bridging the gap with a hierarchical survey

2022

View full text Add to dashboard Cite

show abstract

“…Without complying with any psychological models, Bujnowski et al (2020) addressed a task that could fall in this category. Neutral sentences were transferred into "cute" ones, i.e.…”

Section: Personality Traitsmentioning

confidence: 99%

“…There exists many examples of this stream of research. Shang et al (2019), for instance, paraphrased old Chinese poems, Bujnowski et al (2020) and Carlson et al (2018), switched between prose styles of various versions of the Bible ("Then Samuel gave him an account of everything, keeping nothing back" → "And Samuel told all things, and did not hold back"); Xu et al (2012), Jhamtani et al (2017), andHe et al (2020) matched the style of Shakespearean plays, thus transferring Early Modern English into contemporary language ("I will bite thee by the ear for that jest" → "I'll bite you by the ear for that joke") or vice versa ("Send thy man away" → "Send your man away"). A similar goal is addressed by Pang and Gimpel (2019), but with Dickens' literature, while Krishna et al (2020) performed style transfer with different styles and attributes which include, but are not limited to, the transformation of tweets into Shakespearean-like texts, Shakespearean texts into Joyce's writings a , Joyce authored texts into Bible-styled ones, and Bible verses into poems.…”

Section: Literaturementioning

confidence: 99%

From Theories on Styles to their Transfer in Text: Bridging the Gap with a Hierarchical Survey

Troiano¹,

Aswathy²,

Klinger³

2021

Preprint

View full text Add to dashboard Cite

Humans are naturally endowed with the ability to write in a particular style. They can, for instance, rephrase a formal letter in an informal way, convey a literal message with the use of figures of speech, edit a novel mimicking the style of some well-known authors. Automating this form of creativity constitutes the goal of style transfer. As a natural language generation task, style transfer aims at re-writing existing texts, and specifically, it creates paraphrases that exhibit some desired stylistic attributes. From a practical perspective, it envisions beneficial applications, like chat-bots that modulate their communicative style to appear empathetic, or systems that automatically simplify technical articles for a non-expert audience. Style transfer has been dedicated several style-aware paraphrasing methods. A handful of surveys give a methodological overview of the field, but they do not support researchers to focus on specific styles. With this paper, we aim at providing a comprehensive discussion of the styles that have received attention in the transfer task. We organize them into a hierarchy, highlighting the challenges for the definition of each of them, and pointing out gaps in the current research landscape. The hierarchy comprises two main groups. One encompasses styles that people modulate arbitrarily, along the lines of registers and genres. The other group corresponds to unintentionally expressed styles, due to an author's personal characteristics. Hence, our review shows how the groups relate to one another, and where specific styles, including some that have never been explored, belong in the hierarchy. Moreover, we summarize the methods employed for different stylistic families, hinting researchers towards those that would be the most fitting for future research.

show abstract

A Study on Korean Language Honorific Translation using Transformers

Jeong¹,

Kim²,

Kim³

et al. 2021

jkiit

View full text Add to dashboard Cite

An Empirical Study on Multi-Task Learning for Text Style Transfer and Paraphrase Generation

Cited by 3 publications

References 33 publications

From theories on styles to their transfer in text: Bridging the gap with a hierarchical survey

From theories on styles to their transfer in text: Bridging the gap with a hierarchical survey

From Theories on Styles to their Transfer in Text: Bridging the Gap with a Hierarchical Survey

A Study on Korean Language Honorific Translation using Transformers

Contact Info

Product

Resources

About