Shakespearizing Modern Language Using Copy-Enriched Sequence to Sequence Models
By Harsh Jhamtani, Varun Prashant Gangal, Eduard H. Hovy, Eric Nyberg. 2017
190 citations recorded by the source corpus.
Variations in writing styles are commonly used to adapt the content to a specific context, audience, or purpose. However, applying stylistic variations is still by and large a manual process, and there have been little efforts towards automating it. In this paper we explore automated methods to transform text from modern English to Shakespearean English using an end to end trainable neural model with pointers to enable copy action. To tackle limited amount of parallel data, we pre-train embeddings of words by leveraging external dictionaries mapping Shakespearean words to modern English words as well as additional text. Our methods are able to get a BLEU score of 31+, an improvement of ≈ 6 points above the strongest baseline. We publicly release our code to foster further research in this area.
Subfields: Artificial Intelligence, Literature and Literary Theory, Machine Translation, Natural Language Generation
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
- Percy Liang
- Mohit Bansal
- Graham Neubig
- Luke S. Zettlemoyer
- Jianfeng Gao
- Yejin Choi
- Noah A. Smith
- Quoc V. Le
In Literature and Literary Theory
- Sudha Rao
- Joel R. Tetreault
- Percy Liang
- Eduard H. Hovy
- Anna Rogers
- Rada Mihalcea
- Anna Rumshisky
- Anders Søgaard
In Machine Translation
- Sudha Rao
- Joel R. Tetreault
- Percy Liang
- Rada Mihalcea
- Anders Søgaard
- Hal Daumé Iii
- Orhan Firat
- Jason Riesa
In Natural Language Generation
- Sudha Rao
- Joel R. Tetreault
- Ruslan Salakhutdinov
- Percy Liang
- Hal Daumé Iii
- Ehud Reiter
- Jiwei Li
- Rick Dale
Citations in the corpus
- Fair Transfer of Multiple Style Attributes in Text (2019, background): “Most of the existing style transfer techniques have focussed on individual and granular styles; for example, altering the sentiment of a review from negative to positive style, transferring a tweet from male to female style, and converting the style of text from one author to other [5], [7], [11], [”
- (Male, Bachelor) and (Female, Ph.D) have different connotations: Parallelly Annotated Stylistic Language Dataset with Multiple Personas (2019, background): “Style transfer with parallel corpus: Few recent works use parallel text for style transfer be-tween modern and Shakespearean text (Jhamtani et al., 2017), sarcastic and literal tweets (Peled and Reichart, 2017), and formal and informal text (Heylighen and Dewaele, 1999; Rao and Tetreault, 2018).”
- Answer-based Adversarial Training for Generating Clarification Questions (2019, background): “…(Bahdanau et al., 2015; Luong et al., 2015), summarization (Nallapati et al., 2016), dialog (Bor-des et al., 2016; Li et al., 2016a; Serban et al., 2017), textual style transfer (Jhamtani et al., 2017; Rao and Tetreault, 2018) and question answering (Yin et al., 2016; Serban et al., 2016).”
- TextSETTR: Label-Free Text Style Extraction and Tunable Targeted Restyling (2021, methodology): “Examples can be seen in Jhamtani et al. (2017) and Carlson et al. (2018).”
- FRMT: A Benchmark for Few-Shot Region-Aware Machine Translation (2022, background): “Earlier work in this space leverages supervised parallel data (Jhamtani et al., 2017).”
- Adversarial Decomposition of Text Representation (2018, methodology): “th the losses from the discriminator and the motivator: L total( E; G) = L rec +L adv +L motiv 4.5 Comparison with previous work In contrast to the proposals ofXu et al. (2012),Carlson et al.(2017),Jhamtani et al. (2017), our solution does not require a parallel corpus. Furthermore, unlike the model”
- Deep Learning for Text Style Transfer: A Survey (2020, methodology): “A possible fix to consider is to combine BLEU with PINC (Chen and Dolan 2011) as in paraphrasing (Xu et al. 2012; Jhamtani et al. 2017).”
- Text Style Transfer: A Review and Experiment Evaluation (2020, methodology): “[25] trained a Seq2Seq model with a pointer network on a parallel corpus and applied the model to translate modern English phrases to Shakespearean English.”
- Challenges and Strategies in Cross-Cultural NLP (2022, background): “For example, Jhamtani et al. (2017) transform text from modern English to Shakespearean English.”
- Dear Sir or Madam, May I Introduce the GYAFC Dataset: Corpus, Benchmarks and Metrics for Formality Style Transfer (2018, methodology): “, 2012) and neural machine translation (NMT) (Jhamtani et al., 2017) models to the task.”
- IMaT: Unsupervised Text Attribute Transfer via Iterative Matching and Translation (2019, background): “In the past few years, attribute-controlled NLG has witnessed renewed interest by researchers working on neural approaches to generation (Hu et al., 2017; Jhamtani et al., 2017; Melnyk et al., 2017; Mueller et al., 2017; Zhang et al., 2018; Prabhumoye et al., 2018; Niu and Bansal, 2018).”
- StylePTB: A Compositional Benchmark for Fine-grained Controllable Text Style Transfer (2021, background): “, 2020), writing styles (Jhamtani et al., 2017; Syed et al., 2020; Jin et al., 2020) and some other styles (Kang and Hovy, 2019).”