2nd Workshop on Stylistic Variation
Our workshop will be collocated with NAACL HLT 2018, taking place in New Orleans, USA from June 1 - June 6, 2018.
The workshops will be held on June 5 and June 6.
9 March 2018: Paper Due Date
2 April 2018: Notification of Acceptance
16 April 2018: Camera-ready Papers Due
People express themselves in different ways due to their background, their intended audience, the conventions of their language for the genre in question, or just as a matter of personal style.
The associated variation in phonological, lexical, syntactic, or discourse realisation of a particular semantic content have important consequences for low-level NLP tasks such as text normalization, POS tagging, and parsing. At the same time, this variation influences downstream applications such as text simplification, sentiment analysis, information retrieval, or text generation, and directly enables the reverse tasks of predicting variables such as individual speaker, speaker demographics, target audience, genre, language, etc.
Leveraging these stylistic differences in NLP can be beneficial for machines and humans alike, while on the other hand opening a range of ethical questions. An informed discussion on these timely topics requires gathering at one table researchers who encounter stylistic variation directly or indirectly in their work - from the design of computational methods to the interpretation of results in other fields - and collecting and aggregating the current development, challenges and opportunities.
The overarching questions that motivate this workshop are:
1. To what extent it is possible or desirable to go beyond superficial, uninterpretable, task-specific stylistic features to deeper, broader, more systematic, and more psychologically-plausible conceptualizations of stylistic variation.
2. To what extent does stylistic variation indirectly impact applied tasks historically treated as stylistically uniform, and to what extent is an isolation of the stylistic factor desirable in these applications.
3. To what extent recent advances in related areas such as distributional semantics can be applied to better capture stylistic variation and which new challenges arise with using and interpreting these approaches.
For purposes of the workshop, “stylistic variation” includes variation in phonological, lexical, syntactic, or discourse realization of particular semantic content, due to differences in extralinguistic variables such as individual speaker, speaker demographics, target audience, genre and so on.
Call for Papers
We welcome submissions including the proposed topics and applications of interest from the following non-exhaustive list.
- Evidence for or against targeted approaches to stylistic variation
- Interpretability of computational models of style
- Effects of stylistic variation on downstream tasks
- General methods for differentiating style from semantics/topic
- Style-aware natural language generation
- Domain adaptation across stylistically distinct domains
- Capturing style in distributional vector space models
- Stylistic lexicon acquisition
- Challenges in the interpretation (and overinterpretation) of stylistic differences
- Speaker identification in text and speech
- Challenges of annotating style
- Quantification of genre differences
- Stylistic features for mental health applications
- Literary stylistics (author and character profiling)
- Rhetoric (e.g. stylistic choice in political speeches, etc.)
- Authorship attribution, stylistic segmentation, intrinsic plagiarism detection
- Identifying trustworthiness and deception
- Text normalization
- Modelling of demographics and personality
- Politeness and other linguistic manifestations of social power
- Stylistically-informed sentiment analysis (e.g. sarcasm, hate speech)
- Readability, complexity, and simplification
- Learner language (e.g. fluency, use of collocations, stylistic appropriateness, etc.)
We accept regular long (8 pages + references) and short papers (4 pages + references) following the NAACL 2018 format.
Submit your papers via START: https://www.softconf.com/naacl2018/Style-Var18/user/scmd.cgi?scmd=submitPaperCustom&pageid=0
Lucie Flekova, Amazon Research
Julian Brooke, Thomson Reuters
Thamar Solorio, University of Houston
Moshe Koppel, Bar-Ilan University
Nikolaos Aletras (Amazon Research)
Yves Bestgen (Université catholique de Louvain)
Alberto Barrón-Cedeño (Qatar Computing Research Institute)
Walter Daelemans (University of Antwerp)
Jacob Eisenstein (Georgia Tech)
Roger Evans (University of Brighton)
Alexander Gelbukh (Instituto Politécnico Nacional)
Adam Hammond (San Diego State University)
Graeme Hirst (University of Toronto)
Dirk Hovy (University of Copenhagen)
Eduard Hovy (Carnegie Mellon University)
Ekaterina Kochmar (Cambridge University)
Vasileios Lampos (University College London)
Dominique Legallois (Université Paris 3, Sorbonne-Nouvelle)
Manuel Montes-y-Gomez (Instituto Nacional de Astrofísica, Optica y Electronica)
Dong Nguyen (Alan Turing Institute)
Umashanthi Pavalanathan (Georgia Institute of Technology)
Ellie Pavlick (University of Pennsylvania)
Barbara Plank (University of Groningen)
Martin Potthast (Leipzig University)
Vinod Prabhakaran (Computer Science, Stanford)
Daniel Preotiuc-Pietro (University of Pennsylvania)
Emily Prud’hommeaux (Rochester Institute of Technology)
Sudha Rao (University of Maryland)
Paolo Rosso (Universitat Politècnica de València)
Maarten Sap (University of Washington)
Andrew H. Schwartz (Stony Brook University)
Anders Soegaard (University of Copenhagen)
Benno Stein (Bauhaus-Universität Weimar)
Joel Tetreault (Grammarly)
Sandra Uitdenbogerd (RMIT University)
Sowmya Vajjala (Iowa State University)
Svitlana Volkova (Pacific Northwest National Laboratory)
Wei Xu (Ohio State Unversity)
Marcos Zampeiri (University of Cologne)
Prof. Rada Mihalcea, University of Michigan
Prof. James W. Pennebaker, University of Texas at Austin
Previous year: https://sites.google.com/site/workshoponstylisticvariation/program