EACL 2012 Joint Workshop of LINGVIS & UNCLH


Visualization of Linguistic Patterns


and


Uncovering Language History from Multilingual Resources


April 23-24, 2012, Avignon, France


Submission deadline: February 3, 2012



The overall aim of the workshop is to explore how methods developed in computational linguistics, statistics and computer science can help linguists in exploring various language phenomena. The workshop focuses particularly on two special subtopics: 
1) visualization of linguistic patterns (LINGVIS); 
2) usage of multilingual resources in computational historical linguistics (UNCLH). 


Visualization of Linguistic Patterns

The aim of the first subtopic of this joint workshop is to combine techniques developed in the vibrant fields of Information Visualization (InfoVis) and Visual Analytics with methodology and analyses from theoretical and computational linguistics in order to allow for a novel perspective on linguistic data and patterns. We aim to bring together researchers interested in combining methods and insights from the fields of Visual Analytics and Linguistics: despite the fact that statistical methods for language analysis have proliferated in the last two decades, computational linguistics has so far only marginally availed itself of techniques from InfoVis and Visual Analytics (e.g., Collins et al. 2009, Collins 2010, Honkela et al. 1995, Mayer et al. 2010a,b, Rohrdantz et al. 2011, Neumann et al. 2007). Besides standard visualization techniques such as bar charts, scatterplots or line charts, a large number of advanced novel methods have been developed within these fields. Prominent examples are treemaps, showing hierarchical data, pixel displays, or the sophisticated visualizations of graphs that are an intuitive and beneficial way of modeling interactions. The workshop aims to investigate how complex linguistic questions can profit from such visual analysis. 

We especially encourage submission on the following topics:
  • using visual analysis to probe change in language over time (historical linguistics) and differences in structure across languages (typology) 
  • experimenting with methods for the exploration of high dimensional spaces like vector spaces for analyses of, for example, lexical semantic or thematic information 
  • exploring of linguistic patterns in terms of integrating geo-spatial locations with hierarchical data structures and temporal dimensions 
  • designing multifactorial visual analyses of the interacting linguistic factors. 

Uncovering Language History from Multilingual Resources

The second subtopic of the joint workshop focuses on the usage of multilingual resources in computational historical linguistics. In the past 20 years, the application of quantitative methods in historical linguistics has received increasing attention among linguists (e.g. Dunn et al. 2005, Heggarty et al. 2010, McMahon and McMahon 2006), computational linguists (e.g. Kondrak 2001, Hall and Klein 2010), and evolutionary anthropologists (e.g. Gray and Atkinson 2003). Due to the application of these quantitative methods, the field of historical linguistics is undergoing a renaissance. One of the main problems that researchers face is the limited amount of suitable data, often falling back on relatively restricted 'Swadesh type' wordlists. One solution is to use synchronic data, like dictionaries or texts, which are available for many languages. For example, in Kondrak (2001), vocabularies of four Algonquian languages were used in the task of automatic cognate identification. Another solution employed by Snyder et al. (2010) is to apply a non-parametric Bayesian framework to two non-parallel texts in the task of text deciphering. Although very promising, these approaches have so far only received modest attention. Thus, many questions and challenges in the automatization of language resources in computational historical linguistics remain open and ripe for investigation. 

We especially encourage submissions related to the following topics: 
  • computational approaches that uncover sound correspondences and sound change 
  • automatic identification of cognates and/or loanwords across languages 
  • linguistically-informed n-gram comparisons, e.g. using flexible n-gram length or using more advanced sound similarities between languages 
  • comparison of Wordnet structures from different languages 
  • treatment of dictionaries as translation graphs and comparison of graph structures between dictionaries 
  • exploration of meaning shifts as instantiated through differences in usage in texts



Invited Speakers:

Christopher Collins (University of Ontario Institute of Technology)
Grzegorz Kondrak (University of Alberta) 

Important Dates:

Deadline for submission: February 3, 2012
Notification of acceptance: February 24, 2012
Revised version of papers: March 9, 2012
Workshop: April 23-24, 2012

Organizing Committee:

LINGVIS:
Miriam Butt (Universität Konstanz)        
Sheelagh Carpendale (University of Calgary)        
Gerald Penn (University of Toronto)                         

UNCLH:
Jelena Prokić (LMU Munich)
Michael Cysouw (LMU Munich)
Thomas Mayer (LMU Munich)
Steven Moran (LMU Munich)

Program Committee:

Quentin Atkinson (University of Auckland)
Christopher Collins (University of Ontario)
Chris Culy (University of Tübingen)
Dan Dediu (MPI Nijmegen)
Michael Dunn (MPI Nijmegen) 
Sheila Embleton (York University, Toronto) 
Simon Greenhill (University of Auckland)
Harald Hammarström (University of Nijmegen) 
Wilbert Heeringa (Meertens Institute, Amsterdam)
Gerhard Heyer (University of Leipzig)
Eric Holman (UCLA) 
Gerhard Jäger (University of Tübingen) 
Daniel Keim (University of Konstanz)
Tibor Kiss (University of Bochum)
Jonas Kuhn (University of Stuttgart)
John Nerbonne (University of Groningen) 
Anke Lüdeling (Humboldt University, Berlin) 
Don Ringe (University of Pennsylvania) 
Hinrich Schütze (University of Stuttgart)
Tandy Warnow (University of Texas at Austin) 
Søren Wichmann (EVA MPI, Leipzig) 

Contact:
LINGVIS: Annette Hautli (annette.hautli@uni-konstanz.de)
UNCLH: Jelena Prokić (unclh2012@gmail.com)

Submission Instructions:

All submissions must be submitted electronically as PDF via the EACL submission system

https://softconf.com/eacl2012/LINGVIS-UNCLH/

All papers must follow the two-column format of EACL proceedings. Authors are strongly recommended to use the style files available on the conference web site

http://eacl2012.org/information-for-authors/index.html

Papers may consist of up to eight (8) pages of content and any number of additional pages containing references only. We invite different submission modalities: 
(i) regular long papers (8 content pages + 1 page for references); 
(ii) short papers (4 content pages + 1 page for references). 
In addition, authors can specify whether they also want to be considered for poster presentations.

As the reviewing will be blind, papers must not include the authors' names and affiliations.