SIG:CMC/Technical Meeting on CMC at DARIAH VCC 2014

NOTE: The content of this page is preliminary and tentative. The pages describes a proposal and tentative program for a technical meeting on issues related with the modeling of CMC corpora submitted by members of the CMC-SIG for the 4th DARIAH-EU VCC meeting 2014 in Rome.

= PROPOSAL =

Proposal for a technical meeting on the topic

Models and tools for structuring & annotating corpora of social media / computer-mediated communication

to be held during the 4th General Virtual Competency Centre (VCC) meeting of DARIAH-EU, Rome, 17-19 September 2014

Corpora of computer-mediated communication (CMC) are a desideratum for many scholars in the Humanities who are interested in doing empirical research of language use and of emerging communicative genres on the Internet and in social media applications. Important steps for building such corpora and for representing them in an interoperable way are


 * to create models for the representation of CMC (including text-only as well as multimodal genres) which comply with a standard or de-facto standard in the field of Humanities;
 * to adapt tools for boilerplate removal and natural language processing to the structural and linguistic peculiarities of these genres (e.g., tools for converting raw CMC data into a standard format; tools for linguistic processing such as tokenizers, POS taggers, parsers etc.).

Researchers at a European level are already aware that many of the challenges in building CMC corpora in the Humanities are the same for every language; therefore CMC corpus projects for different languages can benefit from sharing knowledge and experiences with each other and from facing the challenges as a joint task. Since 2013, a group of corpus projects from France, Germany, Italy and the Netherlands has started to exchange expertise and experiences in building CMC corpora (= the network "Buildng and annotating CMC corpora", https://wiki.itmc.tu-dortmund.de/cmc/) and to jointly work on a proposal for an extension to the TEI standard which is adapted to the particularities of a broad range of CMC genres (= the TEI special interest group on CMC, http://www.tei-c.org/Activities/SIG/). The DARIAH technical meeting will gather a restricted number a researchers, coming from different European countries, involved in projects aiming at building, structuring and annotating CMC corpora – including:


 * CoMeRe: Project for a corpus of French CMC: http://comere.org
 * Dortmund Chat Corpus: http://www.chatkorpus.tu-dortmund.de
 * DeRiK: project for a reference corpus of German CMC: www.tinyurl.com/derik-llc
 * SoNaR ?
 * Web2CorpusIT, Pilot Corpus of Italian Computer-Mediated Communication: http://www.glottoweb.org/web2corpus/
 * 'Wikipedia corpora in DeReKo / IDS Mannheim: www.ids-mannheim.de/dereko

The expected outcomes of the meeting are, amongst other:


 * an advanced proposal for representing CMC genres in TEI (which subsequently shall be presented as a proposal to the TEI community in 2015);
 * a first exchange about experiences in automatically structuring and processing CMC data and a concept for a common platform (to be set up in 2015) for the documentation and exchange of NLP tools and annotation experiments with other projects and research groups interested in building CMC corpora in different languages;
 * plans for international scientific events (extended workshops, conferences) around these topics in 2015/16.

=TENTATIVE SCHEDULE / PROGRAM=


 * September 17, 2014: short intro/presentation (3-5 mins) on the work of our community
 * September 18, 2014: technical meeting (1,5 + 1,5 + 2 hours).

'''Precise times for the meeting and presentations will be fixed after DARIAH weill have decided about the proposal. All titles given below are working titles.'''

Technical meeting pt. I: Current state of building & modeling corpora (1,5 hrs)

 * A presentation from CoMeRé ?
 * A presentation by Isabella and/or Isabelle and Axel ?
 * Harald Luengen & Eliza Margareta (IDS Mannheim): Experiences with modeling Wikipedia corpora in TEI
 * N.N.

Technical meeting pt. II: NLP for CMC / social media corpora (1,5 hrs)
(using the Wikipedia corpora of the IDS)
 * members of the KobRA project (M. Beißwenger/C. Pölitz) on adapting machine learning methods for the analysis and annotation of CMC features in Wikipedia corpora
 * a presentation from someone who's dealing with "NLP4CMC" issues in the French community?
 * ? A report from the project of preparing a shared task on the linguistic annotation of CMC (= intermediate results from developing a tagset and guidelines for POS tagging of German CMC) (members of the Empirikom network)
 * N.N.

Technical meeting pt. III: schedule for further work on standards and joint scientific activities (2 hrs)

 * schedule for extending the standards (e.g., the next step in the TEI-CMC SIG)
 * plans for scientific events in 2015-16 focused on annnotaons, NLP4CMC, work on wikipedia & other genres
 * platform for the documentation and exchange of NLP tools and annotation experiments?