Full Title: Analysis of Linguistic VAriation for BEtter Tools
Short Title: ALVABET

Date: 09-Sep-2024 - 14-Apr-2024
Location: Paris, France
Contact Person: Mathieu Dehouck
Meeting Email: [email protected]
Web Site: https://llcd2024.sciencesconf.org/resource/page?id=18&forward-action=page&forward-controller=resource&lang=en

Linguistic Field(s): Computational Linguistics; Historical Linguistics; Morphology; Syntax; Text/Corpus Linguistics

Call Deadline: 30-Apr-2024

Meeting Description:

LLcD (Language and Languages at the crossroad of Disciplines) is a new event for linguists, NLP practitioners, and any researcher interested in language and languages to meet up and exchange.

ALVABET is a workshop collocated with the event.

The aim of this workshop is to question the cross-contributions of Natural Language Processing and variation analysis in the fields of morphosyntax and syntax.

Call for Papers:

Deadline Extension : April 30th

Call for Papers: Analysis of Linguistic VAriation for BEtter Tools (ALVABET) within the LLcD 2024 Conference (https://llcd2024.sciencesconf.org/)

Workshop

Variation plays a particularly important role in linguistic change, since every change stem from a state of variation; but each state of variation does not necessarily end up with a change: the new variant can disappear, or variation can linger but in different contexts. Access to sufficient amounts of data and their quantification, in order to detect the emergence of new variants as precisely as possible, and the recession or even disappearance of others, is a precious tool for the study of variations, whatever their dimensions (diachronic, diatopic, …) and in whatever field (syntax, morphology, …). The appearance of large corpora has thus renewed the study of variation. NLP has contributed largely to this renewal, providing tools for the enrichment and the exploration of these corpora. In return, linguistic analysis can help explain some of these errors and thus deepen the picture where performance metrics tend to flatten out everything under a single number, or even help improve the performances.

NLP annotation tools, such as syntactic parsers and morphological taggers, reach great performances nowadays when they are applied on similar data to those seen during their development. However, they quickly drop as the target data diverges from those of the training scenario. This raises a number of issues when it comes to using automatically annotated data to perform linguistic studies.

This workshop aims at exploring bilateral contributions between Natural Language Processing and variation analysis in the fields of morphosyntax and syntax, from diachronic and diatopic perspectives but also from genre, domain or form of writing, without any restriction on the languages of interest.

We warmly welcome submissions dealing with the issues and contributions of applying NLP to variation analysis :

• Quantification of variation along its different dimensions (both external and internal ones as well as in interaction with each other);

• Impact of annotation errors on the study of marginal structures (emergent or recessing);

• Syntactic variation when it is induced by semantic changes.

But also submissions dealing with the contributions of variation analysis to NLP:

• Variation mitigation (spelling standardisation...);

• Domain adaptation (domain referring here to any variation dimension);

• Error analysis (in and out of domain) in light of known variation phenomena, amongst which (de-)grammaticalisation;

• The evolution of grammatical categories and its impact on prediction models;

• The place of variation studies in NLP in the large language model era.

These themes are only suggestions and the workshop will gladly host any submission that deals substantially with the reciprocal contributions between NLP and variation analysis in the mentioned fields.

Full workshop description: https://llcd2024.sciencesconf.org/data/pages/WS12Eng.pdf

Important Dates

Apr 30, 2024: deadline for abstract submission for the workshop

May 15, 2024: Notification

Sep 9-11: Conference

Submissions

Abstracts must clearly state the research questions, approach, method, data and (expected) results. They must be anonymous: not only must they not contain the presenters' names, affiliations or addresses, but they must avoid any other information that might reveal their author(s). They should not exceed 500 words (including examples, but excluding bibliographical references).

Abstracts will be assessed by two members of the Scientific Committee and (one of) the workshop organizers.

On behalf of the ALVaBeT organizing committee,