sophia.ananiadou AT manchester.ac.uk, nathalie.friburger AT univ-tours.fr, sophie.rosset AT limsi.fr
We wish to invite papers on new research relating to Named Entities (NEs), their recognition (NER) or the extraction of NE relations (NERelX), one of the most widely studied areas in information extraction as they are useful for several NLP applications (information retrieval, question-answering, machine translation, summarisation, etc.). NEs include proper nouns but also entities expressed through other nominal expressions, such as multi-word units, classified into types which may be coarse or fine-grained according to domain or user requirements. Despite years of research, NER still includes several challenges, such as correct classification, resolution of ambiguity, synonym detection, coreference and variability (e.g., acronyms, orthography). Several methods have been used to improve the prediction of correct classes, ranging from rule-based and dictionary-based approaches, to semi-supervised, and unsupervised machine learning techniques. Evaluation depends on the existence of gold (or silver) standards and domain specificity (e.g. genes, proteins, symptoms in health records, etc). Evaluation of NREelX is even more complex when considered in the end-to-end case. How can we evaluate NERelX while taking into account the errors coming from previous analysis steps?
We invite submissions on all topics relating to NE, NER and NE relation extraction, including:
Definition and typology of NEs, multi-word units
Domain and document adaptation methods in NER (abstracts, full papers, wikipedia, domain specific documents, new social media like twitter, online threads, spoken documents, etc.)
Detecting NE spans and structural analysis of NEs (NE parsing)
Cross-document coreference and entity linking
NE Tracking through time, social or geographical groups, intra- and inter-document NE tracking etc.
Normalisation aspects of NE (coreference, disambiguation)
Recognising NEs in general language and special domains
Guidelines and annotation tools of NE resources, NE corpora
Cross-language aspects in NE extraction
Evaluation, comparison and critical assessment of tools
NE and NLP applications dependant or based on NEs
TO NOTE
IMPORTANT DATES
April 15, 2013: Deadline for abstract
April 30, 2013: Deadline for submission
July 2013: Notification to authors
Autumn 2013: Publication
THE JOURNAL
TAL (Traitement Automatique des Langues / Natural Language Processing) is an international journal published by ATALA (French Association for Natural Language Processing, http://www.atala.org) since 1959 with the support of CNRS (National Centre for Scientific Research). It has moved to an electronic mode of publication, with printing on demand.
