Research on Language and Computation

Ambiguity and semantic judgments

Special issue edited by Massimo Poesio and Ron Artstein

Status (updated 25 September 2008)

The compiled issue has been sent to the publisher.

List of Articles:

Scope

This is the original call for papers. The issue is already in press, and we cannot accept new submissions.

We invite articles for a special issue on ambiguity and semantic judgments from a computational, theoretical and psychological perspective. Much research in computational linguistics assumes that tasks have a single answer: word sense disambiguation looks for an unambiguous sense in context, anaphora resolution algorithms look for a unique antecedent, question-answering systems look for the best answer, semantic role labeling identifies the most appropriate role, and so on. Yet theoretical and psychological evidence show that ambiguity is abundant, and semantic annotation tasks often display disagreements between coders which are the result of genuine ambiguity rather than annotation error.

We are interested in ambiguity, broadly defined. On the one hand, there are cases where ambiguities constitute clearly distinct interpretations, preserved despite the context. On the other hand, there are instances of underspecification which may or may not be construed as ambiguous given a context. And in between there may be cases where different modes of processing give rise to differences of emphasis which may or may not warrant classifying as ambiguities. All these shades of variation, and the disputes they give rise to, call for more empirical study of matters of ambiguity, especially as they pertain to semantic judgments used in corpus annotation and computational implementation.

For this special issue we are looking for high-quality, original, full-length journal articles on any aspect pertaining to ambiguity and semantic judgment. We especially welcome articles on the following topics:

Revised versions

This information is only relevant for authors who have been notified that their manuscript has been accepted for publication.

Style

Please consult the journal’s instructions for authors for matters of style.

Format

Please prepare your revised version in LaTeX, using the accompanying class and style files. These files are based on the old Kluwer classes, and probably retain some of the old bugs. There is limited documentation within the files, but we do provide an example submission as LaTeX source and PDF output. If you have trouble with the files or with using LaTeX in general, write to Ron.

Length

We don’t have final figures for the page counts, but we expect the page limit to be around 20 pages in the final format. We will update you when we have the final figures.

Links

The workshop on Ambiguity in Anaphora was held at ESSLLI 2006, 7–11 August 2006, Málaga, Spain.

The editors were involved in the Arrau project (Anaphora Resolution and Underspecification).