Enhancing and using electronic dictionaries

A SIGLEX-endorsed COLING workshop

Michael Zock (Limsi, CNRS) & Patrick St. Dizier (Irit, CNRS)

Sunday 29th August 2004

Introduction

A dictionary is a vital component of any natural language processing system. Its modern, digital form has considerable potential, especially if it is extended and built in a way compatible with the needs and habits of the average language user. There are many ways to make an electronic dictionary useful for people in their daily tasks of processing language. One could assist

1. reading and writing: adding a transliterator and a morphological generator/parser to a dictionary would put the needed information at the distance of a mouse click. Imagine someone trying to look up the meaning (or translation) of a word in a script he can’t read.
2. language learning: combining dictionaries with a parametrizable flashcard system and a goal-driven exercice generator could help the memorization and automation of words and basic syntactic patterns. In such a system choosing a goal would trigger syntactic templates, filling the latter with words would yield (simple) sentences.
3. lexical access (navigation): structuring the dictionary in a similar way as the human mind, i.e. building an associative network akin to WordNet, but with many more links in particular at the syntagmatic axis, could assist the writer not only in finding new ideas (brainstorming), but also the word s/he is looking for. Within this framework, word access amounts to entering and navigating in a huge, associative network. To build such a tool one could extract associations from an encyclopedia, label and add them as links to a resource like WordNet.

As one can see, there are numerous ways to enhance dictionaries. While the new hardware offers many, sometimes surprising opportunities for novel uses, seizing them requires some rethinking. This is the goal of this one-day workshop.

In particular, we’d like to discuss interesting extensions and enhancements of electronic dictionaries. For example, one could consider merging different, thesaurus-like dictionaries and see what kind of conceptual and navigational aids might be added to support the language user: what are his needs, what information is he looking for? Actually, a focus shift might be necessary to move from the data (content and size of the dictionary) to their organisation and access. As, what is a huge dictionary good for, if one cannot find the word one is looking for?

Target audience

The aim of this workshop is to bring together leading researchers involved in the building of electronic dictionaries to discuss modifications of existing resources in line with the users’ needs (i.e. how to capitalize on the advantages of the digital form). Given the breadth of the questions, we welcomed reports on work from many perspectives, including, but not limited, to linguistics, computer science, psycholinguistics, language learning, and ergonomics. We requested that each contribution addresses computational aspects.

Preliminary program

Back to the top of the page

Posters