Title:
Needles in a haystack and how to find them? The case of lexical access
Date:
24/02/2015
Town:
Pisa
Venue:
ILC-CNR – Aula Seminari IBF SG 5
Description:
Word access is an obligatory step in language production. In order to achieve his communicative goal, a speaker/writer needs not only to have something to say, he must also find the word(s) expressing the message he wants to convey. Yet, knowing a word, i.e. having it stored in a data-base (human mind, dictionary), does not mean that one is able to access it in time. This is a case where computers (electronic dictionaries) can be of great help, especially when the person is engaged in writing.
I will present in my talk some ideas of how a dictionary could be enhanced to support a writer to find the word he is looking for. To this end I suggest to add an index based on the notion of association – which is supposed to represent the average person’s knowledge of the world (encyclopedia) – to an existing electronic resource. One way of getting this kind of knowledge is to extract it from a well balanced corpus (to capture both general and episodic knowledge). An alternative is to draw on an existing Association Thesaurus like the E.A.T. (Edinburgh Association Thesaurus). In both cases we have a semantic network within which search takes place but clustering the results to avoid drowning the user under a huge (unstructured) list of words remains a problem.
I will present in my talk some ideas of how a dictionary could be enhanced to support a writer to find the word he is looking for. To this end I suggest to add an index based on the notion of association – which is supposed to represent the average person’s knowledge of the world (encyclopedia) – to an existing electronic resource. One way of getting this kind of knowledge is to extract it from a well balanced corpus (to capture both general and episodic knowledge). An alternative is to draw on an existing Association Thesaurus like the E.A.T. (Edinburgh Association Thesaurus). In both cases we have a semantic network within which search takes place but clustering the results to avoid drowning the user under a huge (unstructured) list of words remains a problem.
Programme:
- Introduction
Vito Pirrelli, ILC-CNR (5′) - Needles in a haystack and how to find them? The case of lexical access
Michael Zock, LIF-CNRS (40′) - Discussion (15′)
Promotional Material:
Presentations: