Lexical Access, a Search-Problem

Our work is confined to word access, that is, we present here our ideas of how to improve electronic dictionaries in order to help language producers (speaker/writer) to find the word they are looking for. Our approach is based on psychological findings (representation, storage and access of information in the human mind), observed search strategies and typical navigational behavior. If one agrees with the idea that lexical access (word finding) is basically a search problem, then one may still want to find out where and how to search. While the space, i.e. the semantic map in which search takes place is a resource problem,— any of the following could be used: dictionary, corpus, thesauraus, etc. or a mix of them,— its exploration is typically a search problem. Important as it may be, the building of a high quality resource is not the focus of this work, we rely on an existing one, and while we are concerned with its quality, we will be mostly concerned here with search methods, in order to determine the best. 1 Problem: find the needle in a haystack One of the most vexing problems in speaking or writing is that one knows a given word, yet one fails to access it when needed. This kind of search failure, often referred to as dysnomia or Tip of the Tongue-problem, occurs not only in communication, but also in other activities of everyday life. Being basically a search problem it is likely to occur whenever we look for something that exists in real world (objects) or our mind: dates, phone numbers, past events, peoples’ names, or you just name it. As one can see, we are concerned here with the problem of words, or rather, how to find them in the place where they are stored: the human brain, or an external resource, a dictionary. Our work being confined to lexical access, we would like to develp a semantic map and a compass to help language producers to find the word they are looking for. More precisely, we try to build an index and a navigational tool allowing people to access words no matter how incomplete their conceptual input may be. Our approach is based on psychological findings concerning the mental lexicon (Aitchison, 2003; Levelt et al., 1999), i.e. storage and access of information in the human mind, observed search strategies and typical navigational behavior. 2 Consider the following elements before attempting an engineering solution Before conceiving a roadmap leading to an engineering solution it may be useful to consider certain points. The list here below is by no means complete, neither is the following discussion. Nevertheless we believe that the following points are worth consideration: features of the mental lexicon, how to build and use the resource, searching, ranking and weights, interface problems. For reasons of space constraints we will touch briefly only upon some of these points. Our main goal is the enhancement of electronic dictionaries to help speakers or writers to find