Wheat or chaff? A Compound Selection Model Based on Look-Up Data

Abstract

Which compounds should be included in general-purpose dictionaries is often an open question that is answered with a case-by-case consideration of all compounds above a certain corpus frequency threshold. Another way to determine which compounds should be listed, is to examine which compounds, or rather which compound properties, are in demand by the users. This study uses look-up data from the two officially sanctioned, general-purpose dictionaries of Norwegian (Bokmålsordboka and Nynorskordboka) to derive an explicit compound selection model that performs with comparable sensitivity and specificity as the traditional procedure. These findings demonstrate that it is indeed possible to arrive at a fully operational and explicit compound selection model that meets the needs of users. With such a tool at their disposal, lexicographers would be able to separate the wheat from the chaff in the boundless field that is the compound lexicon of North Germanic Languages.