Semantic change computation: A successive approach

Abstract

The prevalence of creativity in the emergent online media language calls for more effective computational approach to semantic change. This paper advocates the successive view of semantic change and proposes a successive framework for automatic semantic change detection. The framework measures Word Status of a word in a time unit with entropy, forms a time series data with the Word Statuses obtained from successive time units, and applies curve-fitting to obtain change pattern over the time series data. Experiments with the framework show that change pattern, the speed of change in particular, can be successfully related to classical semantic change categories such as broadening, narrowing, new word coining, metaphorical change, and metonymic change. By transforming the task of semantic computation into change pattern detection, the framework makes a plausible platform for semantic change investigation. © Springer International Publishing Switzerland 2013.