Text analysis for automatic identification of definitional linguistic markers in Costa Rican gastronomy recipes

Abstract

The analysis of definitional contexts allows to classify and systematize the definitional information belonging to a specific domain, and then to identify standards for the forms in which words and terms are defined in this domain. This paper describes the process implemented to automate the analysis of definitional contexts in the gastronomy domain in Costa Rica. The automation was done by using computational tools for natural language processing. The automation enables analysis of large quantities of data and results in less time than required by manual analysis. Automation consists of two modules, the first one is for the classification of documents in texts with or without recipes and the second one is for the identification of recipe ingredients based on formal linguistic patterns. Key Words: linguistic analysis of recipes, analysis of definitional contexts, definitional patterns, definitional markers, natural language processing.