Recognition and normalization of temporal expressions in Serbian medical narratives

Abstract

The temporal dimension emerges as one of the essential concepts in the field of medicine, providing a basis for the proper interpretation and understanding of medically relevant information, often recorded only in unstructured texts. Automatic processing of temporal expressions involves their identification and formalization in a language understandable to computers. This paper aims to apply the existing system for automatic processing of temporal expressions in Serbian natural language texts to medical narrative texts, to evaluate the system’s efficiency in recognition and normalization of temporal expressions and to determine the degree of necessary adaptation according to the characteristics and requirements of the medical domain.