An Italian corpus of clinical speech data

Corpus di Parlato Patologico Italiano in Pazienti Affetti da Schizofrenia Farmacoresistente (CIPP-TRS)

Abstract

The Corpus di Parlato Patologico Italiano in Pazienti Affetti da Schizofrenia Farmacoresistente (CIPP-TRS) is a spoken language corpus consisting of semi-spontaneous speech samples collected from adults diagnosed with treatment-resistant schizophrenia, non-treatment-resistant schizophrenia, and healthy controls. The corpus includes approximately 15 minutes of recorded interviews per participant, elicited through semi-structured interviews, for a total of over seven hours of annotated audio data. All speakers are adult native speakers of Italian, with recordings reflecting a Neapolitan regional variety. The dataset is enriched with detailed sociolinguistic metadata, including age, years of schooling, and clin- ical status, enabling investigations into language variation across adulthood and young old age. Orthographic transcriptions are aligned with audio and annotated in XML-TEI format, allowing systematic analysis of disfluencies, pauses, peri- and paralinguistic phenomena, and other features of spoken interaction. The CIPP-TRS corpus is designed to support linguistic, phonetic, and interactional research, with particular relevance for studies on aging, clinical populations, and spoken language variability. In line with FAIR data principles (Wilkinson et al. 2016), the corpus is structured to ensure transparent metadata documentation, standardized encoding (XML-TEI), and clear licensing conditions to facilitate discoverability, accessibility, interoperability across platforms, and reuse in future research.