Search within the transcribed texts and listen to the voices of the speakers
A new resource for spoken Italian
The KIParla Corpus collects more than 150 h of conversations, recorded in Bologna, Turin and other cities, transcribed and aligned with audio. The search interface allows results to be filtered by sociodemographic parameters, such as age, gender, city of origin, and educational qualification.
Different speakers, different situations
What speakers? Young people, adults and the elderly with different regional backgrounds, more and less educated. What situations? Semi-structured interviews, spontaneous conversations, table conversations, lectures, exams, student reception.
Texts and voices: who is speaking?
The corpus can be queried through the NoSketchEngine interface. The results of language searches are aligned with the audio: not only will you find what you are looking for, you can listen to it! Transcripts are anonymized, but speaker and situation metadata are accessible as search filters. Access the corpus!
A project that grows in time and space
The corpus is designed to grow over time, increasing the collection points and types of communicative situations represented: navigate through the modules. If you would like to collaborate with us, please contact us!