Data were collected using the
YouTube Data Tools, with the use of the following search queries:
For each query, the first 200 videos were collected.
From
the collection of videos, they were removed:
→ Duplicate videos;
→ The
videos in languages other than Italian;
→
Videos that did not explicitly mention the urban environment or a
specific city;
→ Videos that
did not explicitly mention biodiversity in the title or
description.
The transcript of each video was generated through the use of the
VOSK Models
speech recognition tool.
Animal mentions were counted with
the help of
Videogrep.
The same tool was utlized to generate Supercuts.
The
list of animal mentions was then unified following the following
criteria:
→ Plurals and
singulars have been rendered in the singular;
→
Words such as 'etti' and 'ini' remained unchanged, e.g.
'tortoiseshell' and 'sparrow';
→
The feminine and masculine names remained separate;
→
The contracted forms such as 'to the eagle' and 'of the eagle' have
been unified into 'eagle';
→
The contracted forms such as 'a bee,' 'of the bee,' and 'the bee' have
been unified into 'bee.'
→
Feminine forms such as 'hen' and 'hen' remained separate.