The name (Meletē)ToPān v.0.1 is based on the Greek principle μελέτη
τὸ πᾶν which roughly translate to "take into care everything". I decided
for the name because Topic-Modelling performs well on large amounts of
logically structured chunks of texts and it helps selecting the
interesting bits in a large corpus of text by technically having looked
at everything. The butterfly in the logo is of the species Melete. The
original photograph is by Didier Descouens and he has licensed it under
CC BY-SA 4.0. I changed the image for the logo slightly. I'd strongly
suggest to start with the original
if you want to use it, but you can also use this now slightly modified
logo under CC BY-SA 4.0 license as I am required to share it under the
same license as the original image
ToPān is Topic-Modelling for everyone: from people without
programming knowledge to people that want to build teaching and
text-reuse tools and apps based on Topic-Modelling data without having
to develop their own tool or having to majorly restructure their textual
data. ToPān is made to be shared and used. That is why I tried to
modularise ToPān in a way that in each step you could ingest your own
data. It works best however, if you work your way from left to right:
from "Data Input" to "LDA Tables" (please find more details under
"Instructions"). ToPān works best with files that are structured
according to the CTS/CITE architecture.
ToPān is also still under active development. This is an alpha
release. More features will be added and you are encouraged to roadtest
ToPān and send me feedback or report bugs.
No comments:
Post a Comment