UHH
>
Informatik
>
NatS
>
ProSub16 Web
>
FromTranscriptToSubtitles
(14 Apr 2016,
ArneKoehn
)
P
rint version
content-unaware
max. ~80 / 110 Characters (Length)
Punctuation at the end of the sentence -> expectation/hint for the next sentence.
at the start too? i did... ...this
"Als oberstes Verfassungsorgan hat dieses Gericht nämlich:"
"Anteil an der obersten Staatsleitung,"
"Das heißt nicht, dass es damit die Gewaltenfrage ignorieren dürfte," [...]/ ...
"und sich in Kernkompetenzen der Legislative"
"oder der Exekutive einmischen könnte."
delete fillers
syntax
split when a clause is over/part of a sentence -> meaningful subtitles
add punctuation
un-delete fillers (keep the ones that actually help if they do)
more difficult stuff
capitalization
"meine Damen und Herren" deleted (no information)
[OFF-TOPIC] mark/delete/both
[AUDIENCE QUESTION] else its just a weird pause (mark)
rearrange words (more or less) -> problematic if you hear & understand the speaker
repairs: correct smaller ones but not larger ones
timing
span subtitles to show them all the time
--
ArneKoehn
- 14 Apr 2016
ProSub16
Log In
Navigation
Research
Teaching
Publications
Download
People
Jobs
Contact
Changes
Courses
-
CIT 831 Natural Language Processing
More
Research
Constraint Dependency Grammar
Hamburg Dependency Treebank
Spoken Wikipedia Corpus
Construction Grammar
NatsWiki
Main
User
Sandbox
System
Copyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki?
Send feedback