background image
DigiCULT
.
Info
13
T
RIS (Trials Support) co-ordinates
and promotes the results of twenty-
five recently completed EC-funded trial
projects which tested the application of
new technologies to enhance access to
collections held in museums, libraries, and
archives.TRIS has published material to
publicise the results of these projects on its
Website at: http://www.trisweb.org/tris/
trisportalpro/material/default.asp.The
Website is regularly updated with news
and events related to cultural heritage, and
has a new section on European and inter-
national projects which can be found at:
http://www.trisweb.org/tris/trisportal
pro/related/default.asp. Additions to this
section are welcome. Information on
TRIS can be accessed in English, French,
German, and Italian from: http://www.tris
web.org/tris/trisportalpro/tris/tris_proj
ect_sum.asp.
A
new document, "The Value Chain
for Digital Learning Content in
England", is now available for download-
ing from:
T
he Value Chain was created by
Simulacra (http://www.simulacrame
dia. com) for the Department for Educa-
tion and Skills in the UK, and is a con-
ceptual model that is intended to help
organisations to work out how they fit
into the process of creating, publishing,
managing, distributing and using digital
content. It aims to provide a commonly
understood model and language for the
Digital Learning Content market and rep-
resents the value that is added to digital
learning content through the key stages
from conception to use.The key roles
undertaken by organisations involved in
these stages are also examined.
V
ALUE
C
HAIN FOR
D
IGITAL
L
EARNING
C
ONTENT
LINK TO PDF DOCUMENT
BACK TO PAGE 1
BACK TO PAGE 1
T
RIALS
S
UPPORT FOR
E
UROPEAN
P
ROJECTS
previously restricted development and
research in this area. However, the compu-
tational demands of language processing
can now be met in many parts of the
world and the full potential of language
technologies is finally being realised.
H
LTs can be classified in various ways
but are generally divided into two
broad areas: natural language processing
(NLP) and speech processing. Broadly,
speech processing refers to the medium of
language the human ability to hear or
produce spoken language whereas NLP
refers to the message, that is, comprehen-
sion of the meaning of what is being said.
SPEECH PROCESSING
S
peech processing can be further subdi-
vided into different ways of recognising
and producing speech. For a computer to
convert spoken language into text, it needs
to isolate the words used (e.g. from back-
ground noise, speaker elision) and `recognise'
them, that is, acoustically analyse vocal
sounds and produce the appropriate signifier
of written language which represents it.This
requires the machine to not only `know' all
of the words being spoken, but also to be
able to `hear' correctly each individual word
and, in the case of homonyms, understand
which version of that sound (e.g. right,
write, rite) is being used. In addition to con-
verting speech into text, however, there are
additional analyses that can be performed on
vocal input, such as speaker recognition/ver
ification, identification and classification.
S
peaker recognition/verification
:
this technology allows a computer to
verify the identity of a person based on his
or her voice.The user typically repeats a
certain phrase and the computer compares
this with saved sound files either to identi-
fy which of its saved speakers is most likely
to be speaking (recognition) or to confirm
whether the speaker is a specific individual
(verification). Research continues into
tion inherent in disparate languages while
preserving the cultural significance of our
multilingual world.
T
he requirements of HLT (powerful
computer processing capacity, high
speed and widespread connectivity) have
I
n our global village, the convergence of
information society technologies has
created a demand for both universal com-
munication and the infrastructure that can
support it. Human Language Technology
(HLT) is an umbrella term for the innova-
tion, research and development which can
help human beings to interact with and
through computers in a more natural way,
that is, using their own languages.These
technologies can increase `understanding'
between humans and machines and can
help people communicate with one anoth-
er, removing the barriers to communica-
H
UMAN
L
ANGUAGE
T
ECHNOLOGIES
E
XPLAINED
"Human language technologies
will eventually allow anyone to use
a computer simply by talking and
listening to it."
D
AISY
A
BBOTT
DigiCULT Forum Researcher