|
|
| Communications
| IVR and Voice
Portals | Custom solutions | In-IVR
| In-IVR™- Demo|
|
 |
In-IVR™- Demo
|
There are some already
developed services that demonstrate the In-IVR™ platform
operation. Here you can download the corresponding audio files.
|
|
1) Horoscopes
|
This
service provides to the caller updated information on various
Zodiac signs. The system asks the user what sign the person
wishes to know about, receiving the answer (for example "Leo",
or "Aquarius") by voice.
The way the questions are asked to the callers
changes, so that the usage or clarifications the machine requests
about what has been said does not become tedious.
The technologies used to obtain up-to-date
contents from the SAPO Internet site of known astrologist
Paulo Cardoso are ASR,
TTS and IIWC™.
The texts of the page are converted automatically by a speech
synthesiser.
The IBE™
technology cam can be also used in this service, but for the
demonstration purposes, taking into account a possibility
of using telephones with loudspeakers, we chose not to use
it to make sure it would not interfere with ASR.
For users that do not use loudspeakers the inclusion of IBE™
would not interfere with ASR.
[back]
|
2) Questionnaires
 |
This
service allows the callers to answer a questionnaire about
their "psychological profile". The callers answer
the questions by DTMF,
and in the end they get a score that corresponds to a given
psychological profile which is spoken to them via IVR.
The technologies used for getting the question/answer content
from the Web are DTMF,
TTS, IBE™
and IIWC™. The
texts of the page are converted automatically by a speech
synthesiser. [back]
|
3) Weather
Forecast
|
This
service allows the users to know the weather forecasts for
a given location at their choice. For that they choose a region
by DTMF, and then a city.
It is possible to associate localisation platforms to this
service so that the users are not obliged to navigate through
menus. ASR can also be used for the choice of the city.
The technologies used to obtain the contents from the Web
site are DTMF, TTS,
IBE™ and IIWC™.
The texts of the page are converted automatically by a speech
synthesiser. [back]
|
4) Traffic
Information
|
This
service provides a constantly updated traffic report from
a Web site.
Whenever there is an accident or a traffic
jam, the service includes the pertinent information and makes
it available from the menu, so that the user can get more
details about occurrence. This helpful service justifies well
a small amount of money spent on the call, being especially
useful during the rush hours (of course, it must be used with
an earpiece or a "hands-free" adaptor).
The technologies used are DTMF,
TTS, IBE™,
and IIWC™ to get
the data from a web site. The texts of the page are converted
automatically by a speech synthesiser. [back]
|
5) News
|
This
service is very similar to the traffic reports service, but
is devoted to the digest of the general national and international
news.
It should be noted that here the usage of
IBE™ technology
enable the common text-to-speech (TTS) systems to be introduced,
overcoming the drawback of reading the news by monotone voices,
which are typically deficient in expressiveness.
The technologies used to get the up-to-date
news from a Web page are DTMF,
TTS, IBE™
and IIWC™. The
texts of the page are converted automatically by a speech
synthesiser. [back]
|
6) Online
Shopping
|
This
service intends to demonstrate how the IVR
technology can be used for tasks varying from moderate to
high complexity, based on the strong potential of the IIWC™
technology.
The user chooses products thus forming a
"cart" or "basket" for subsequent home
delivery.
The technologies used to get the list of
products and prices from the Hypermarket Jumbo Web site are
DTMF, TTS
and IIWC™. The
texts of the page are converted automatically by a speech
synthesiser. [back]
|
7) Conventional
Weather Forecast
|
This
example shows how one can develop services of moderate complexity
with "conventional concatenation technology", as
long as it is supported by the ICOT™ optimisation and
by the IBE™ technology.
The database is not real (so by now it snows
in Alentejo): the demonstration mainly focuses on the voice
technology.
The technologies used are DTMF,
ICOT™, and IBE™.
[back]
|
| 8) Voice SMS |
There
is a service for the Short Message Service (SMS) delivery,
making possible to send an SMS message with the destination
number to IVR. Then the
IVR service makes a call to the provided destination, reading
the message contents to the addressee.
For example, the user sends a message:
"213100400#Bom dia mãe. Parabéns
pelo teu aniversário do teu filho Joaquim." ("213100400#Good
morning, mother. Congratulations on your Birthday from your
son Joaquim.")
The IVR
makes a call to 213100400 and reads the message "Bom
dia mãe. Parabéns pelo teu aniversário
do teu filho Joaquim. Repito... Bom dia mãe..."
using the TTS technology.
Additionally we provide an access to the
service by e-mail, just to demonstrate how e-mails are read
by TTS: the user sends
an e-mail message to the IVR
service, which then makes a call to the requested number and
reads the message. [back]
|
| Tech Glossary |
IVR
- Interactive Voice Response: Technology for establishing
voice dialogues by telephone. Actually, INOV In-IVR™
product enhances the concept, providing complete computer
telephony integration (CTI) platform, capable of interacting
with several information systems and being used as a Web portal.
DTMF - Dual-Tone
Multi-Frequency: a standard industrial technology that enables
the digits to be introduced from the phone keypad.
TTS - Text-To-Speech:
Automatic conversion from text to speech without the need
for specific recordings. This technology relies on the RealSpeak
algorithms from L&H, currently owned by ScanSoft. INOV
holds two licenses for usage of European (not Brazilian) Portuguese.
ASR - Automatic
Speech Recognition: Recognition of words. This technology
relies on Nuance's algorithms, for which INOV holds two licenses
for usage; integration with Philips is also available.
IIWC™ -
INOV Instant Web Contents: INOV proprietary technology that
allows the IVR to access any Web page available on the Internet,
and use its contents (or part) in one or more services. These
contents may be automatically updated with the desired periodicity
(from 10 seconds to several days), or on-demand when the service
needs the information.
IBE™ - INOV
Background Environment: INOV proprietary technology that allows
the IVR to recreate a given environment by using several special
effects, like background music, equalisation, reverberation,
delay, etc. to make better impression and/or to improve perception
in hostile environments (for example, cellular phones used
un automobiles).
ICOT™ - INOV Concatenated
Optimisation Technology: INOV proprietary technology that
allows the IVR to use pre-recorded sentences making the concatenations
as soft and undistinguishable as possible. It is essentially
an off-line technology.
API - Application
Programming Interface: Interface for the programming of additional
services and/or functionalities to the platform.
DLL - Dynamic Link
Library: External library of functionalities that can be used
by the platform.
[back]
|
| General Description |
The
In-IVR™ platform is a natural application of the accumulated
know-how within INOV-INESC in the last 10 years. It is a multi-service,
modular, scalable, robust and versatile platform, qualifying
as telecom-grade.
Thanks to its modularity, the platform can
be used with an almost absolute independence from the hardware,
making it very simple to install hundreds of thousand services
in one or more machines, with one or more telephone entries
per machine.
The platform contains multiple statistics
and alarm mechanisms, including auto-recovery systems, which
gives it irreprehensible up-times even for the most demanding
applications of the telecommunications industry.
The platform easily integrates with any other
INOV system, like SIMA™
or MultiSERV™,
and even with external systems through a vast range of access
services and protocols.
Currently the platform uses the most recent
speech recognition technologies from Nuance and Philips, either
small or large vocabularies, and also speech synthesis in
European Portuguese (among other languages) from ScanSoft,
constantly evolving to incorporate state-of-the-art technology
in these areas.
It can include a C/C++ API
(Windows), be integrated with external DLL's
or a proprietary interpreted language (script) for easy development
by the client. It includes a simulator so that one can test
the services in ordinary computers (Windows 95/98/NT/2000),
for example, using a laptop.
Besides these possibilities offered by the
platform, INOV offers services of consulting from the specification
to the implementation stage. [back]
|
|
|