acl, eccai and the verbmobil/smartkom consortia german research center for artificial intelligence...
TRANSCRIPT
ACL, ECCAI and the Verbmobil/SmartKom Consortia
German Research Center for Artificial IntelligenceStuhlsatzenhausweg 3, Geb. 43.1
66123 SaarbrückenTel.: (0681) 302-5252/4162
Fax: (0681) 302-5341E-mail: [email protected]
WWW: http://www.dfki.de/~wahlster
Wolfgang Wahlster
ACL Workshop, Hong Kong 7 October 2000
Hosted by DFKI (Thierry Declerck)
http://registry.dfki.de
130 links to software systems (shareware, freeware, open source, trial versions)
all registered software packages are classified
software from academic and industrial groups
Classification according to NSF/EU „State of the Art in Language Technology“
Cooperation with ELRA, LDC, ELSNET, CLASS, TELRI, BAS
The International NL Software Registry of ACL
UkrainADUIS
ItalyAIIA
FranceAFIA
PortugalAPPIA
BulgariaBAIA
U.K.BCS-SGES
SSAISBCzech
RepublikCSKIDenmark
DAISFinlandFAIS
GreeceEETN
GermanyGI/KI
IsraelIAAI
LatviaLANO
LithuaniaLIKS-AIS
NorwayNAIS
HungaryNJSZT
Belgium/ The Netherlands
BNVKI
AustriaÖGAI
RussiaRAAI
SwedenSAIS
SwitzerlandSGAICOSlovenia
SLAIS
Slovak RepublicSSKI SAV Spain
AEPIA
CataloniaACIA
ECCAI MEMBERS
Founded in 1982, currently 26 member societies
Represents more than 6000 AI researchers
The world‘s largest AI society
Web site: WWW.ECCAI.ORG
Every two years: ECAI conference and ACAISummer school
ETAI electronic journal, neo-classical reviewing Hardcopy distributed by Royal Swedish Academy of Sciences
ECCAI: European Coordination Committeeon Artificial Intelligence
UNIVERSITÄT DESSAARLANDES
RUHR-UNIVERSITÄTBOCHUM
Phase 2
UNIVERSITÄTHAMBURG
UNIVERSITÄTKARLSRUHE
UNIVERSITÄTBIELEFELD
TECHNISCHEUNIVERSITÄT
MÜNCHEN
FRIEDRICH-ALEXANDER-UNIVERSITÄT
ERLANGEN-NÜRNBERG UNIVERSITÄTSTUTTGART
RHEINISCHE FRIEDRICHWILHELMS-UNIVERSITÄT
BONNLUDWIG
MAXIMILIANSUNIVERSITÄT
MÜNCHEN
TU-BRAUNSCHWEIG
EBERHARDT-KARLSUNIVERSITÄT
TÜBINGEN W. Wahlster, DFKI
The Verbmobil Consortium
Multilingualand Mobile
CommunicationAssistants
Multimodal Interfaces
SmartKom
Speech-based Web Access to Multilingual
Web pages
WAP Phones WebTV
Multilingual Audio Retrieval
and Audio Mining
Discussions Lecture Notes Organizers
MultilingualIndexing andAnnotation of
Videos
Video Archives News Archives
Call CentersECommerce Mobile Travel Assistance Telephone Translations
Verbmobil
Dialog Translation
International Research Trends in Multilingual Systems
Multilingual Language Technology Speech Recognition, Language Understanding, Language Generation,
and Speech Synthesis
Multilingual Language Technology Speech Recognition, Language Understanding, Language Generation,
and Speech Synthesis
Spontaneous Speech, Robust Processing and Translation, Semantic and Pragmatic Understanding
Verbmobil‘s Massive Data Collection Effort
Transliteration Variant 1Transliteration Variant 2 Lexical OrthographyCanonical PronounciationManual Phonological Segmentation
Automatic Phonological SegmentationWord SegmentationProsodic SegmentationDialog ActsNoises
Superimposed SpeechSyntactic CategoryWord CategorySyntactic FunctionProsodic Boundaries
The so-called Partitur (German word for musical score)orchestrates fifteen strata of annotations
3,200 dialogs (182 hours)with 1,658 speakers79,562 turnsdistributed on56 CDs, 21.5 GB
Treebank: 85,000 Trees
79,562 turns
1,520,000 words
76,210 dialog acts
Collection Sites: All over Germany
American English: CMU
Japanese: ATR
Mandarine: Philips Taiwan
The Verbmobil Corpora
Machine Learningfor the Integration of Statistical Properties into
Symbolic Models for Speech Recognition, Parsing,Dialog Processing, Translation
TranscribedSpeech Data
SegmentedSpeech
with ProsodicLabels
AnnotatedDialogs withDialog Acts
Treebanks &Predicate-ArgumentStructures
AlignedBilingualCorpora
HiddenMarkovModels
Neural Nets,MultilayeredPerceptrons
ProbabilisticAutomata
ProbabilisticGrammars
ProbabilisticTransfer
Rules
Extracting Statistical Properties from Large Corpora
MediaInterfaceEuropean Media LabUinv. Of
MunichUniv. ofStuttgart
Saarbrücken
Aachen
Dresden Berkeley
Stuttgart
MunichUniv. ofErlangen
Heidelberg
Main ContractorProject Management
TestbedSoftware Integration
DFKISaarbrücken
The SmartKom Consortium:
Project Budget: $ 34 MProject Duration: 4 years
DAIMLERCHRYSLERUlm
SmartKom: Intuitive Multimodal Interaction
W. Wahlster, DFKI
User(s)
MediaAnalysis
Design
Media Fusion
OutputRendering
Representation and Inference
UserModel
DiscourseModel
DomainModel
TaskModel
MediaModels
InteractionManagement
MediaAnalysis
InputProcessing
Information
Applications
People Intention
RecognitionMediaDesign
Ap
pli
cati
on
In
terf
aceDiscourseModeling
UserModelin
g
PresentationDesign
Language
Graphics
Gesture
Biometrics
Language
Graphics
Gesture
AnimatedPresentation
Agent
The Architecture of the SmartKom Agent (cf. Maybury/Wahlster 1998)
W. Wahlster, DFKI
SmartKom-Home/Office:A Versatile Agent-based Interface
SmartKom-Public: A Multimodal
CommunicationBooth
SmartKom-Mobile: A Handheld
CommunicationAssistant
Media Analysis
Kernel ofSmartKomInterface
Agent
Interaction Management
ApplicationManage-
ment
MediaDesign
SmartKom: A Transportable and Transmutable Interface Agent
W. Wahlster, DFKI
Smartcard/ Credit Cardfor authentication and billing
Docking stationfor PDA/Notebook/Camcorderhigh speed and broadbandwidth Internet connectivity
High-resolution scanner
Loudspeaker
Room microphone
Face-tracking camera
Virtual touchscreenprotected against vandalism
Multipoint video conferencing
SmartKom-Public:A MultimodalCommunication Booth
W. Wahlster, DFKI
SmartKom Domains: EPG and Cinema Selection
Multimodal SelectionOf Cinema
Multimodal Selection of Seats
Camera
GPS
Microphone
Loudspeaker
Stylus-Activated Sketch Pad
WearableComputeServer
Docking Stationfor Car PC
Biosensorfor Authentication
& Emotional Feedback
GSM for Telephone,Fax, Internet Connectivity
SmartKom-Mobile: A Handheld Communication Assistant
W. Wahlster, DFKI
SmartKom‘s Data Collection of Multimodal Dialogs
User
Side-viewCamera
Face-tackingCamera withMicrophone
EnvironmentalNoise
MicrophoneArray
Screen
ProjectedWebpage
Face-tackingCamera
LoudspeakerMicrophone
Array
User
Bird’s-eyeCamera LCD
Beamer
SIVIT-Camera
BAS in Munich (Prof. Tillmann, Dr. Schiel)
68 GB of transliterated speech data 590 hours of speech, 2500 different speakers
works for Verbmobil and SmartKom consortium
treebanks, gesture and dialog act annotations
very large corpora of GSM phone dialogs multi-channel recordings of spontaneous dialogs WOZ data
one year exclusive use for German labs and industry
after one year distribution by ELRA and then by LDC
The German Center for Language Resources