1. CHART or PICTURE INTEGRATING SEMANTIC WEB TO IMPROVE ONLINE Marta Gatius Meritxell González TALP Research Center (UPC) They are friendly and easy to use. They support different types of interactions: menus, form filling and commands. They can tackle complex questions. They can be integrated with other modes. They are adaptable to different types of users: languages, skills, ages and cultural sensitivity. Voice mode improves web usability, especially for people with impairments The system consists of three layers: 1 Presentation. It controls interaction with the user. VoiceXML a. The Audio Web Server connected with a VoiceXML Platform Platform acting as Voice Browser. b. The Text Server and the Natural Language Analyzer for text interaction. 2 Dialogue Management. It controls the dialogue flow. a. The Application Manager, controlling the language and the service identification. b. The Dialogue Manager, controlling dialogue. 3 Knowledge Layer. It manages data for the application. a. The Ontology Manager, for ontology access. b. Action & Query Manager, for backend access. CHART or PICTURE The system uses ontologies modelling the service tasks. The service tasks are represented by a set of concepts described by attributes. The attributes represent the information required/resulting when performing these tasks (input and output parameters). Domain and task-restricted ontologies are used for improving The communication: the dialogue flow and language processing The obtaining of the linguistic resources for new services standard SRGS - Grammars following the standard SRGS for voice - Context-free g - Context-free grammars and lexicon for text standard language OWL Ontologies are implemented in the standard language OWL DIALOGUE SYSTEMS FOR ACCESSING THE WEB ARCHITECTURE OF THE SYSTEM THE USE OF ONTOLOGIES THE DIALOGUE SYSTEM The dialogue system was developed in the context of the European project HOPS (http://www.hops-fp6.org) It was designed to enhance the usability and accessibility of online public services and information. It integrates language and semantic web technologies. The current implementation of the system supports speech (through the telephone) text mode (through the web) in several languages: English, Spanish, Catalan and Italian. In FADA the service registry is a network of nodes keeping information about the services. It provides to the system a more fault tolerant nature. Registration and lease renewal life-cycle Service proxy look-up mechanism Following a Service Oriented Architecture The components of the system are interrelated between themselves using FADA (Federated Advanced Directory Architecture) Application Manager Audio Web Server Text Server Natural Language Analyzer Dialogue Manager Ontology Manager Action & Query Manager Common Backend Access Torino Backend Barcelona Backend Camden Backend vxml grxml Applications Dialogues & Templates Ontologies Data Bases grammars txml Fada Communication

2. AND LANGUAGE TECHNOLOGIES PUBLIC SERVICES Sheyla Militello Pablo Hernández Loquendo S.p.A SUN Microsystems The Cultural Agenda is an informational service giving information about the cultural events in a city. Two possible consulting tasks are allowed: - Consult the list of events matching a specific information - Consult specific information about a particular event These two tasks are represented in separated concepts. The attributes describing these concepts are those parameters of the query back-end involved in the communication. 1. 1. Dupriez, C. Multilingual Vocal Access to Databases: the Central Role of Ontology Management in VocaBase. Sixth International Protégé Workshop, England, 2003. 2. 2. Milward, D., Beveridge, M. Ontology-Based Dialogue Systems. Third Workshop on Knowledge and Reasoning in Practical Dialogue Systems, 2003. 3. 3. Wahlster, W. SmartKom: Symmetric Multimodality in an Adaptive and Reusable Dialogue Shell. Human Computer Interaction Status Conference, 2003. 4. 4. Wilcock, G., Jokinen, K. Generating Responses and Explanations from RDF/XML and DAML+OIL. Third Workshop on Knowledge and Reasoning in Practical Dialogue Systems, 2003. Ontologies are specially appropriate in dialogue systems: 1. In complex and changing domains 2. In mutli-modal and/or multilingual systems 3. In multi-domain systems In the Hops system ontologies modelling service tasks and domain knowledge are used for Dialogue management modelling Language processing Language generation Ontologies are used to generate the systems messages, the grammars and the lexicon adapted to Each service The speech and text modes The different languages SAMPLE SERVICE 1: THE LARGE OBJECTS COLLECTION SAMPLE SERVICE 2 THE CULTURAL AGENDA CONCLUSIONS REFERENCES The three tasks performed by the transactional service Large Objects Collection are represented in three separated concepts: Information, Collection and Cancellation. The collection task sets a date for collection. The attributes describing this concept correspond to the service back-end parameters. The attribute objectype represents the type of the object to be collected. It is linked to a specific furniture taxonomy providing all its possible values. It is used to obtain the service-restricted linguistic resources. CULTURAL AGENDA EVENT LIST EVENT LIST event type locationdateparticipantprice SINGLE EVENT SINGLE EVENTlocationtitleparticipantdatedescription Integrating the ontologies used in a general e-government infrastructure. Incorporating service discovering and composition techniques. FUTURE WORK