Intelligent Dialogue Systems

Master in Computer Science
Specialization: Methods and Models in Artificial Intelligence
SYLLABUS

Academic year 2011- 2012

   Departament Home



Cod: MMIA221
Titular curs: Prof.univ.dr.N.Tandareanu
Forma de invatamant: master
Year 2
2nd Sem., Course: 28h, Lab.: 28h
Credits: 8
Profil: computer science
Tip disciplina: compulsory
Objectives:
  • Initiation into research on the design and implementation of intelligent dialogue systems.
  • Learning specific fundamental notions of dialogue systems.
  • Learning the architecture of a system of dialogue, the role of each component.
  • Learning new methods of knowledge representation suitable for dialogue systems.
  • Implement the components of a dialogue systems.
Contents
I. Generalities on dialog systems.
  1. Types of dialogue systems.
  2. Demands a spoken dialogue system.
  3. Modular architecture of a voice dialogue system. Description of modules, the connection between these modules.
  4. Intelligent dialogue systems. Features and requirements.
  5. Architecture of intelligent dialogue system. Description of modules and their connection.
  6. Intelligent relevant dialogue systems: Darpa Communicator, ARISE, Mercury, UAH
II. Question-answering systems.
  1. The definition of a question-answering system as a particular dialogue system.
  2. Case study: intelligent query a relational data base.
    1. Interface communication, grammar for user-system communication.
    2. Communication via text or voice.
    3. Extracting semantics and its translation into operations on the database.
    4. Generation system's response
  3. Conditional schemas.
    1. Conditional binary relations.
    2. Definition of a conditional schema. Formal description of its components.
    3. Conditional graphs. Conditional graph generated bu a conditional schema.
    4. Knowledge processing by conditional schemas. Answer mapping.
  4. Question-answering systems based on conditional schemas.
    1. Architecture of a question-answering system based on conditional schema: modules, information flow.
    2. Implement a question-answering system based on conditional schema: modules, information flow.
III. Dialogue systems based on semantic schemas.
  1. Mathematical background
    1. ω - trees.
    2. Embedding ω -trees.
    3. Equivalence relation between ω -trees. Equivalence classes of ω -trees.
    4. Partial order relation on the factor set of equivalence classes.
    5. Local greatest elements in factor set.
    6. Structures ω -template .
  2. Master-slave systems of semantic schemas.
    1. Master schema and slave schemas in cooperating systems.
    2. Formal computations in a slave schema. Algebraic properties of the formal computations via ω -trees.
    3. Semantic computations in a slave schema.
    4. Formal computations in a master schema. Algebraic properties of the formal computations via ω -trees.
    5. Semantic computations in a master schema. Cooperation master-slave.
    6. Design of dialogue systems by means of master-slave systems.
Bibliography:
  1. Z.C.Carrion: - On the Development of Adaptive and Portable Spoken Dialogue Systems: Emotion Recognition, Language Adaptation and Field Evaluation, Univ. Granada (file pdf)
  2. Laura Monceaux et al.: - The query answering system PRODICOS (file pdf)
  3. Susan McRoy et al. : - Building Intelligent Dialog Systems, (file pdf)
  4. Susan McRoy et al.: - A Natural Language Generation Component for Dialog Systems, (file pdf)
  5. N.Tandareanu, M. Ghindeanu: - Towards a Mathematical Modelling of Conditional Knowledge, Research Notes in Artificial Intelligence and Digital Communications, Vol.103, 3nd Romanian Conference on Artificial Intelligence and Digital Communications, Craiova, June 2003, p.5-15 (file pdf)
  6. N.Tandareanu: - Cooperating Structures Based on Semantic Schemas in Knowledge Representation, Research Notes in Artificial Intelligence and Digital Communications, Vol.108, 8th International Conference on Artificial Intelligence and Digital Communications, p.1-21, 2008 (file pdf)
  7. N.Tandareanu, M.Colhon: - Conditional graphs generated by conditional schemas, Annals of the University of Craiova, Mathematics and Computer Science Series, Vol. 36(1), 2009,p.1-11, (file pdf)
  8. N.Tandareanu, C.Zamfir: - Algebraic properties of ω-trees (I), Annals of the University of Craiova, Mathematics and Computer Science Series, Vol. 37 (1) 2010 (file pdf)
  9. N.Tandareanu, M.Colhon: - Inference Mechanism in Conditional Schemas, Annals of the University of Craiova, Mathematics and Computer Science Series, Vol. 37 (1) 2010 (file pdf)
Lab Works
Ultima actualizare: Sept. 2010