Knowledge-based media coordination in intelligent user interfaces -- Some notes on cumulative reasoning -- A family of three-valued autoepistemic logics -- Viewpoints subsume beliefs, truth and situations -- Automating meta-theory creation and system extension -- Implementing exceptions in inheritance by concept transforming actions -- Approximate entailment -- Negation as failure and constraints through abduction and defaults -- A hybrid system with datalog and concept languages -- Optimal search for conjunctive goals using constraints -- FSS-WASTL interactive knowledge acquisition for a semantic lexicon -- Hybrid encoding: Constraints on addressing structure -- Uncertainty and cyclic dependencies a proposal and a network implementation -- Abductive reasoning in a multi-theory framework -- The abstract interpretation of hybrid rule/frame-based systems -- Taxonomic reasoning in configuration tasks -- Representation and use of teleological knowledge in the multi-modeling approach -- Towards the integration of different knowledge sources in model-based diagnosis -- Integrating statistics, numerical analysis and dependency-recording in model-based diagnosis -- Version-space induction with multiple concept languages -- Knowledge compilation to speed up numerical optimization -- New perspectives about default hierarchies formation in learning classifier systems -- Use of a causal model to learn diagnostic knowledge in a real domain -- Extending inverse resolution to build up abstractions -- Using word association for syntactic disambiguation -- A computational model of tense selection and its experimentation within an intelligent tutor -- An efficient context-free parsing algorithm with semantic actions -- From scopings to interpretation: The semantic interpretation within the AlFresco system -- Parsing dependency grammars -- A system based on neural architectures for the reconstruction of 3-D shapes from images -- Retrieval of high-level data structures from stereo vision data -- High-level and low-level computer Vision: Towards an integrated approach -- An associative link from geometric to symbolic representations in artificial vision -- A theory of sensor-based robot navigation using local information -- Insights into cooperative group design: Experience with the LAN designer system -- Direct interaction among active data structures: A tool for building AI systems -- Introducing knowledge representation techniques in database models -- Self-organizing maps: A new digital architecture -- Electrical equipment maintenance training: An its application in industrial environment -- Temporal and spatial data and default assumptions in DIPSY-E system -- Visual indexing with an attentive system -- On neural network programming -- A logic for the representation of spatial knowledge -- Truth maintenance in approximate reasoning -- A strategy for design and development of complex knowledge-systems -- Efficient compilation of first order predicates -- A flexible tool for assumption-based user modeling -- KL: A neural model for capturing structure in speech -- Lexical discrimination within a multilevel semantics approach -- Metalevel representation of analogical inference -- Semantic interpretation of copulative sentences -- Typicality for plausible reasoning -- A distributed image-understanding system.

This book collects the scientific papers presented at the 2nd Congress of the Italian Association for Artificial Intelligence, held in Palermo in October 1991. It displays the state of the art of both Italian and European scientific research in AI. The book begins with an invited paper by W. Wahlster et al. The bulk of the book is then divided into five parts on: - Knowledge representation (18 papers), - Knowledge acquisition (5 papers), - Natural language (5 papers), - Perception and robotics (5 papers), - Architecture and technologies (5 papers). A section containing short papers completes the book. The high quality of the papers reflects massive research activity mainly devoted to the theoretical aspects of AI, but clearly aimed at consolidating the results already achieved. Several contributions are oriented to the technological aspects of AI.