KI-96: Advances in Artificial Intelligence 20th Annual German Conference on Artificial Intelligence Dresden, Germany, September 17 - 19, 1996, Proceedings

This book constitutes the refereed proceedings of the 20th Annual German Conference on Artificial Intelligence, KI-96, held in Dresden, Germany, in September 1996.
The book presents three invited papers together with 23 revised full research papers, nine posters and four systems demonstrations. All contributions were carefully selected from a larger number of excellent submissions, of which a considerable part had been submitted from abroad. The papers address many subareas of AI; however, there is a certain focus on new develpments in theorem proving, knowledge representation, and reasoning.

Rational models of normal logic programs.- Massively parallel reasoning about actions.- DisLoP: A disjunctive logic programming system based on PROTEIN theorem prover.- Abstraction as a heuristic to guide planning.- Equational theorem proving using taxonomic constraints.- Planning for distributed theorem proving: The teamwork approach.- A revision of Dependency-Directed Backtracking for JTMS.- A compiler-interpreter-system for decoding the user's intention within a speech understanding application.- Cyclic rules in Linear Connection Proofs.- Optimized nearest-neighbor classifiers using generated instances.- Towards a totally distributed meeting scheduling system.- Representations as basis of cognitive processes.- Improving the functionality of a text-to-speech system by adding morphological knowledge.- Solving hard combinatorial problems with GSAT — A case study.- Using rippling for equational reasoning.- Reasoning about action and change: Defeasible observations and actions with abnormal effects.- Falldaten: Case-based reasoning for the diagnosis of technical devices.- Integration of prosodic and grammatical information in the analysis of dialogs.- Application of a genetic algorithm for plausible justification of observations.- Managing multiple contexts efficiently.- Inductive learning in symbolic domains using structure-driven recurrent neural networks.- Belief revision in a nonclassical logic.- Avoiding combinatorial explosion in automatic test generation: Reasoning about measurements is the key.- Case retrieval nets: Basic ideas and extensions.- Putting default logics in perspective.- EULE2: A prototypical knowledge-based decision support system for the performance of office tasks.- When to prove theorems by analogy?.- Probabilistic knowledge representation andreasoning at maximum entropy by SPIRIT.- Knowledge based image understanding by iterative optimization.- Knowledge organization using the development system FAENSY.- A uniform proof procedure for classical and non-classical logics.- Qualitative reasoning under uncertainty with symbolic probabilities.- Narratives with concurrent actions: The mutual exclusion problem (extended abstract).- Tools for autonomous agents.- A concept language extended with different kinds of transitive roles.- Representation in auditory cortex.- A terminological qualification calculus for preferential reasoning under uncertainty.- Syntactic disambiguation by using categorial parsing in a DOOD framework.- Constructive disjunction revisited.