Sokoban: Evaluating standard single-agent search techniques in the presence of deadlock -- A heuristic incremental modeling approach to course timetabling -- Planning strategy representation in DoLittle -- Establishing logical connectivity between query keywords and database contents -- Test-driving tanka: Evaluating a semi-automatic system of text analysis for knowledge acquisition -- An object indexing methodology as support to object recognition -- Oracles and assistants: Machine learning applied to network supervision -- A common multi-agent testbed for diverse seamless personal information networking applications -- A hybrid genetic algorithm for the vehicle routing problem with time windows -- Maintaining genetic diversity in genetic algorithms through co-evolution -- The effect of genetic operator probabilities and selection strategies on the performance of a genetic algorithm -- The impact of external dependency in genetic programming primitives -- Quality control in the concept learning process -- Grapheme generation in learning to read english words -- A trainable bracketer for noun modifiers -- Lessons learned in the development and implementation of a bilingual nationally accessible knowledge-based system -- Selecting the next action with constraints -- Poker as a testbed for AI research -- Fault prediction in the telephone access loop using a neural network -- Learning english syllabification rules -- Characterizing tractable CSPs -- An attribute redundancy measure for clustering -- On the complexity of VLSI-friendly neural networks for classification problems -- String clustering and statistical validation of clusters -- Finding partitions for learning control of dynamic systems -- A relational modeling of cognitive maps -- Distance constraint arrays: A model for reasoning on intervals with qualitative and quantitative distances -- Revising Timegraph-II -- Temporally invariant junction tree for inference in dynamic bayesian network -- Utility theory-based user models for intelligent interface agents -- A hybrid convergent method for learning probabilistic networks -- Relational concepts and the fourier transform: An empirical study -- ELEM2: A learning system for more accurate classifications -- Sequential instance-based learning -- Predicate invention from a few examples.

This book constitutes the refereed proceedings of the 12th Biennial Conference of the Canadian Society for Computational Studies of Intelligence, AI'98, held in Vancouver, BC, Canada in June 1998. The 28 revised full papers presented together with 10 extended abstracts were carefully reviewed and selected from a total of more than twice as many submissions. The book is divided in topical sections on planning, constraints, search and databases; applications; genetic algorithms; learning and natural language; reasoning; uncertainty; and learning.