Efficient Inference and Computation of Optimal Alternatives for Preference Languages Based On Lexicographic Models
This work addresses computational efficiency in preference reasoning for AI systems, though it appears incremental as it builds on existing lexicographic models with specific algorithmic improvements.
The paper tackles the problem of preference inference and computation of optimal alternatives for lexicographic preference models, showing that testing consistency and computing optimal sets is polynomial for a specific preference language LpqT that supports various statement types.
We analyse preference inference, through consistency, for general preference languages based on lexicographic models. We identify a property, which we call strong compositionality, that applies for many natural kinds of preference statement, and that allows a greedy algorithm for determining consistency of a set of preference statements. We also consider different natural definitions of optimality, and their relations to each other, for general preference languages based on lexicographic models. Based on our framework, we show that testing consistency, and thus inference, is polynomial for a specific preference language LpqT, which allows strict and non-strict statements, comparisons between outcomes and between partial tuples, both ceteris paribus and strong statements, and their combination. Computing different kinds of optimal sets is also shown to be polynomial; this is backed up by our experimental results.