Fingerprint indexing for paramodulation and rewriting

Authors:
Stephan Schulz
Affiliations:
Institut für Informatik, Technische Universität München, München, Germany
Venue:
IJCAR'12 Proceedings of the 6th international joint conference on Automated Reasoning
Year:
2012

Citing 5
Cited 1

Experiments with discrimination-tree indexing and path indexing for term retrieval

Journal of Automated Reasoning
Term Indexing

Term Indexing
The design and implementation of VAMPIRE

AI Communications - CASC
E - a brainiac theorem prover

AI Communications - CASC
The TPTP Problem Library and Associated Infrastructure

Journal of Automated Reasoning

Simple and efficient clause subsumption with feature vector indexing

Automated Reasoning and Mathematics

Quantified Score

Hi-index	0.00

Visualization

Abstract

Indexing is critical for the performance of first-order theorem provers. We introduce fingerprint indexing, a non-perfect indexing technique that is based on short, constant length vectors of samples of term positions ("fingerprints") organized in a trie. Fingerprint indexing supports matching and unification as retrieval relations. The algorithms are simple, the indices are small, and performance is very good in practice. We demonstrate the performance of the index both in relative and absolute terms using large-scale profiling.