In this work we present a novel word lattice filtering algorithm which removes redundancy from lattices. We use the filtering algorithm to analyze lattices obtained from dynamic network and transducer-based LVCSR decoders from several sites regarding size, coverage, and redundancy. We show that our filtering algorithm reduces the size of lattices by 30 to 90% without degrading the oracle word error rate.
Bibliographic reference. Nolden, David / Soltau, Hagen / Povey, Daniel / Ghahremani, Pegah / Mangu, Lidia / Ney, Hermann (2014): "Removing redundancy from lattices", In INTERSPEECH-2014, 656-660.