SRILM Manual Pages

Papers and Tutorials

• A. Stolcke, SRILM - An Extensible Language Modeling Toolkit, in Proc. Intl. Conf. Spoken Language Processing, Denver, Colorado, September 2002.
  Gives an overview of SRILM design and functionality.
• D. Jurafsky, Language Modeling, Lecture 11 of his course on "Speech Recognition and Synthesis" at Stanford.
  Excellent introduction to the basic concepts in LM.
• J. Goodman, The State of The Art in Language Modeling, presented at the 6th Conference of the Association for Machine Translation in the Americas (AMTA), Tiburon, CA, October, 2002.
  Tutorial presentation and overview of current LM techniques (with emphasis on machine translation).
• K. Kirchhoff, J. Bilmes, and K. Duh, Factored Language Models Tutorial, Tech. Report UWEETR-2007-0003, Dept. of EE, U. Washington, June 2007.
  This report serves as both a tutorial and reference manual on FLMs.
• S. F. Chen and J. Goodman, An Empirical Study of Smoothing Techniques for Language Modeling, Tech. Report TR-10-98, Computer Science Group, Harvard U., Cambridge, MA, August 1998 (original postscript document).
  Excellent overview and comparative study of smoothing methods. Served as a reference for many of the methods implemented in SRILM.

FAQ

Programs

These are the top-level executables that are currently part of SRILM:

ngram-count(1)

count N-grams and estimate language models

ngram-merge(1)

merge N-gram counts

ngram(1)

apply N-gram language models

ngram-class(1)

induce word classes from N-gram statistics

disambig(1)

disambiguate text tokens using an N-gram model

hidden-ngram(1)

tag hidden events between words

nbest-lattice(1)

rescore N-best lists and lattices

nbest-optimize(1)

optimize score combination for N-best word error minimization

nbest-mix(1)

interpolate N-best posterior probabilities

segment(1)

segment text using N-gram language model

segment-nbest(1)

rescore and segment N-best lists using N-gram language models

anti-ngram(1)

count posterior-weighted N-grams in N-best lists

multi-ngram(1)

build multiword N-gram models

lattice-tool(1)

manipulate word lattices

nbest-pron-score(1)

score pronunciations and pauses in N-best hypotheses

Utility Scripts

Additional tools implemented as scripts:

training-scripts(1)

miscellaneous conveniences for language model training

lm-scripts(1)

manipulate N-gram language models

ppl-scripts(1)

manipulate perplexities

pfsg-scripts(1)

create and manipulate finite-state networks

nbest-scripts(1)

rescore and evaluate N-best lists

select-vocab(1)

select a maximum-likelihood vocabulary from a mixture of corpora

File Formats

Some of the data formats used by SRILM:

ngram-format(5)

ARPA backoff N-gram models

classes-format(5)

Word class definitions

pfsg-format(5)

Decipher(TM) probabilistic finite-state grammars

nbest-format(5)

N-best hypotheses lists

LM Library Classes

These are some of the basic classes of the SRILM library. Note that this list is woefully incomplete, as this part of the documentation is largely yet to be written.

LM(3)

Generic language model

Vocab(3)

Vocabulary indexing for SRILM

Prob(3)

Probabilities for SRILM

File(3)

Wrapper for stdio streams

Back to the SRILM home page.