Markov Chain
From lingwiki
A Markov Chain is a collection of random variables, X0, X1, X2, ... , which hold the property that given the present state being analyzed, future states are conditionally independent from past states. Formally put,
More simply, the value at time i depends only on the value at the preceding time, and on nothing that went on before.
[edit] Use of Markov Chains in Speech Recognition
Probabilistic functions of Markov Chains are called Hidden Markov Models, which are widely used in Acoustic Modeling, Language Modeling, and linguistic decoding processes of speech recognition.