Some results on stochastic language modeling
Renato De Mori, Roland Kühn · 1991
The paper will discuss three issues. The first is the derivation of precise probability scores for partial hypotheses containing islands, in the context of a Stochastic-Context-Free-Grammar (SCFG) for Language Modeling (LM). The second issue is the possibility of adding a cache component to a LM. This component alters the expected probability of words to reflect the speaker's patterns of word use. Finally, the idiosyncratic properties of dialogue are being studied; this work will indicate how knowledge about the discourse state can be incorporated into the LM and into the semantic component.