愉快吗
发表于 2025-3-28 16:36:26
https://doi.org/10.1007/978-3-642-61898-7s paper proposes that Levelt’s coordination rule does not necessarily entail that repairs are produced along the lines of this rule. In order to illustrate this, a connectionist language-production model is introduced, which produces both coordination-like repairs and syntactic repairs with the same mechanism.
Impugn
发表于 2025-3-28 22:30:06
http://reply.papertrans.cn/24/2373/237204/237204_42.png
群居动物
发表于 2025-3-29 00:19:43
http://reply.papertrans.cn/24/2373/237204/237204_43.png
Muscularis
发表于 2025-3-29 06:03:20
http://reply.papertrans.cn/24/2373/237204/237204_44.png
Observe
发表于 2025-3-29 09:40:01
Conference proceedings 1993s to dataand numerical processing, has provided new means of empirical dataacquisition and the application of mathematical models of adequatecomplexity.This volume contains the papers read at .QUALICO 91., and providesa representative overview of the state of the art in quantitativelinguistic research.
智力高
发表于 2025-3-29 11:30:39
http://reply.papertrans.cn/24/2373/237204/237204_46.png
PHONE
发表于 2025-3-29 16:08:20
http://reply.papertrans.cn/24/2373/237204/237204_47.png
碎片
发表于 2025-3-29 20:57:58
Statistical Language Modelling Using a Cache Memory the adaptation to those fluctuations, we added a dynamic component, the cache memory, which uses the word frequencies of the recent past to update the static word probabilities. Compared to a usual bigram language model we achieved an improvement of perplexity of 8% and 23%, respectively, depending on the heterogeneity of the data.
Shuttle
发表于 2025-3-30 00:07:59
http://reply.papertrans.cn/24/2373/237204/237204_49.png
半导体
发表于 2025-3-30 07:48:39
Vergleichende Anatomie der Wirbeltiere a word bigram model, the use of clustered equivalence classes in a bigram class model leads to a significant improvement, as measured by the perplexity. Depending on the size of the training material, the automatically clustered word classes are even better than manually determined categories.