On the dynamic adaptation of stochastic language models

By Reinhard Kneser, Volker Steinbiss. 1993

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Natural Language Processing

Citations in the corpus

Unbounded cache model for online language modeling with open vocabulary (2017, methodology): “Other adaptive language models have been proposed in the past: Kneser and Steinbiss [35] and, Iyer and Ostendorf [26] dynamically adapt the parameters of their model to recent history using different weight interpolation schemes.”
Improving Neural Language Models with a Continuous Cache (2016, background): “This makes this model particularly interesting for domain adaptation (Kneser & Steinbiss, 1993).”
Continual Adaptation for Efficient Machine Communication (2019, background): “With this said, the extent of adaptation in humancomputer dialogue is known to be affected by human participants’ expectations about the artificial agent (Branigan et al., 2011; Koulouri et al., 2016), potentially including expectations about whether it will be adaptive or not.”
Segmenting DNA sequence into 'words' based on statistical language model (2012, background): “More research works about the adaptation of n-gram models could refer to (22,23).”
Contextual Language Model Adaptation for Conversational Agents (2018, unknown): “Hence, they do not model longer-range dependencies that may vary between different subsets of the training data such as speaking style, vocabulary and topics of conversation [1].”
Two decades of statistical language modeling: where do we go from here? (2000, unknown): “[64] introduced yet another adapta tion scheme.”
Statistical language model adaptation: review and perspectives (2004, background): “The others are smaller models trained on the relevant portions of B (Kneser and Steinbiss, 1993).”
Generalizing and Hybridizing Count-based and Neural Language Models (2016, background): “Also conceptually similar is work on adaptation of n-gram LMs, which start with n-gram probabilities (Della Pietra et al., 1992; Kneser and Steinbiss, 1993; Rosenfeld, 1996; Iyer and Ostendorf, 1999) and adapt them based on the distribution of the current document, albeit in a linear model.”
Lightweight Adaptive Mixture of Neural and N-gram Language Models (2018, methodology): “In (Kneser and Steinbiss, 1993) information about ground truth performance of several n-gram models on previous 400 words is used to predict the optimal interpolation weights at each position.”