A survey of smoothing techniques for ME models

By Stanley F. Chen, R. Rosenfeld. 2000

246 citations recorded by the source corpus.

How researchers see this paper

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

In Artificial Intelligence

Citations in the corpus

Regularizing Text Categorization with Clusters of Words (2016, background): “Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 1827–1837, Austin, Texas, November 1-5, 2016. c©2016 Association for Computational Linguistics”
Random forests and the data sparseness problem in language modeling (2007, methodology): “However, the similarity also indicates that it is unlikely for the maximum entropy models to improve upon interpolated Kneser–Ney models if only n-gram features are used, as shown by Chen and Rosenfeld (2000) and Goodman (2004).”
A maximum entropy approach to species distribution modeling (2004, background): “Other priors studied for maxent are Gaussian ( Chen & Rosenfeld, 2000 ) and exponential (Goodman, 2003).”
Offline recognition of unconstrained handwritten texts using HMMs and statistical language models (2004, background): “Several discounting strategies are available in the literature and their performances are essentially equivalent [24].”
Accurate Information Extraction from Research Papers using Conditional Random Fields (2004, unknown): “Here we explore two key practical issues: (1) regularization, with an empirical study of Gaussian (Chen and Rosenfeld, 2000), exponential (Goodman, 2003), and hyperbolic-L1 (Pinto et al.”
A Bayesian Mixed Effects Model of Literary Character (2014, methodology): “This work has since been extended to a Bayesian setting by applying both a Gaussian prior (Chen and Rosenfeld, 2000), which dampens the impact of any individual feature, and sparsity-inducing priors (Kazama and Tsujii, 2003; Goodman, 2004), which can drive many feature weights to 0.”
Inequality Maximum Entropy Classifier with Character Features for Polyphone Disambiguation in Mandarin TTS Systems (2007, methodology): “Gaussian Smoothing Several smoothing algorithms have been proposed to overcome data sparseness and Chen and Rosenfeld [11] demonstrate that Gaussian smoothing performs as well as or better than all others.”
Deep belief nets for natural language call-routing (2011, methodology): “The MaxEnt models are trained using the improved iterative scaling algorithm [15] with Gaussian prior smoothing [14] using a single universal variance parameter of 2.”
Question classification with log-linear models (2006, background): “However, maximum likelihood parameters for log-linear models have a tendency to overfit the data [ 2 ].”
Linguistic Structured Sparsity in Text Categorization (2014, background): “Since the seminal work of Chen and Rosenfeld (2000), the importance of regularization in discriminative models of text— including language modeling, structured prediction, and classification—has been widely recognized.”
Controlled user evaluations of information visualization interfaces for text retrieval: Literature review and meta-analysis (2008, unknown): “Second, the data-sparseness problem encountered in DA recognition is also relevant to other NLP and IR tasks ( Chen & Rosenfeld, 2000; Kazama & Tsujii, 2005).”
Active Learning Literature Survey (2009, background): “Parameter regularization (Chen and Rosenfeld, 2000; Goodman, 2004) can help control this effect somewhat, and it doesn’t appear to be a significant problem in practice.”
Joint-sequence models for grapheme-to-phoneme conversion (2008, methodology): “As both models make use the same contextual information ( M -grams), smooth probabilities of unseen events and preserve lower-order marginals, similar levels of performance can be expected (Chen and Rosenfeld, 2000).”
Efficient Large-Scale Distributed Training of Conditional Maximum Entropy Models (2009, background): “Conditional maximum entropy models [1, 3], conditional maxent models for short, also known as multinomial logistic regression models, are widely used in applications, most prominently for multiclass classification problems with a large number of classes in natural language processing [1, 3] and…”