Trainable grammars for speech recognition

By James K. Baker. 1979

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

Structured Attention Networks (2017, methodology): “It is possible to calculate the marginal probability of each edge p(zij = 1 |x, q) for all i, j inO(n3) time using the inside-outside algorithm (Baker, 1979) on the data structures of Eisner (1996).”
Recurrent Neural Network Grammars (2016, background): “Syntactic language models jointly generate a syntactic structure and a sequence of words (Baker, 1979; Jelinek and Lafferty, 1991).”
Learning to Map Sentences to Logical Form: Structured Classification with Probabilistic Categorial Grammars (2005, methodology): “Expectations of this type can again be calculated using dynamic programming, us-ing a variant of the inside-outside algorithm (Baker, 1979), which was originally formulated for probabilistic context-free grammars.”
Accurate Unlexicalized Parsing (2003, background): “In the early 1990s, as probabilistic methods swept NLP, parsing work revived the investigation of probabilistic context-free grammars (PCFGs) (Booth and Thomson, 1973; Baker, 1979). However, early results on the utility of PCFGs for parse disambiguation and language modeling were somewhat disappoint”
Grammar Variational Autoencoder (2017, background): “By assigning probabilities to each production rule in the grammar, it is possible to deﬁne a probability distribution over parse trees (Baker, 1979; Booth & Thompson, 1973).”
Parameter Learning of Logic Programs for Symbolic-Statistical Modeling (2001, methodology): “We now compare the graphical EM algorithm with the Inside-Outside algorithm (Baker, 1979; Lari & Young, 1990).”
Two decades of statistical language modeling: where do we go from here? (2000, methodology): “These transition probabilities can be estimated from annotated corpora using the Inside-Outside algorithm [40], an Estimation-Maximization (EM) algorithm (see [41]).”
Learning syntactic structure (2007, methodology): “This algorithm, called the inside-outside algorithm (Baker 1979; Lari and Young 1990), begins with some initial setting of the rule probabilities and re-estimates these probabilities on a corpus until a local maximum of the corpus likelihood is reached.”
Adaptive computation and machine learning (1998, unknown): “Learning algorithms for SCFGs of one sort or the other are described in [25, 345, 459, 460, 156].”
Intricacies of Collins’ Parsing Model (2004, background): “crude proxy for the outside probability of a chart item (see (Baker, 1979; Lari and Young, 1990) for full descriptions of the Inside–Outside algorithm).”
Variational Bayesian Grammar Induction for Natural Language (2006, methodology): “Since Lari and Young [11] empirically showed the possibility of statistical induction of PCFGs using the Inside-Outside algorithm [2], parameter-estimationbased grammar induction has received a great deal of attention.”
Corpus-Based Induction of Syntactic Structure: Models of Dependency and Constituency (2004, methodology): “…to argue empirically against the poverty of the stimulus (Clark, 2001), to use induction systems as a ﬁrst stage in constructing large treebanks (van Zaanen, 2000), to build better language models (Baker, 1979; Chen, 1995), and to examine cognitive issues in language learning (Solan et al., 2003).”
From Natural Language Specifications to Program Input Parsers (2013, methodology): “The algorithm works similarly to the inside algorithm (Baker, 1979), except that we do not assume the tree is binary.”
First- and Second-Order Expectation Semirings with Applications to Minimum-Risk Training on Translation Forests (2009, methodology): “Then the algorithm of Figure 2 reduces to the classical inside algorithm (Baker, 1979) and computes Z.”
Natural Language Grammar Induction Using a Constituent-Context Model (2001, methodology): “These systems t ypically then attempt to find the grammar production parameters 2 which maximize the likelihoodP(S|2) using the inside-outside algorithm [1], which is an efficient (dynami c programming) instance of the EM algorithm [8] for PCFGs.”
Survey of the State of the Art in Human Language Technology (2010, methodology): “In that case, iterative procedures may be used to find the transition probabilities that maximize the probability that the training corpus was observed (Dempster, Laird, et al., 1977; Baum & Petrie, 1966; Baker, 1979).”
Sequence-to-Sequence Learning with Latent Neural Grammars (2021, unknown): “However, we observe that the inner summation t ∈T ( y ) p θ ( t | s ) = p θ ( y | s ) can be computed with the usual inside algorithm [9] in , where S is the source length and T is the target length.”
Deep Learning in Medical Imaging: General Overview (2017, background): “Many researchers (37-40) have developed technologies that enable the recognition and translation of the spoken language into text by computerized devices, including smart technologies and robotics.”