Word learning as Bayesian inference.

By Fei Xu, Josh Tenenbaum. 2007

1,001 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Cultural Studies, Developmental and Educational Psychology

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

In Cultural Studies

In Developmental and Educational Psychology

Citations in the corpus

Learning with Hierarchical-Deep Models (2013, background): “…learns to learn novel concepts from very few training example by learning low-level generic features, high-level features that capture correlations among low-level features, and a category hierarchy for sharing priors over the high-level features that are typical of different kinds of concepts.”
Machine Teaching for Bayesian Learners in the Exponential Family (2013, background): “We focus on Bayesian learners because they are widely used in both machine learning and cognitive science [7, 23, 24] and because of their predictability: they react to any teaching examples in D by performing Bayesian updates.”
Learning Semantic Image Representations at a Large Scale (2014, methodology): “A key aspect of this problem is determining the degree to which the concept should be generalized [116] when multiple concepts are logically consistent with the given examples: for example, consider the concepts represented by examples in Figure 4.”
Troubles with Bayesianism: An introduction to the psychological immune system (2018, background): “…not hard to see: it allows for the possibility of a single mental mechanism—Bayesian updating—to unify mental processes as diverse as word learning (Xu & Tenenbaum, 2007), belief updating (Bennett, 2015), conditional reasoning (Oaksford & Chater, 1994), the development of moral judgments…”
Rational integration of noisy evidence and prior semantic expectations in sentence interpretation (2013, background): “This prediction holds generally for a wide range of statistically sensible noise models, and follows from the Bayesian size principle (24, 25): a deletion only requires a particular word to be randomly selected from a sentence, whereas an insertion requires its selection from (a subset of) the produ”
The Emergence of Organizing Structure in Conceptual Representation (2016, background): “These representations can also provide scaffolding for learning higher-level concepts with direct structural interpretations, such as using a tree structure to help learn the world “primate” (Xu & Tenenbaum, 2007).”
The rule of three: How the third event signals the emergence of a streak. (2007, result): “After three examples of a new word, people can generalize its definition to new objects (Tenenbaum & Xu, 2000), a result consistent with work by Kareev (1995, 2000) on humans’ tendency to pick up patterns in small samples.”
Infants rapidly learn word-referent mappings via cross-situational statistics. (2008, methodology): “One is the formulation and evaluation of hypotheses (e.g., Tenenbaum & Xu, 2000).”
Children's imitation of causal action sequences is influenced by statistical and pedagogical evidence. (2011, background): “…are surprisingly sophisticated in their use of statistical information such as conditional probabilities in a range of domains, from phonology (Saffran, Aslin, & Newport, 1996), to visual perception (Fiser & Aslin, 2002; Kirkham, Slemmer, & Johnson, 2002), to word meaning (Xu & Tenenbaum, 2007).”
Discovering linguistic structures in speech: models and applications (2014, background): “For instance, in [186], children were tested to decide which objects belong to the set of elephants and which do not after hearing the word elephant paired with an exemplar.”
Memory, Show the Way: Memory Based Few Shot Word Representation Learning (2018, background): “Such ability to efﬁciently acquire representation from small data, namely fast mapping, is thought to be the hallmark of human intelligence that a cognitive plausible agent should strive to reach (Xu and Tenenbaum, 2007; Lake et al., 2015).”
A Probabilistic Computational Model of Cross-Situational Word Learning (2010, background): “A richer representation that reﬂects the diﬀerences between (syntactic or semantic) categories would make it possible to study such learning preferences in our model (see Xu and Tenenbaum, 2007, for one such study).”
Word learning emerges from the interaction of online referent selection and slow associative learning (2012, methodology): “Bayesian models (Xu & Tenenbaum, 2007), in their focus on how interacting constraints lead to accurate inferences, off-load many constraints onto priors (and add new ones).”
Reconstructing constructivism: Causal models, Bayesian learning mechanisms and the theory theory (2012, positive): “Even infants make these inferences (Gweon, Tenenbaum, & Schulz, 2010; Xu & Tenenbaum, 2007).”
Learning from others: children's construction of concepts. (2009, background): “…can track low-level statistical cues with much greater accuracy than had been previously realized (e.g., Gopnik et al. 2004, Saffran et al. 1996, Xu & Tenenbaum 2007), and in part due to new empiricist models that provide a more detailed and realistic appreciation of children’s concepts…”