AnalogySpace: Reducing the Dimensionality of Common Sense Knowledge

By Robyn Speer, Catherine Havasi, Henry Lieberman. 2008

202 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.

Citations in the corpus

A semantic matching energy function for learning with multi-relational data (2013, methodology): “Other directions have also been proposed derived from probabilistic matrix factorization for multi-dimensional data (Chu and Ghahramani 2009) or by adapting dimensionality reduction techniques such as SVD (Speer et al. 2008; Cambria et al. 2009).”
Representing General Relational Knowledge in ConceptNet 5 (2012, neutral): “That information is important in some applications such as generating natural-language questions to ask, as the AnalogySpace system (Speer et al., 2008) does with ConceptNet data; it is also very important so that if we change the normalization process one day, the original data is not lost and…”
Commonsense Knowledge Base Completion (2016, methodology): “Our goals are similar to those of the Analogy-Space method (Speer et al., 2008), which uses matrix factorization to improve coverage of Concept-Net.”
AffectiveSpace: Blending Common Sense and Affective Knowledge to Perform Emotive Reasoning (2009, background): “AnalogySpace [9] is a way of representing a Common Sense knowledge base in a multidimensional vector space.”
Typed Tensor Decomposition of Knowledge Bases for Relation Extraction (2014, background): “For instance, Speer et al. (2008) aimed to create vectors of latent components for representing concepts in a common sense knowledge base us-ing SVD. Franz et al. (2009) proposed TripleRank to model the subject - predicate - object RDF triples in a tensor, and then applied the CP decomposition to…”
Modelling Relational Data using Bayesian Clustered Tensor Factorization (2009, methodology): “Models based on matrix or tensor factorization [18, 19, 3] ha ve the potential of making better predictions than interpretable models of similar complexity, as w e demonstrate in our experimental results section.”
Coarse Word-Sense Disambiguation Using Common Sense (2010, background): “AnalogySpace (Speer, Havasi, and Lieberman 2008) is a matrix representation of ConceptNet that is “smoothed” using dimensionality reduction.”
Commonsense Reasoning for Natural Language Understanding: A Survey of Benchmarks, Resources, and Approaches (2019, methodology): “AnalogySpace was originally applied to ConceptNet (Liu & Singh, 2004), and is included with the latest release of ConceptNet.”
Conversational Neuro-Symbolic Commonsense Reasoning (2020, methodology): “Examples of commonsense reasoning engines are: AnalogySpace [36,37] that uses dimensionality reduction, and [6] that uses the event calculus formal language.”
Common Sense Reasoning for Detection, Prevention, and Mitigation of Cyberbullying (2012, methodology): “This truncated SVD represents the approximation AAk = UkIkVkT.”