Inferring a Semantic Representation of Text via Cross-Language Correlation Analysis

By Alexei Vinokourov, John Shawe-Taylor, Nello Cristianini. 2002

303 citations recorded by the source corpus.

Subfields: Natural Language Processing, Representation & Self-Supervised Learning

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

In Representation & Self-Supervised Learning

Citations in the corpus

Beyond CCA: Moment Matching for Multi-View Models (2016, methodology): “Examples of such data include, for instance, representation of some text in two languages (e.g., Vinokourov et al., 2002) or images aligned with text data (e.g., Hardoon et al., 2004; Gong et al., 2014).”
Learning Continuous Phrase Representations for Translation Modeling (2014, background): “As a result, variants of such models for cross-lingual scenarios have been proposed so that words in different languages are projected into the shared latent vector space (Dumais et al. 1997; Platt et al. 2010; Vinokourov et al. 2002; Yih et al. 2011; Gao et al. 2011; Huang et al. 2013; Zou et al. 2”
Domain Adaptation for Visual Applications: A Comprehensive Survey (2017, background): “We can also have multi-view situations even when the views have the same data modalities (textual, visual, audio), such as in the case of parallel texts in different languages [236, 79], photos of the same person taken across different poses, illuminations and expressions [261, 195, 196, 170].”
Bayesian Canonical correlation analysis (2013, methodology): “The Bayesian CCA as presented by Wang (2007) and Klami and Kaski (2007) models the viewspecific variation with a free covariance parameter.”
Deep Canonical Correlation Analysis (2013, methodology): “CCA and KCCA have been used for unsupervised data analysis when multiple views are available (Hardoon et al., 2007; Vinokourov et al., 2003; Dhillon et al., 2011); learning features for multiple modalities that are then fused for prediction (Sargin et al.”
On the Role of Correlation and Abstraction in Cross-Modal Multimedia Retrieval (2014, methodology): “Multimodal integration has also been applied to retrieval tasks including audiovisual content [42], [43].”
A web-based kernel function for measuring the similarity of short text snippets (2006, background): “Previous research has also tried to address learning a semantic representation for a document by using cross-lingual techniques [18].”
Learning Discriminative Projections for Text Similarity Measures (2011, background): “Canonical Correlation Analysis (CCA) (Vinokourov et al., 2003) ﬁnds projections that maximize the cross-covariance between the projected vectors.”
Learning Semantic Representations for the Phrase Translation Model (2013, methodology): “2010) and canonical correlation analysis (Vinokourov et al. 2002) have been applied to cross-lingual IR, where documents in different languages are projected into the shared latent concept space.”
Clickthrough-based latent semantic models for web search (2011, background): “[30] Vinokourov, A.”
Canonical Correlation Analysis: An Overview with Application to Learning Methods (2004, unknown): “[19] with improved results.”
From Dependence to Causation (2016, background): “Consider for instance of document translation (Vinokourov et al., 2002), where the training data is a collection of documents in two different languages, let us say English and Spanish.”
Translingual Document Representations from Discriminative Projections (2010, methodology): “The closest previous work to this paper is the use of Canonical Correlation Analysis (CCA) to ﬁnd projections for multiple languages whose results are maximally correlated with each other (Vinokourov et al., 2003).”
Learning Multi-View Neighborhood Preserving Projections (2011, background): “…exists regarding the integration of multi-view data, and the methods that do exists are often tailored with specific application domains in mind: Vinokourov et al. (2002) and Hardoon et al. (2004) use canonical correlation analysis to construct a joint feature space from image and text data.”
On Deep Multi-View Representation Learning (2015, methodology): “CCA and KCCA have long been the workhorse for multi-view feature learning and dimensionality reduction ( Vinokourov et al., 2003; Kakade and Foster, 2007; Socher and Li, 2010; Dhillon et al., 2011).”