This Title All WIREs
How to cite this WIREs title:
WIREs Comp Stat

Implicit translation

Full article on Wiley Online Library:   HTML PDF

Can't access this content? Tell your librarian.

Abstract This article describes a joint embedding approach to matching documents in multiple languages. The approach is general, operating on dissimilarity matrices, and is thus applicable to any problem of fusing disparate information. We apply it to a problem of implicit translation, in which documents in two different languages are matched. WIREs Comp Stat 2012, 4:28–34. doi: 10.1002/wics.181 This article is categorized under: Data: Types and Structure > Text Data Statistical Learning and Exploratory Methods of the Data Sciences > Text Mining

Manifold matching. Given observations from two different conditions—two sensor modalities, two different sensors, two different environmental conditions (outside versus inside for example)—where we observe only a subset of the classes under one of the conditions, we wish to develop a method for classifying new observations. Given the ‘manifold matching’ function φ the problem is straightforward. Without this function (which may not even exist in some real‐world situations), a methodology must be developed (as illustrated in the figure) to map the observations together into a space in which inference can be performed.

[ Normal View | Magnified View ]

Browse by Topic

Statistical Learning and Exploratory Methods of the Data Sciences > Text Mining
Data: Types and Structure > Text Data

Access to this WIREs title is by subscription only.

Recommend to Your
Librarian Now!

The latest WIREs articles in your inbox

Sign Up for Article Alerts