Reference:

Jaakko J. Väyrynen and Timo Honkela. Comparison of independent component analysis and singular value decomposition in word context analysis. In Timo Honkela, Ville Könönen, Matti Pöllä, and Olli Simula, editors, Proceedings of AKRR'05, International and Interdisciplinary Conference on Adaptive Knowledge Representation and Reasoning, pages 135–140. Helsinki University of Technology, Laboratory of Computer and Information Science, June 2005.

Abstract:

In earlier studies we have been able show that Independent Component Analysis is able to extract automatically meaningful linguistic features. The emergent syntactic and semantic features are based on an analysis of the words in their contexts in a large corpus. We have also shown that there is a reasonably strong correlation between traditional features and categories defined by linguists and the emergent features. In this article, we introduce a new measure for comparing the emergent and the traditionally defined features. We apply this measure to compare the emergent features produced by Singular Value Decomposition (SVD) and Independent Component Analysis (ICA). The conclusion is that the ICA-based features correspond to the human intuitions much more closely than the SVD-based features not only in a visual inspection but also in a systematic and principled comparison.

Suggested BibTeX entry:

@inproceedings{Vayrynen05AKRR05,
    author = {Jaakko J. V{\"{a}}yrynen and Timo Honkela},
    booktitle = {Proceedings of {AKRR'05}, International and Interdisciplinary Conference on Adaptive Knowledge Representation and Reasoning},
    editor = {Timo Honkela and Ville K{\"{o}}n{\"{o}}nen and Matti P{\"{o}}ll{\"{a}} and Olli Simula},
    month = {June},
    pages = {135--140},
    publisher = {Helsinki University of Technology, Laboratory of Computer and Information Science},
    title = {Comparison of Independent Component Analysis and Singular Value Decomposition in Word Context Analysis},
    year = {2005},
}

See www.researchgate.net ...