Similarity metrics within a point of view

Publisher:
Springer Link
Publication Type:
Conference Proceeding
Citation:
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2011, 7052 LNCS pp. 13 - 24
Issue Date:
2011-11-02
Full metadata record
Files in This Item:
Filename Description Size
QI_SimiliarMetrics.pdfPublished version223 kB
Adobe PDF
Vector space based approaches to natural language processing are contrasted with human similarity judgements to show the manner in which human subjects fail to produce data which satisfies all requirements for a metric space. This result would constrains the validity and applicability vector space based (and hence also quantum inspired) approaches to the modelling of cognitive processes. This paper proposes a resolution to this problem, by arguing that pairs of words imply a context which in turn induces a point of view, so allowing a subject to estimate semantic similarity. Context is here introduced as a point of view vector (POVV) and the expected similarity is derived as a measure over the POVV's. Different pairs of words will invoke different contexts and different POVV's. We illustrate the proposal on a few triples of words and outline further research. © 2011 Springer-Verlag.
Please use this identifier to cite or link to this item: