Automated complementary association learning from web documents

Abstract

In recent years, much effort has been put in finding semantic associations between items. One type of these associations is complementary association between items which determines an item that its usage is interrelated with the use of an associated or paired item such that a demand for one generates demand for the other. This association has many applications in the field of economics and marketing. This paper presents a novel contribution in this area, proposing an automatic and unsupervised method for acquiring complementary associations between products in a product catalog, framed in the context of domain ontology learning, using the Web as corpus. The paper also discusses how obtained associations can be automatically evaluated against WordNet and presents encouraging results for several categories of products. © 2009 Praise Worthy Prize S.r.l. - All rights reserved.