FINDING AND CLASSIFYING PRODUCT RELATIONSHIPS USING INFORMATION FROM THE PUBLIC WEB
Daniel Schuster, Till M. Juchheim, Alexander Schill · 2010
Relationships between products such as accessory or successor products are hard to find on the Web or have to be inserted manually in product information systems. Finding and classifying such relations automatically using information from the public Web only offers great value for customers and vendors as it helps to improve the buying process at low cost. We present and evaluate algorithms and methods for product relationship extraction on the Web requiring only a set of clustered product names as input. The solution can be easily implemented in different product information systems most useful in but not necessarily restricted to the application domain of online shopping.