Faceted search with a large amount of properties. Exploring opportunities for faceted search for an e-commerce application with a large amount of dynamic properties

Abstract

Qalixa.com is a online e-commerce meta search engine that consolidate articles from a growing number of retailers from most lines of retail business. The articles in store entail a variety of meta data types such as properties, tags, category structures and free text. In this project, the author try to find ways to exploit the meta data and facilities available through a modern Enterprise search solution to create suggestions on how to build a User interface with sufficient competence to deal with this complex search space in an empowering manner, yet represent a simple enough experience for the end user to allow overview as well as a simple as possible learning curve. A few concepts are prototyped and a variety of options of future work are analyzed. Characteristic issues with applying faceted search to the particular case are found and analyzed to some detail. Concepts that are tested to some extent are selectable facet filters, grouping of properties with clustering, fuzzy logic in facet search and implicit ordering of documents given filter order.