Semantic Content Search on IKEA.com
摘要
In this paper, we present an approach to content search. The aim is to increase customer engagement with content recommendations on IKEA.com . As an alternative to Boolean search, we introduce a method based on semantic textual similarity between content pages and search queries. Our approach improves the relevance of search results by a 2.95% increase in click-through rate in an online A/B test.