The Carrot2 clustering engine caught my interest today. It’s an “open source framework for building search clustering engines”. Carrot2 will basically take search results, analyze these on the fly, and group them according to most common concepts and keywords. Clusters can then be used for navigation and visualization (see the demo at http://search.carrot2.org/stable/search). Grokker (http://www.grokker.com) is using parts of this framework, apparently. Carrot2 can be integrated as a Java API, or you can interface with the standalone clustering server via XML or JSON. I hope I get to play with this technology sometime soon.
amazon bjørn olstad business goals concept composition concept design daniel tunkelang decision-making design patterns don tapscott efficiency enterprise search exploratory search faceted search facets FAST FF09 freebase google information retrieval ingenious search interaction design microsoft NLP paradox of choice personalization powerset recommendations recommenders relevancy satisficing searchme searchnuggets semantic seo sharepoint slideshare social topic pages traffic twitter usability user experience user needs visualization yggdrasil08