Blogging was where we began, and how we built our company so we have preserved this archive to show how our thinking developed over a decade of developing the use of social technology inside organisations

Associative Link Analysis in data mining


Judith Meskill’s knowledge notes points us to social networking – the link is the thing…, which pulls together a series of three recent articles from DM Review “The Premier Publication for Business Intelligence and Analytics” – that investigate the so-called Small World concept and how it might be used in data mining applications
Briefly, the first article looks at the Small World property of complex networked systems, and recalls Barabási’s analysis of hubs, nodes and links per node with Web sites and the role of “weak links” and “tipping points” in networks with high degrees of the SW property. The second looks at how this kind of social network analysis can be applied to data mining customer information and it suggests a moethodology the author calls Associative Link Analysis that can derive visual representations of aspects of associative relationships between events in the data set. Finally, the third article in the series discusses metrics than can be used to measure quantify these aspects, suggests strategies and tactics that such analysis can support (e.g. averting, maximising or maintaining levels of the SW property in netwokrs for different reasons), and recommends ways to implement these strategies in data warehousing and business ntelligence operations
It is heady often seemingly obscure stuff, but it shows how data mining and business inteligence operations are trying to operationalise the hitherto esoteric theories and observations of social network analysts in a large-scale commercial setting.

Comments are closed.