“You are what you tweet” is the new adage on ubiquitous social media nowadays. Social network data, including one’s own social network postings and those of network neighbors, can provide a good picture of a user’s interests. Therefore, folksonomy, or collaborative tagging, helps in classifying and categorizing content on the Internet.

This study uses preferences, together with frequency, recency, and duration, to garner information, which is something that has not been done previously. Thus, this paper is a major and unique contribution to knowledge on the issue. Frequency represents the strength of the user’s preference; recency, closeness to the current time period; and duration, the continuity of the user’s preference for tag-based information. These three factors strengthen information obtained from blogs, commercial products, movies, music, news, and photographs.

Data used in this study was collected from Delicious.com, a popular social bookmarking website. A robot was designed to crawl Delicious.com collecting tags, items, active users, and friend networks. The results of this study seem to indicate that “the hybridization of a user’s preferences with frequency, recency, and duration[, as well as information about friends and neighbors,] plays an important role” in identifying preferences in a much better way than traditional collaborative recommendation systems.

The authors are aware of research limitations brought about by the fact that a relatively small number of users were tested in this study compared to the large number of users on the Delicious.com website.

Those in the marketing field would be very interested in the results of this study as a way of determining user preferences for products that they introduce to various audiences, since finding similar users that have similar preferences is an important aspect of moving their products.