Leveraging browsing patterns for topic discovery and photostream recommendation

In photo-sharing websites and in social networks, pho-tographs are most often browsed as a sequence: users who view a photo are likely to click on those that fol-low. The sequences of photos (which we call photo-streams), as opposed to individual images, can there-fore be considered to be very important content units in their own right. In spite of their importance, those se-quences have received little attention even though they are at the core of how people consume image content. In this paper, we focus on photostreams. First, we per-form an analysis of a large dataset of user logs contain-ing several million pageviews, examining navigation patterns between photostreams. Based on observations from the analysis, we build a stream transition graph to analyze common stream topic transitions (e.g., users of-ten view “train ” photostreams followed by “firetruck” photostreams). We then implement two stream recom-mendation algorithms, based on collaborative filtering and on photo tags, and report the results of a user study involving 40 participants. Our analysis yields interest-ing insights into how people navigate between photo-streams, while the results of the user study provide use-ful feedback for evaluating the performance and charac-teristics of the recommendation algorithms.