Box has vast amounts of content in its stores, and as it begins to apply artificial intelligence and machine learning to make it easier to surface, it also wants to expose each piece of content and how it relates to other content and users. To help achieve that, the company announced the Box Graph today at BoxWorks. "The Box Graph enables customers to predictively recommend content that might be relevant to you. It would be impossible to do this without a graph," Jeetu Patel, Chief Product Officer at Box, told TechCrunch. The idea of a graph was first popularized by Facebook, which discussed an individual's social graph, the connections they had between friends on the social network.

Online content have become an important medium to disseminate information and express opinions. With their proliferation, users are faced with the problem of missing the big picture in a sea of irrelevant and/or diverse content. In this paper, we addresses the problem of information organization of online document collections, and provide algorithms that create a structured representation of the otherwise unstructured content. We leverage the expressiveness of latent probabilistic models (e.g., topic models) and non-parametric Bayes techniques (e.g., Dirichlet processes), and give online and distributed inference algorithms that scale to terabyte datasets and adapt the inferred representation with the arrival of new documents. This paper is an extended abstract of the 2012 ACM SIGKDD best doctoral dissertation award of Ahmed [2011].