The Holy Grail of Transactions: Scalability & Instruction Locality
Traditional online transaction processing (OLTP) systems face two major challenges while trying to utilize modern hardware: (1) exploiting the abundant thread-level parallelism given by multicores and (2) taking advantage of the aggressive micro-architectural features. On…
Opinion Dynamics in Social Networks
Understanding how users form and update their opinions through interactions with their friends and neighbors in social networks is a fundamental problem with applications to online advertising, recommendations, and marketing. In our research, we carried…
Beyond voice recognition: Windows Phone’s Cortana anticipates your needs
Cortana, the world’s first truly personal digital assistant, is available soon on Windows Phone 8.1. Powered by Bing, Cortana is driven by state-of-the-art algorithms using natural language, machine learning, and contextual signals that benefit from…
Planetary Predictions
Societies and governments around the world want to know how the biosphere is likely to change and what we can do to avoid or mitigate against it. Current models used to provide that information, though,…
Quick Answers From Large Data
Tempe is an interactive system for exploring large data sets. It accelerates faster machine learning by facilitating quick, iterative feature engineering and data understanding. Tempe is a combination of three technologies: Trill: a high-speed, temporal,…