Material Detail

BoltzRank: Learning to Maximize Expected Ranking Gain

BoltzRank: Learning to Maximize Expected Ranking Gain

This video was recorded at 26th International Conference on Machine Learning (ICML), Montreal 2009. Ranking a set of retrieved documents according to their relevance to a query is a popular problem in information retrieval. Methods that learn ranking functions are difficult to optimize, as ranking performance is typically judged by metrics that are not smooth. In this paper we propose a new listwise approach to learning to rank. Our method creates a conditional probability distribution over rankings assigned to documents for a given query, which allows for gradient ascent optimization of the expected value of some performance measure. The rank probabilities take the form of a Boltzmann distribution, based on an energy function that depends on a scoring function composed of individual and... Show More


  • User Rating
  • Comments
  • Learning Exercises
  • Bookmark Collections
  • Course ePortfolios
  • Accessibility Info

More about this material


Log in to participate in the discussions or sign up if you are not already a MERLOT member.