Current LETOR baselines

• Ranking SVM
• RankBoost
• AdaRank
• Multiple hyperline ranker
• FRank
• ListNet

The performance of the above baseline algorithms are summarized here.

For the experimental setting, please refer to the paper: Tie-Yan Liu, Tao Qin, Jun Xu, Wenying Xiong and Hang Li, LETOR: Benchmark dataset for research on learning to rank for information retrieval, LR4IR 2007, in conjunction with SIGIR 2007.

Submit a LETOR baseline

If you want to add your own algorithm to LETOR as a baseline, you may need to provide the following data

• An excel file with a similar format to that provided in LETOR
• A html page describing your baseline using this template.
• A model file, and a test script so that we can validate your results on the testset if necessary
• Scores outputted by your model for each testing document.
• A link to the document (or the paper) describing how your method was implemented

Then, please ZIP these files and send it to letor@microsoft.com.