SoftRank: optimizing non-smooth rank metrics
The DOI matches its Crossref record; 14 of the 14 references it cites were found in a registry or an index.
Abstract
We address the problem of learning large complex ranking functions. Most IR applications use evaluation metrics that depend only upon the ranks of documents. However, most ranking functions generate document scores, which are sorted to produce a ranking. Hence IR metrics are innately non-smooth with respect to the scores, due to the sort. Unfortunately, many machine learning algorithms require the gradient of a training objective in order to perform the optimization of the model parameters,and because IR metrics are non-smooth,we need to find a smooth proxy objective that can be used for training. We present a new family of training objectives that are derived from the rank distributions of documents, induced by smoothed scores. We call this approach SoftRank. We focus on a smoothed approximation to Normalized Discounted Cumulative Gain (NDCG), called SoftNDCG and we compare it with three other training objectives in the recent literature. We present two main results. First, SoftRank yields a very good way of optimizing NDCG. Second, we show that it is possible to achieve state of the art test set NDCG results by optimizing a soft NDCG objective on the training set with a different discount function
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Probability, Random Variables, and Stochastic Processes.
Abrahams, J., & Papoulis, A. (1984) · Journal of the American Statistical Association · doi:10.2307/2288754 · Cited by 16,226
Optimizing search engines using clickthrough data
Joachims, T. (2002) · Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining · doi:10.1145/775047.775067 · Cited by 3,873
Learning to rank using gradient descent
Burges, C., Shaked, T., Renshaw, E., Lazier, A., Deeds, M., Hamilton, N., & Hullender, G. (2005) · Proceedings of the 22nd international conference on Machine learning - ICML '05 · doi:10.1145/1102351.1102363 · Cited by 2,836 · cited 14×
“RankNet [4] is a probabilistic model for pairwise preferences.”
BackgroundMethodInfluentialLearning to rank: from pairwise approach to listwise approach
Cao, Z., Qin, T., Liu, T.-Y., Tsai, M.-F., & Li, H. (2007) · Proceedings of the 24th international conference on Machine learning · doi:10.1145/1273496.1273513 · Cited by 1,960 · cited 1×
“A more recent approach [5] sets out to define a probability distribution over rankings: in other words, the event space consists of document permutations.”
BackgroundIR evaluation methods for retrieving highly relevant documents
Järvelin, K., & Kekäläinen, J. (2017) · ACM SIGIR Forum · doi:10.1145/3130348.3130374 · Cited by 1,444
Large Margin Rank Boundaries for Ordinal Regression
Herbrich, R., Graepel, T., & Obermayer, K. (2000) · Advances in Large-Margin Classifiers · doi:10.7551/mitpress/1113.003.0010 · Cited by 1,074
Improving web search ranking by incorporating user behavior information
Agichtein, E., Brill, E., & Dumais, S. (2006) · Proceedings of the 29th annual international ACM SIGIR conference on Research and development in information retrieval · doi:10.1145/1148170.1148177 · Cited by 1,041
(2006) · doi:10.7551/mitpress/7503.003.0029 · Cited by 753
Simple BM25 extension to multiple weighted fields
Robertson, S., Zaragoza, H., & Taylor, M. (2004) · Proceedings of the thirteenth ACM international conference on Information and knowledge management · doi:10.1145/1031171.1031181 · Cited by 701
Self-citationCrammer, K., & Singer, Y. (2002) · Advances in Neural Information Processing Systems 14 · doi:10.7551/mitpress/1120.003.0087 · Cited by 538
Gaussian Processes for Ordinal Regression
Chu, W., & Ghahramani, Z. (2005) · Journal of Machine Learning Research · Cited by 514
Caruana, R., Baluja, S., & Mitchell, T. M. (1995) · neural information processing systems · Cited by 126
Optimisation methods for ranking functions with multiple parameters
Taylor, M., Zaragoza, H., Craswell, N., Robertson, S., & Burges, C. (2006) · Proceedings of the 15th ACM international conference on Information and knowledge management - CIKM '06 · doi:10.1145/1183614.1183698 · Cited by 115
Self-citationImproving Web Search Ranking by Incorporating User Behavior
Agichtein, E., Brill, E., & Dumais, S. (2006) · Cited by 76
Cited by
Learning to rank: from pairwise approach to listwise approach
Cao, Z., Qin, T., Liu, T.-Y., Tsai, M.-F., & Li, H. (2007) · Proceedings of the 24th international conference on Machine learning · doi:10.1145/1273496.1273513 · Cited by 1,960
Learning to rank for information retrieval
Liu, T.-Y. (2010) · Proceedings of the 33rd international ACM SIGIR conference on Research and development in information retrieval · doi:10.1145/1835449.1835676 · Cited by 1,856
Learning to Rank for Information Retrieval
Liu, T.-Y. (2009) · Foundations and Trends® in Information Retrieval · doi:10.1561/1500000016 · Cited by 1,568
Learning diverse rankings with multi-armed bandits
Radlinski, F., Kleinberg, R., & Joachims, T. (2008) · Proceedings of the 25th international conference on Machine learning - ICML '08 · doi:10.1145/1390156.1390255 · Cited by 503
LETOR: A benchmark collection for research on learning to rank for information retrieval
Qin, T., Liu, T.-Y., Xu, J., & Li, H. (2010) · Information Retrieval · doi:10.1007/s10791-009-9123-y · Cited by 461
Metadata
- Authors
- Taylor, M., Guiver, J., Robertson, S., & Minka, T.
- Venue
- Proceedings of the international conference on Web search and web data mining - WSDM '08
- Publisher
- ACM Press
- Published
- 2008
- Pages
- 77
- OpenAlex
- W2059001985
- Cited by
- 322
- References
- 14
- Type
- conference abstract
- Language
- en
- Open access
- closed
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.