Future Work - Learning to Rank for Information Retrieval - page 247

20.9 Beyond Ranking
In recent years, there has been a trend in commercial search engines that goes be-
yond the pure relevance-based ranking of documents in their search results. For
example, the computational knowledge engine WolframAlpha.com , the decision en-
gine Bing.com , the universal search in Google.com , all try to provide rich presenta-
tion of search result to users.
When the ranked list is no longer the desired output, the learning-to-rank tech-
nologies need to be refined: the change of the output space will naturally lead to
the change of the hypothesis space and the loss function, as well as the change of
the learning theory. On the other hand, the new search scenario may be decomposed
into several sub ranking tasks and many key components in learning to rank can still
be used. This may become a promising future work for all the researchers currently
working on learning to rank, which we would like to call learning to search rather
than learning to rank .
Overall, this topic is just a stage-wise summary of the hot research field of learn-
ing to rank. Given the fast development of the field, we can foresee that many new
algorithms and theories will gradually arrive. We hope that this topic will motivate
more people to work on learning to rank, so as to make this research direction have
more impact in both the information retrieval and machine learning communities.
References
1. Chang, Y., Liu, T.Y.: Future directions in learning to rank. JMLR: Workshop and Conference
Proceedings 14 , 91-100 (2011)
2. Bickel, S., Scheffer, T.: Dirichlet-enhanced spam filtering based on biased samples. In: Ad-
vances in Neural Information Processing Systems 19 (NIPS 2006) (2007)
3. Cambazoglu, B., Zaragoza, H., Chapelle, O., Chen, J., Liao, C., Zheng, Z., Degenhardt, J.:
Early exit optimizations for additive machine learned ranking systems. In: Proceedings of
the Third ACM International Conference on Web Search and Data Mining (WSDM 2010),
pp. 411-420 (2010)
4. Chang, E., Zhu, K., Wang, H., Bai, H., Li, J., Qiu, Z., Cui, H.: Parallelizing support vector
machines on distributed computers. In: Platt, J., Koller, D., Singer, Y., Roweis, S. (eds.) Ad-
vances in Neural Information Processing Systems 20 (NIPS 2007), pp. 257-264. MIT Press,
Cambridge (2008)
5. Chu, C.T., Kim, S.K., Lin, Y.A., Yu, Y., Bradski, G., Ng, A., Olukotun, K.: Map-reduce for
machine learning on multicore. In: Schölkopf, B., Platt, J., Hoffman, T. (eds.) Advances in
Neural Information Processing Systems 19 (NIPS 2006), pp. 281-288. MIT Press, Cambridge
(2007)
6. Dong, A., Chang, Y., Zheng, Z., Mishne, G., Bai, J., Zhang, R., Buchner, K., Liao, C., Diaz, F.:
Towards recency ranking in web search. In: Proceedings of 3rd ACM International Conference
on Web Search and Data Mining (WSDM 2010) (2010)
7. Guiver, J., Snelson, E.: Bayesian inference for Plackett-Luce ranking models. In: Proceedings
of the 26th International Conference on Machine Learning (ICML 2009), pp. 377-384 (2009)
8. Inagaki, Y., Sadagopan, N., Dupret, G., Liao, C., Dong, A., Chang, Y., Zheng, Z.: Session
based click features for recency ranking. In: Proceedings of 24th AAAI Conference on Artifi-
cial Intelligence (AAAI 2010) (2010)