Learning to rank with partially-labeled data. Duh, K. & Kirchhoff, K. In Proceedings of the 31st annual international ACM SIGIR conference on Research and development in information retrieval, of SIGIR '08, pages 251–258, New York, NY, USA, July, 2008. Association for Computing Machinery. ![Learning to rank with partially-labeled data [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper doi abstract bibtex Ranking algorithms, whose goal is to appropriately order a set of objects/documents, are an important component of information retrieval systems. Previous work on ranking algorithms has focused on cases where only labeled data is available for training (i.e. supervised learning). In this paper, we consider the question whether unlabeled (test) data can be exploited to improve ranking performance. We present a framework for transductive learning of ranking functions and show that the answer is affirmative. Our framework is based on generating better features from the test data (via KernelPCA) and incorporating such features via Boosting, thus learning different ranking functions adapted to the individual test queries. We evaluate this method on the LETOR (TREC, OHSUMED) dataset and demonstrate significant improvements.
@inproceedings{duh_learning_2008,
address = {New York, NY, USA},
series = {{SIGIR} '08},
title = {Learning to rank with partially-labeled data},
isbn = {978-1-60558-164-4},
url = {https://doi.org/10.1145/1390334.1390379},
doi = {10.1145/1390334.1390379},
abstract = {Ranking algorithms, whose goal is to appropriately order a set of objects/documents, are an important component of information retrieval systems. Previous work on ranking algorithms has focused on cases where only labeled data is available for training (i.e. supervised learning). In this paper, we consider the question whether unlabeled (test) data can be exploited to improve ranking performance. We present a framework for transductive learning of ranking functions and show that the answer is affirmative. Our framework is based on generating better features from the test data (via KernelPCA) and incorporating such features via Boosting, thus learning different ranking functions adapted to the individual test queries. We evaluate this method on the LETOR (TREC, OHSUMED) dataset and demonstrate significant improvements.},
urldate = {2021-10-15},
booktitle = {Proceedings of the 31st annual international {ACM} {SIGIR} conference on {Research} and development in information retrieval},
publisher = {Association for Computing Machinery},
author = {Duh, Kevin and Kirchhoff, Katrin},
month = jul,
year = {2008},
keywords = {boosting, information retrieval, kernel principal components analysis, learning to rank, transductive learning},
pages = {251--258},
}
Downloads: 0
{"_id":"qk5Frxerc9RTocD8z","bibbaseid":"duh-kirchhoff-learningtorankwithpartiallylabeleddata-2008","author_short":["Duh, K.","Kirchhoff, K."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","address":"New York, NY, USA","series":"SIGIR '08","title":"Learning to rank with partially-labeled data","isbn":"978-1-60558-164-4","url":"https://doi.org/10.1145/1390334.1390379","doi":"10.1145/1390334.1390379","abstract":"Ranking algorithms, whose goal is to appropriately order a set of objects/documents, are an important component of information retrieval systems. Previous work on ranking algorithms has focused on cases where only labeled data is available for training (i.e. supervised learning). In this paper, we consider the question whether unlabeled (test) data can be exploited to improve ranking performance. We present a framework for transductive learning of ranking functions and show that the answer is affirmative. Our framework is based on generating better features from the test data (via KernelPCA) and incorporating such features via Boosting, thus learning different ranking functions adapted to the individual test queries. We evaluate this method on the LETOR (TREC, OHSUMED) dataset and demonstrate significant improvements.","urldate":"2021-10-15","booktitle":"Proceedings of the 31st annual international ACM SIGIR conference on Research and development in information retrieval","publisher":"Association for Computing Machinery","author":[{"propositions":[],"lastnames":["Duh"],"firstnames":["Kevin"],"suffixes":[]},{"propositions":[],"lastnames":["Kirchhoff"],"firstnames":["Katrin"],"suffixes":[]}],"month":"July","year":"2008","keywords":"boosting, information retrieval, kernel principal components analysis, learning to rank, transductive learning","pages":"251–258","bibtex":"@inproceedings{duh_learning_2008,\n\taddress = {New York, NY, USA},\n\tseries = {{SIGIR} '08},\n\ttitle = {Learning to rank with partially-labeled data},\n\tisbn = {978-1-60558-164-4},\n\turl = {https://doi.org/10.1145/1390334.1390379},\n\tdoi = {10.1145/1390334.1390379},\n\tabstract = {Ranking algorithms, whose goal is to appropriately order a set of objects/documents, are an important component of information retrieval systems. Previous work on ranking algorithms has focused on cases where only labeled data is available for training (i.e. supervised learning). In this paper, we consider the question whether unlabeled (test) data can be exploited to improve ranking performance. We present a framework for transductive learning of ranking functions and show that the answer is affirmative. Our framework is based on generating better features from the test data (via KernelPCA) and incorporating such features via Boosting, thus learning different ranking functions adapted to the individual test queries. We evaluate this method on the LETOR (TREC, OHSUMED) dataset and demonstrate significant improvements.},\n\turldate = {2021-10-15},\n\tbooktitle = {Proceedings of the 31st annual international {ACM} {SIGIR} conference on {Research} and development in information retrieval},\n\tpublisher = {Association for Computing Machinery},\n\tauthor = {Duh, Kevin and Kirchhoff, Katrin},\n\tmonth = jul,\n\tyear = {2008},\n\tkeywords = {boosting, information retrieval, kernel principal components analysis, learning to rank, transductive learning},\n\tpages = {251--258},\n}\n\n\n\n","author_short":["Duh, K.","Kirchhoff, K."],"key":"duh_learning_2008","id":"duh_learning_2008","bibbaseid":"duh-kirchhoff-learningtorankwithpartiallylabeleddata-2008","role":"author","urls":{"Paper":"https://doi.org/10.1145/1390334.1390379"},"keyword":["boosting","information retrieval","kernel principal components analysis","learning to rank","transductive learning"],"metadata":{"authorlinks":{}},"html":""},"bibtype":"inproceedings","biburl":"https://bibbase.org/zotero/mh_lenguyen","dataSources":["XJ7Gu6aiNbAiJAjbw","XvjRDbrMBW2XJY3p9","3C6BKwtiX883bctx4","5THezwiL4FyF8mm4G","RktFJE9cDa98BRLZF","qpxPuYKLChgB7ox6D","PfM5iniYHEthCfQDH","iwKepCrWBps7ojhDx"],"keywords":["boosting","information retrieval","kernel principal components analysis","learning to rank","transductive learning"],"search_terms":["learning","rank","partially","labeled","data","duh","kirchhoff"],"title":"Learning to rank with partially-labeled data","year":2008}