Learning Rules from Incomplete Examples via Implicit Mention Models

Janardhan Rao Doppa, Mohammad Shahed Sorower, Mohammad Nasresfahani, Jed Irvine, Walker Orr, Thomas G. Dietterich, Xiaoli Fern, Prasad Tadepalli
Proceedings of the Asian Conference on Machine Learning, PMLR 20:197-212, 2011.

Abstract

We study the problem of learning general rules from concrete facts extracted from natural data sources such as the newspaper stories and medical histories. Natural data sources present two challenges to automated learning, namely, radical incompleteness and systematic bias. In this paper, we propose an approach that combines simultaneous learning of multiple predictive rules with differential scoring of evidence which adapts to a presumed model of data generation. Learning multiple predicates simultaneously mitigates the problem of radical incompleteness, while the differential scoring would help reduce the effects of systematic bias. We evaluate our approach empirically on both textual and non-textual sources. We further present a theoretical analysis that elucidates our approach and explains the empirical results.

Cite this Paper

BibTeX
@InProceedings{pmlr-v20-doppa11, title = {Learning Rules from Incomplete Examples via Implicit Mention Models}, author = {Doppa, Janardhan Rao and Sorower, Mohammad Shahed and Nasresfahani, Mohammad and Irvine, Jed and Orr, Walker and Dietterich, Thomas G. and Fern, Xiaoli and Tadepalli, Prasad}, booktitle = {Proceedings of the Asian Conference on Machine Learning}, pages = {197--212}, year = {2011}, editor = {Hsu, Chun-Nan and Lee, Wee Sun}, volume = {20}, series = {Proceedings of Machine Learning Research}, address = {South Garden Hotels and Resorts, Taoyuan, Taiwain}, month = {14--15 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v20/doppa11/doppa11.pdf}, url = {https://proceedings.mlr.press/v20/doppa11.html}, abstract = {We study the problem of learning general rules from concrete facts extracted from natural data sources such as the newspaper stories and medical histories. Natural data sources present two challenges to automated learning, namely, radical incompleteness and systematic bias. In this paper, we propose an approach that combines simultaneous learning of multiple predictive rules with differential scoring of evidence which adapts to a presumed model of data generation. Learning multiple predicates simultaneously mitigates the problem of radical incompleteness, while the differential scoring would help reduce the effects of systematic bias. We evaluate our approach empirically on both textual and non-textual sources. We further present a theoretical analysis that elucidates our approach and explains the empirical results.} }
Endnote
%0 Conference Paper %T Learning Rules from Incomplete Examples via Implicit Mention Models %A Janardhan Rao Doppa %A Mohammad Shahed Sorower %A Mohammad Nasresfahani %A Jed Irvine %A Walker Orr %A Thomas G. Dietterich %A Xiaoli Fern %A Prasad Tadepalli %B Proceedings of the Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2011 %E Chun-Nan Hsu %E Wee Sun Lee %F pmlr-v20-doppa11 %I PMLR %P 197--212 %U https://proceedings.mlr.press/v20/doppa11.html %V 20 %X We study the problem of learning general rules from concrete facts extracted from natural data sources such as the newspaper stories and medical histories. Natural data sources present two challenges to automated learning, namely, radical incompleteness and systematic bias. In this paper, we propose an approach that combines simultaneous learning of multiple predictive rules with differential scoring of evidence which adapts to a presumed model of data generation. Learning multiple predicates simultaneously mitigates the problem of radical incompleteness, while the differential scoring would help reduce the effects of systematic bias. We evaluate our approach empirically on both textual and non-textual sources. We further present a theoretical analysis that elucidates our approach and explains the empirical results.
RIS
TY - CPAPER TI - Learning Rules from Incomplete Examples via Implicit Mention Models AU - Janardhan Rao Doppa AU - Mohammad Shahed Sorower AU - Mohammad Nasresfahani AU - Jed Irvine AU - Walker Orr AU - Thomas G. Dietterich AU - Xiaoli Fern AU - Prasad Tadepalli BT - Proceedings of the Asian Conference on Machine Learning DA - 2011/11/17 ED - Chun-Nan Hsu ED - Wee Sun Lee ID - pmlr-v20-doppa11 PB - PMLR DP - Proceedings of Machine Learning Research VL - 20 SP - 197 EP - 212 L1 - http://proceedings.mlr.press/v20/doppa11/doppa11.pdf UR - https://proceedings.mlr.press/v20/doppa11.html AB - We study the problem of learning general rules from concrete facts extracted from natural data sources such as the newspaper stories and medical histories. Natural data sources present two challenges to automated learning, namely, radical incompleteness and systematic bias. In this paper, we propose an approach that combines simultaneous learning of multiple predictive rules with differential scoring of evidence which adapts to a presumed model of data generation. Learning multiple predicates simultaneously mitigates the problem of radical incompleteness, while the differential scoring would help reduce the effects of systematic bias. We evaluate our approach empirically on both textual and non-textual sources. We further present a theoretical analysis that elucidates our approach and explains the empirical results. ER -
APA
Doppa, J.R., Sorower, M.S., Nasresfahani, M., Irvine, J., Orr, W., Dietterich, T.G., Fern, X. & Tadepalli, P.. (2011). Learning Rules from Incomplete Examples via Implicit Mention Models. Proceedings of the Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 20:197-212 Available from https://proceedings.mlr.press/v20/doppa11.html.

Related Material