Intimate Learning: A Novel Approach for Combining Labelled and Unlabelled Data
Abstract
This paper introduces a new bootstrapping method closely related to co-training and scoped-learning. The method is tested on a Web information extraction task of learning course names from web pages in which we use very few labelled items as seed data (10 web pages) and combine with an unlabelled set (174 web pages). The overall performance improved the precision/recall from 3.11%/0.31 % for a baseline EM-based method to 44.7%/44.1 % for intimate learning. 1 Intimate learning The expensive nature of labelling data for machine learning methods and the lack of success in using purely unsupervised methods have motivated the study of learning methods
Cite
Text
Shi and Sarkar. "Intimate Learning: A Novel Approach for Combining Labelled and Unlabelled Data." International Joint Conference on Artificial Intelligence, 2005.Markdown
[Shi and Sarkar. "Intimate Learning: A Novel Approach for Combining Labelled and Unlabelled Data." International Joint Conference on Artificial Intelligence, 2005.](https://mlanthology.org/ijcai/2005/shi2005ijcai-intimate/)BibTeX
@inproceedings{shi2005ijcai-intimate,
title = {{Intimate Learning: A Novel Approach for Combining Labelled and Unlabelled Data}},
author = {Shi, Zhongmin and Sarkar, Anoop},
booktitle = {International Joint Conference on Artificial Intelligence},
year = {2005},
pages = {1634-1635},
url = {https://mlanthology.org/ijcai/2005/shi2005ijcai-intimate/}
}