Cross-training
Published 24 August 2003
Sunita Sarawagi, Soumen Chakrabarti, Shantanu Godbole
Citations50
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
The distributional and discriminative algorithms for cross-training are given and it is shown that cross- training can discover and exploit probabilistic relations between two taxonomies for more accurate classification.
Abstract
Classification is a well-established operation in text mining. Given a set of labels A and a set DA of training documents tagged with these labels, a classifier learns to assign labels to unlabeled test documents. Suppose we also had available a different set of labels B, together with a set of documents DB marked with labels from B. If A and B have some semantic overlap, can the availability of DB help us build a better classifier for A, and vice versa? We answer this question in the affirmative by proposing cross-training: a new approach to semi-supervised learning in presence of multiple label sets. We give distributional and discriminative algorithms for cross-training and show, through extensive experiments, that cross-training can discover and exploit probabilistic relations between two taxonomies for more accurate classification.
Keywords
Computer Science
Journal of the Royal Statistical Society Series B (Statistical Methodology)Maximum Likelihood from Incomplete Data Via the <i>EM</i> Algorithm
49,657 Citations1977A. P. Dempster, N. M. Laird +1 more
Machine LearningMultitask Learning
6,312 Citations1997Rich Caruana
Prior work on MTL is reviewed, new evidence that MTL in backprop nets discovers task relatedness without the need of supervisory signals is presented, and new results for MTL with k-nearest neighbor and kernel regression are presented.
International Conference on Neural Information ProcessingAdvances in kernel methods: support vector learning
5,815 Citations1999Bernhard Schölkopf, Christopher J. C. Burges +1 more
Support vector machines for dynamic reconstruction of a chaotic system, Klaus-Robert Muller et al pairwise classification and support vector machines, Ulrich Kressel.
Combining labeled and unlabeled data with co-training
5,604 Citations1998Avrim Blum, Tom M. Mitchell
Technical reportsMaking Large-Scale SVM Learning Practical
4,317 Citations2006Thorsten Joachims
This chapter presents algorithmic and computational results developed for SVM light V 2.0, which make large-scale SVM training more practical and give guidelines for the application of SVMs to large domains.
A comparison of event models for naive bayes text classification
3,224 Citations1998Andrew McCallum, Kamal Nigam
It is found that the multi-variate Bernoulli performs well with small vocabulary sizes, but that the multinomial performs usually performs even better at larger vocabulary sizes--providing on average a 27% reduction in error over the multi -variateBernoulli model at any vocabulary size.
Machine LearningText Classification from Labeled and Unlabeled Documents using EM
2,749 Citations2000Kamal Nigam, Andrew Kachites McCallum +2 more
This paper shows that the accuracy of learned text classifiers can be improved by augmenting a small number of labeled training documents with a large pool of unlabeled documents, and presents two extensions to the algorithm that improve classification accuracy under these conditions.
Sequential Minimal Optimization : A Fast Algorithm for Training Support Vector Machines
2,685 Citations1998John Platt
A re-examination of text categorization methods
2,660 Citations1999Yiming Yang, Xin Liu
The results show that SVM, kNN and LLSF signi cantly outperform NNet and NB when the number of positive training instances per category are small, and that all the methods perform comparably when the categories are over 300 instances.
Inductive learning algorithms and representations for text categorization
1,465 Citations1998Susan Dumais, John Platt +2 more
A comparison of the effectiveness of five different automatic learning algorithms for text categorization in terms of learning speed, realtime classification speed, and classification accuracy is compared.
Multitask learning
1,118 Citations1998Rich Caruana
Learning to map between ontologies on the semantic web
891 Citations2002AnHai Doan, Jayant Madhavan +2 more
Glue is described, a system that employs machine learning techniques to find semantic mappings between ontologies and is distinguished in that it works with a variety of well-defined similarity notions and that it efficiently incorporates multiple types of knowledge.
Using Maximum Entropy for Text Classification
756 Citations1999Kamal Nigam, John Lafferty +1 more
This paper uses maximum entropy techniques for text classification by estimating the conditional distribution of the class variable given the document by comparing accuracy to naive Bayes and showing that maximum entropy is sometimes significantly better, but also sometimes worse.
Partially Supervised Classification of Text Documents
516 Citations2002Bing Liu, Wee Sun Lee +2 more
This paper shows that the problem of identifying documents from a set of documents of a particular topic or class P and a large set M of mixed documents, and that under appropriate conditions, solutions to the constrained optimization problem will give good solution to the partially supervised classification problem.
PEBL
277 Citations2002Hwanjo Yu, Jiawei Han +1 more
The Positive Example Based Learning (PEBL) framework for Web page classification is introduced which eliminates the need for manually collecting negative training examples in pre-processing and an algorithm called Mapping-Convergence (M-C) is presented that achieves classification accuracy as high as that of traditional SVM (with positive and negative data).
Discovering Structure in Multiple Learning Tasks: The TC Algorithm.
202 Citations1996Sebastian Thrun, Joseph A. O’Sullivan
The task-clustering algorithm TC clusters learning tasks into classes of mutually related tasks, and outperforms its non-selective counterpart in situations where only a small number of tasks is relevant.
On integrating catalogs
128 Citations2001Rakesh Agrawal, Ramakrishnan Srikant
This work shows how a Naive Bayes classification can be enhanced to incorporate the similarity information present in source catalogs and shows substantial improvement in the accuracy of catalog integration.
Computer Networks and ISDN SystemsAutomatically organizing bookmarks per contents
89 Citations1996Yoelle Maarek, Israel Z. Ben Shaul
A novel approach to bookmark organization is presented that provides automatic classification together with user adaption and is presented as a solution to the need for personalized client-based local URL repositories often called bookmarks.
Computer NetworksPowerBookmarks: a system for personalizable Web information organization, sharing, and management
73 Citations1999Wen‐Syan Li, Quoc Vu +3 more
The notion of bookmark management is extended by introducing the functionalities of hypermedia databases, and PowerBookmarks supports advanced query, classification, and navigation functionalities on collections of bookmarks.
A model of inductive bias learning
25 Citations2000BaxterJonathan
A natural law of succession
11 Citations2002Eric Sven Ristad
A new solution to multinomial estimation using combinatorial priors on strings is presented and it is demonstrated that this solution outperforms standard solutions both in theory and in practice.