Learning to Classify Text from Labeled and Unlabeled Documents

By Kamal Nigam, Andrew McCallum, Sebastian Thrun, Tom Michael Mitchell. 1998

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Machine learning in automated text categorization (2001, neutral): “—the 20 Newsgroups collection, set up by Lang [1995] and used in [Baker and McCallum 1998; Joachims 1997; McCallum and Nigam 1998; McCallum et al. 1998; Nigam et al. 1998; Schapire and Singer 2000].”
Semi-Supervised Self-Training of Object Detection Models (2005, methodology): “However, Nigam in [14] also found that there were instances in which EM did not perform well.”
Forward and Backward Knowledge Transfer for Sentiment Classification (2019, neutral): “Na¨ıve Bayes (NB) text classiﬁer is a generative model (Nigam et al., 1998).”
Unsupervised named-entity extraction from the Web: An experimental study (2005, unknown): “IE systems that use bootstrapping include [37, 1, 6, 33, 12, 9].”
Cluster Kernels for Semi-Supervised Learning (2002, background): “It has been shown experimentally that under certain conditions, the decision function can be estimated more accurately, yielding lower generalization error [1] [4] [6] .”
Partially Supervised Classification of Text Documents (2002, methodology): “Recently, it has been shown in (Nigam et al., 1998) that unlabeled data is helpful in classifier building.”
The Reuters Corpus Volume 1 -from Yesterday’s News to Tomorrow’s Language Resources (2002, background): “An example story is shown in Appendix A (with some truncation of the metadata to conserve space).”
Learning to construct knowledge bases from the World Wide Web (2000, methodology): “Then, average mutual information is I(C;Wi) = H(C) H(CjWi) (7) = Xc2C Pr(c) log(Pr(c)) X vi2fwi;:wigPr(vi)Xc2CPr(cjvi) log(Pr(cjvi)) = X vi2fwi;:wigXc2C Pr(c; vi) log Pr(c; vi) Pr(c) Pr(vi)! (8) This feature selection method has been found to perform best among several alternatives [63], and has bee”
A Brief Survey of Text Mining: Classification, Clustering and Extraction Techniques (2017, background): “Many different variations of multinomial model have been introduced in [76, 97, 101, 106].”
Machine Learning for User Modeling (2001, unknown): “For example, reducing the need for labeled training data through expectation maximization (Nigam et al., 1998) leads to improved predictive performance, but causes a signi¢cant increase in CPU time.”
A Survey of Text Clustering Algorithms (2012, methodology): “In particular, word-clusters [17] [18] and co-training methods [72] can be used in order to improve the classification accuracy of supervised applications with the use of clustering techniques.”
Computing Gaussian Mixture Models with EM Using Equivalence Constraints (2003, background): “Most of the work in the field of semi-supervised learning focused on the c ase of partial labels augmenting a large unlabeled data set [4, 8, 5].”
A Generative Model for Semantic Role Labeling (2003, methodology): “overcome the limited amount of training data, we would ultim ately like to applybootstrapping, in which limited labeled data are combined with unlabeled d ata to produce a more accurate model than that trained on unlabeled data alon e [3, 4].”
Automatic bug triage using text categorization (2004, methodology): “[8] is presented in the following section, followed by the explanation of how we applied it in the bug triage domain.”