Statistical topic models for multi-label document classification

DOI10.1007/S10994-011-5272-5zbMATH Open1243.68248arXiv1107.2462OpenAlexW2074909580MaRDI QIDQ439042FDOQ439042

America Chambers, Mark Steyvers, Timothy N. Rubin, Padhraic Smyth

Publication date: 31 July 2012

Published in: Machine Learning (Search for Journal in Brave)

Abstract: Machine learning approaches to multi-label document classification have to date largely relied on discriminative modeling techniques such as support vector machines. A drawback of these approaches is that performance rapidly drops off as the total number of labels and the number of labels per document increase. This problem is amplified when the label frequencies exhibit the type of highly skewed distributions that are often observed in real-world datasets. In this paper we investigate a class of generative statistical topic models for multi-label documents that associate individual word tokens with different labels. We investigate the advantages of this approach relative to discriminative models, particularly with respect to classification problems involving large numbers of relatively rare labels. We compare the performance of generative and discriminative approaches on document labeling tasks ranging from datasets with several thousand labels to datasets with tens of labels. The experimental results indicate that probabilistic generative models can achieve competitive multi-label classification performance compared to discriminative methods, and have advantages for datasets with many labels and skewed label frequencies.

Full work available at URL: https://arxiv.org/abs/1107.2462

zbMATH Keywords

graphical modelsmulti-label classificationLDAtext classificationtopic modelsdependency-LDAdocument modelingprobabilistic generative models

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Pattern recognition, speech recognition (68T10)

Cites Work

• 10.1162/jmlr.2003.3.4-5.993
• Hierarchical Dirichlet Processes
• 10.1162/153244303322753670
• Reducing multiclass to binary: A unifying approach for margin classifiers
• The nested chinese restaurant process and bayesian nonparametric inference of topic hierarchies
• Feature selection for multi-label naive Bayes classification
• Multilabel classification via calibrated label ranking
• 10.1162/153244303322533188
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)

Cited In (10)

• Two-way Poisson mixture models for simultaneous document classification and word clustering
• A directed topic model applied to call center improvement
• Online multi-label dependency topic models for text classification
• A Bayesian nonparametric model for multi-label learning
• Additive regularization of topic models
• Additive regularization for topic models of text collections
• Joint Concept Correlation and Feature-Concept Relevance Learning for Multilabel Classification
• Robust supervised topic models under label noise
• DOLDA: a regularized supervised topic model for high-dimensional multi-class regression
• Bayesian sparse joint dynamic topic model with flexible lead-lag order

Uses Software

• LIBLINEAR
• RCV1
• MULAN
• Medlda
• OHSUMED

This page was built for publication: Statistical topic models for multi-label document classification

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q439042)