Subset Labeled LDA: A Topic Model for Extreme Multi-label Classification

Papanikolaou, Yannis; Tsoumakas, Grigorios

doi:10.1007/978-3-319-98539-8_12

Cited by 8 publications

(10 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A few Bayesian methods have also been proposed (He et al 2012;Kapoor et al 2012;Jain et al 2017;Gaure et al 2017;Papanikolaou and Tsoumakas 2018), however they cannot be trained on XML datasets and/or they attain poor performance comparing to stateof-the art methods. Our method bears similarities with most of those works.…”

Section: Related Workmentioning

confidence: 99%

Large scale multi-label learning using Gaussian processes

2021

View full text Add to dashboard Cite

We introduce a Gaussian process latent factor model for multi-label classification that can capture correlations among class labels by using a small set of latent Gaussian process functions. To address computational challenges, when the number of training instances is very large, we introduce several techniques based on variational sparse Gaussian process approximations and stochastic optimization. Specifically, we apply doubly stochastic variational inference that sub-samples data instances and classes which allows us to cope with Big Data. Furthermore, we show it is possible and beneficial to optimize over inducing points, using gradient-based methods, even in very high dimensional input spaces involving up to hundreds of thousands of dimensions. We demonstrate the usefulness of our approach on several real-world large-scale multi-label learning problems.

show abstract

Section: Related Workmentioning

confidence: 99%

Large scale multi-label learning using Gaussian processes

2021

View full text Add to dashboard Cite

show abstract

“…Recently, there has been interest in developing distributed linear methods [7,43] which can exploit distributed hardware. From a probabilistic view-point, bayesian approaches for multi-label classification have been developed in recent works such as [21,18] and Labeled LDA [30].…”

Section: Predictive Performance 5 Related Workmentioning

confidence: 99%

Data scarcity, robustness and extreme multi-label classification

Babbar

Schölkopf

2019

Mach Learn

View full text Add to dashboard Cite

The goal in extreme multi-label classification is to learn a classifier which can assign a small subset of relevant labels to an instance from an extremely large set of target labels. Datasets in extreme classification exhibit a long tail of labels which have small number of positive training instances. In this work, we pose the learning task in extreme classification with large number of tail-labels as learning in the presence of adversarial perturbations. This view motivates a robust optimization framework and equivalence to a corresponding regularized objective.Under the proposed robustness framework, we demonstrate efficacy of Hamming loss function for taillabel detection in extreme classification. The equivalent regularized objective, in combination with proximal gradient based optimization, performs better than state-ofthe-art methods on propensity scored versions of preci-sion@k and nDCG@k(upto 20% relative improvement over PFastreXML -a leading tree-based approach and 60% relative improvement over SLEEC -a leading labelembedding approach). Furthermore, we also highlight the sub-optimality of a sparse solver in a widely used package for large-scale linear classification, which is interesting in its own right. We also investigate the spectral properties of label graphs for providing novel insights towards understanding the conditions governing the performance of Hamming loss based one-vs-rest scheme vis-à-vis label embedding methods.

show abstract

“…This is a disadvantage because most news publications have document "tags" (or "labels", loosely speaking, not to be confused with dataset label) that works as a topic, for example, an article about particular flood can have "flood", "disaster", and some tags indicating city location as well (such as "Jakarta"). These tags are valuable and can be used as additional supervision for the LDA, providing a multi-label learning that is explored by many authors [20], [58], [59]. With the introduction of tags as label, the unsupervised nature of LDA becomes supervised in Labeled LDA.…”

Section: Increasing Model Generalizability With Topic Modelingmentioning

confidence: 99%

“…The objective for ATM is to provide topic modeling tool while also solves the memory requirement of LLDA when dealing with a very large number of tags, without sacrificing the coherence of the produced topic sets. LLDA posits a single topic-word distribution for each unique tag (label) that it found in the document, leading to a huge memory requirement for very large number (more than 10,000) of tags, on which case can be considered as an extreme multi-label classification problem [58].…”

Section: Analysis Of the Topic And Event Space: Tying Themes To Geospmentioning

confidence: 99%

Event Geoparser with Pseudo-Location Entity Identification and Numerical Extraction in Indonesian News Corpus

Dewandaru¹,

Widyantoro²,

Akbar³

2020

Preprint

View full text Add to dashboard Cite

One of the most important component of a Geographic Information Retrieval (GIR) is the geoparser, which performs toponym recognition, disambiguation, and geographic coordinate resolution from unstructured text domain. However, news articles which report several events across many place references mentioned in the document is not yet adequately modeled by regular geoparser types where the scope of resolution is either on toponym-level or document-level. The capacity to detect multiple events, geolocate its true locations and coordinates along with their numerical arguments are still missing from modern geoparsers, much less in Indonesian news corpora domain. We propose a novel type event geoparser which integrates an ACE-based event extraction model and provides precise event-level scope resolution. The geoparser casts the geotagging and event extraction as sequence labeling and uses Conditional Random Field with keywords feature obtained using Aggregated Topic Model as a semantic exploration from large corpus, which eventually increases the generalizability of the model. The geoparser also use Smallest Administrative Level feature along with Spatial Minimality-derived algorithm to improve the identification of Pseudo-location entities, resulting 19.4% increase for weighted F1 score. As a side effect of event extraction, the geoparser also extracts various numerical arguments and able to generate thematic choropleth map from a single news story.

show abstract

Subset Labeled LDA: A Topic Model for Extreme Multi-label Classification

Cited by 8 publications

References 14 publications

Large scale multi-label learning using Gaussian processes

Large scale multi-label learning using Gaussian processes

Data scarcity, robustness and extreme multi-label classification

Event Geoparser with Pseudo-Location Entity Identification and Numerical Extraction in Indonesian News Corpus

Contact Info

Product

Resources

About