Lune

ACL Findings2021

Semi-Supervised Data Programming with Subset Selection

Ayush Maheshwari, Oishik Chatterjee, KrishnaTeja Killamsetty, Ganesh Ramakrishnan, Rishabh K. Iyer

Abstract

The paradigm of data programming, which uses weak supervision in the form of rules/labelling functions, and semi-supervised learning, which augments small amounts of labelled data with a large unlabelled dataset, have shown great promise in several text classification scenarios. In this work, we argue that by not using any labelled data, data programming based approaches can yield sub-optimal performances, particularly when the labelling functions are noisy. The first contribution of this work is an introduction of a framework, SPEAR which is a semi-supervised data programming paradigm that learns a joint model that effectively uses the rules/labelling functions along with semi-supervised loss functions on the feature space. Next, we also study SPEAR-SS which additionally does subset selection on top of the joint semi-supervised data programming objective and selects a set of examples that can be used as the labelled set by SPEAR. The goal of SPEAR-SS is to ensure that the labelled data can complement the labelling functions, thereby benefiting from both data-programming as well as appropriately selected data for human labelling. We demonstrate that by effectively combining semi-supervision, data-programming, and subset selection paradigms, we significantly outperform the current state-of-the-art on seven publicly available datasets. 1

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

5 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Fusing Conditional Submodular GAN and Programmatic Weak Supervision cites Robust Weak Supervision with Variational Auto-Encoders
  • Fusing Conditional Submodular GAN and Programmatic Weak Supervision cites Learning from Rules Generalizing Labeled Exemplars
  • Robust Weak Supervision with Variational Auto-Encoders cites Learning from Rules Generalizing Labeled Exemplars
  • Bayesian Coreset Optimization for Personalized Federated Learning cites Coresets for Data-efficient Training of Machine Learning Models
  • GLISTER: Generalization based Data Subset Selection for Efficient and Robust Learning cites Coresets for Data-efficient Training of Machine Learning Models
  • Adaptive Mixing of Auxiliary Losses in Supervised Learning cites Robust Data Programming with Precision-guided Labeling Functions
  • Adaptive Mixing of Auxiliary Losses in Supervised Learning cites Coresets for Data-efficient Training of Machine Learning Models
  • Adaptive Mixing of Auxiliary Losses in Supervised Learning cites Learning from Rules Generalizing Labeled Exemplars
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers5

  1. Training Subset Selection for Weak SupervisionHunter Lang, Aravindan Vijayaraghavan, David A. SontagNeurIPS 2022 · 27 citations
  2. Robust Weak Supervision with Variational Auto-EncodersFrancesco Tonolini, Nikolaos Aletras, Yunlong Jiao, Gabriella KazaiICML 2023 · 7 citations
  3. Adaptive Mixing of Auxiliary Losses in Supervised LearningDurga Sivasubramanian, Ayush Maheshwari, Prathosh AP, Pradeep Shenoy et al.AAAI 2023 · 7 citations
  4. Fusing Conditional Submodular GAN and Programmatic Weak SupervisionKumar Shubham, Pranav Sastry, Prathosh APAAAI 2024 · 3 citations
  5. Bayesian Coreset Optimization for Personalized Federated LearningPrateek Chanda, Shrey Modi, Ganesh RamakrishnanICLR 2024 · 2 citations

Builds on4

  1. Coresets for Data-efficient Training of Machine Learning ModelsBaharan Mirzasoleiman, Jeff A. Bilmes, Jure LeskovecICML 2020 · 494 citations
  2. GLISTER: Generalization based Data Subset Selection for Efficient and Robust LearningKrishnaTeja Killamsetty, Durga Sivasubramanian, Ganesh Ramakrishnan, Rishabh K. IyerAAAI 2021 · 300 citations
  3. Learning from Rules Generalizing Labeled ExemplarsAbhijeet Awasthi, Sabyasachi Ghosh, Rasna Goyal, Sunita SarawagiICLR 2020 · 93 citations
  4. Robust Data Programming with Precision-guided Labeling FunctionsOishik Chatterjee, Ganesh Ramakrishnan, Sunita SarawagiAAAI 2020 · 20 citations

Related papers

  1. Learning to Robustly Aggregate Labeling Functions for Semi-supervised Data ProgrammingAyush Maheshwari, KrishnaTeja Killamsetty, Ganesh Ramakrishnan, Rishabh K. Iyer et al.ACL Findings 2022 · 13 citations
  2. Witan: Unsupervised Labelling Function Generation for Assisted Data ProgrammingBenjamin Denham, Edmund M.-K. Lai, Roopak Sinha, Muhammad Asif NaeemVLDB 2022 · 12 citations
  3. SepLL: Separating Latent Class Labels from Weak Supervision NoiseAndreas Stephan, Vasiliki Kougia, Benjamin RothEMNLP Findings 2022 · 7 citations
  4. Adaptive Ranking-based Sample Selection for Weakly Supervised Class-imbalanced Text ClassificationLinxin Song, Jieyu Zhang, Tianxiang Yang, Masayuki GotoEMNLP Findings 2022 · 3 citations
  5. Inspector Gadget: A Data Programming-based Labeling System for Industrial ImagesGeon Heo, Yuji Roh, Seonghyeon Hwang, Dayun Lee et al.VLDB 2021 · 9 citations
Dusk over the sea between two cliffs drawn in fine vertical lines