NatCat: Weakly Supervised Text Classification with Naturally Annotated Resources

We seek to improve text classification by leveraging naturally annotated data. In particular, we construct a general purpose text categorization dataset (NatCat) from three online resources: Wikipedia, Reddit, and Stack Exchange. These datasets consist of document-category pairs derived from manual curation that occurs naturally by their communities. We build general purpose text classifiers by training on NatCat and evaluate them on a suite of 11 text classification tasks (CatEval). We benchmark different modeling choices and dataset combinations, and show how each task benefits from different NatCat training resources.

cs.CL Topic Modeling Text and Document Classification Technologies Natural Language Processing Techniques Computer Science

View graph Source

Paper

References (52)

01TaxoClass: Hierarchical Multi-Label Text Classification Using Only Class NamesJiaming Shen, Wenda Qiu, Yu Meng et al.2021 · North American Chapter of the Association for Computational Linguistics · 84 citations

02Unsupervised Label Refinement Improves Dataless Text ClassificationZewei Chu, K. Stratos, Kevin Gimpel2020 · Findings · 19 citations In Library

03Hierarchical Metadata-Aware Document Categorization under Weak SupervisionYu Zhang, Xiusi Chen, Yu Meng et al.2020 · Web Search and Data Mining · 30 citations In Library

04Zero-shot Text Classification With Generative Language ModelsRaul Puri, Bryan Catanzaro2019 · arXiv.org · 121 citations In Library

05How to Ask Better Questions? A Large-Scale Multi-Domain Dataset for Rewriting Ill-Formed QuestionsZewei Chu, Mingda Chen, Jiehua Chen et al.2019 · AAAI Conference on Artificial Intelligence · 22 citations In Library

06Hierarchical Transformers for Long Document ClassificationR. Pappagari, Piotr Żelasko, J. Villalba et al.2019 · Automatic Speech Recognition & Understanding · 292 citations In Library

07HuggingFace's Transformers: State-of-the-art Natural Language ProcessingThomas Wolf, Lysandre Debut, Victor Sanh et al.2019 · arXiv.org · 4.2k citations In Library

08ALBERT: A Lite BERT for Self-supervised Learning of Language RepresentationsZhenzhong Lan, Mingda Chen, Sebastian Goodman et al.2019 · International Conference on Learning Representations · 7.6k citations In Library

09Benchmarking Zero-shot Text Classification: Datasets, Evaluation and Entailment ApproachWenpeng Yin, Jamaal Hay, Dan Roth2019 · Conference on Empirical Methods in Natural Language Processing · 692 citations In Library

10StructBERT: Incorporating Language Structures into Pre-training for Deep Language UnderstandingWei Wang, Bin Bi, Ming Yan et al.2019 · International Conference on Learning Representations · 304 citations In Library

11RoBERTa: A Robustly Optimized BERT Pretraining ApproachYinhan Liu, Myle Ott, Naman Goyal et al.2019 · arXiv.org · 31k citations In Library

12Integrating Semantic Knowledge to Tackle Zero-shot Text ClassificationJingqing Zhang, Piyawat Lertvittayakumjorn, Yike Guo2019 · North American Chapter of the Association for Computational Linguistics · 123 citations In Library

Scroll for more · 38 remaining

Similar papers

NATCAT: Weakly Supervised Text Classification with Naturally Annotated ResourcesZewei Chu, Karl Stratos, Kevin Gimpel2021 · Automated Knowledge Base Construction In Library

Minimally Supervised Categorization of Text with MetadataYu Zhang, Meng Yu, Jiaxin Huang et al.2020 · arXiv (Cornell University) · 26 citations In Library

Minimally Supervised Categorization of Text with MetadataYu Zhang, Meng Yu, Jiaxin Huang et al.2020 In Library

Minimally-Supervised Structure-Rich Text Categorization via Learning on Text-Rich NetworksXinyang Zhang, Chenwei Zhang, Luna Dong et al.2021 · arXiv (Cornell University) · 21 citations In Library

Minimally-Supervised Structure-Rich Text Categorization via Learning on Text-Rich NetworksXinyang Zhang, Chenwei Zhang, Xin Dong et al.2021 In Library

An Efficient Nlp Framework for Text Categorization Using Tf-Idf and EcocPriyanshu Birkhani, Riya Garjola, Mehul Manu2025 In Library

Enhancing Robust Text Classification via Category DescriptionXin Gao, Zhengye Zhu, Xu Chu et al.2022 · 2022 IEEE International Conference on Data Mining (ICDM) · 4 citations In Library

Automatic Text ClassificationAndrii Dubovyk, Yevhenii Volynets2025 · NaUKMA Research Papers Computer Science · 0 citations In Library

Text Categorization using Supervised Machine Learning TechniquesIshaan Dawar, Narendra Kumar, Sakshi Negi et al.2023 In Library

Weakly-Supervised Neural Text ClassificationYu Meng, Jiaming Shen, Chao Zhang et al.2018 · International Conference on Information and Knowledge Management · 200 citations In Library

Source of record

arXiv

44B

Welcome.

This panel sticks with you. Pick anything below and it opens right beside it, so you can dig through 60,000-plus records without ever losing your spot here.

Travel 44B

LibraryPapers, policy, standards, statute LabsEvery organization building AI ModelsIntelligence, price, and speed BenchmarksThe evaluation catalog IncidentsAI failures and harms SearchOne field across everything DashboardThe state of AI accountability in NY 44B RegistryThe Art. 44-B compliance portal