Discriminative Topic Mining via Category-Name Guided Text Embedding

Abstract

Mining a set of meaningful and distinctive topics automatically from massive text corpora has broad applications. Existing topic models, however, typically work in a purely unsupervised way, which often generate topics that do not fit users' particular needs and yield suboptimal performance on downstream tasks. We propose a new task, discriminative topic mining, which leverages a set of user-provided category names to mine discriminative topics from text corpora. This new task not only helps a user understand clearly and distinctively the topics he/she is most interested in, but also benefits directly keyword-driven classification tasks. We develop CatE, a novel category-name guided text embedding method for discriminative topic mining, which effectively leverages minimal user guidance to learn a discriminative embedding space and discover category representative terms in an iterative manner. We conduct a comprehensive set of experiments to show that CatE mines high-quality set of topics guided by category names only, and benefits a variety of downstream applications including weakly-supervised classification and lexical entailment direction identification.

arXiv commentWWW 2020. (Code: https://github.com/yumeng5/CatE)

References in corpus (13)

• Supervised Topic Models
• PTE: Predictive Text Embedding through Large-scale Heterogeneous Text Networks
• A Convolutional Neural Network for Modelling Sentences
• Weakly-Supervised Neural Text Classification
• Poincaré Embeddings for Learning Hierarchical Representations
• Poincaré GloVe: Hyperbolic Word Embeddings
• Ultradense Word Embeddings by Orthogonal Transformation
• Learning Continuous Hierarchies in the Lorentz Model of Hyperbolic Geometry
• Spherical Text Embedding
• Hyperbolic Entailment Cones for Learning Hierarchical Embeddings
• Distilled Wasserstein Learning for Word Embedding and Topic Modeling
• Guiding Corpus-based Set Expansion by Auxiliary Sets Generation and Co-Expansion
• Scalable Topical Phrase Mining from Text Corpora

Cited by in corpus (15)

• Hierarchical Topic Mining via Joint Spherical Tree and Text Embedding
• Topic Discovery via Latent Space Clustering of Pretrained Language Model Representations
• Comprehensive Named Entity Recognition on CORD-19 with Distant or Weak Supervision
• CoRel: Seed-Guided Topical Taxonomy Construction by Concept Learning and Relation Transferring
• Keyword Assisted Embedded Topic Model
• Guiding Corpus-based Set Expansion by Auxiliary Sets Generation and Co-Expansion
• TaxoCom: Topic Taxonomy Completion with Hierarchical Discovery of Novel Topic Clusters
• A Survey on Retrieval And Structuring Augmented Generation with Large Language Models
• vONTSS: vMF based semi-supervised neural topic modeling with optimal transport
• Text Classification Using Label Names Only: A Language Model Self-Training Approach
• X-Class: Text Classification with Extremely Weak Supervision
• WOT-Class: Weakly Supervised Open-world Text Classification
• Weakly-Supervised Aspect-Based Sentiment Analysis via Joint Aspect-Sentiment Topic Embedding
• Minimally Supervised Categorization of Text with Metadata
• Empirical Study of Named Entity Recognition Performance Using Distribution-aware Word Embedding