patrickphat / awesome-topic-modeling

All methods, libraries and resources for exploring topics from large datasets

Awesome Topic Modeling

All methods, libraries and resources for exploring topics from large datasets ✨

📚 Introduction

Topic Models are a collection of machine learning models that explores topics in a large set of documents

🏛️ Classical Topic Models

This part includes classical topic models, which heavily based on statistical models and com

• Latent Dirichlet Allocation (LDA) [Paper] [Video] [Article]
• Latent Semantic Analysis (LSA) [Article]
• Non-negative Matrix Factorization (NMF) [Article]

🤖 Neural Topic Models

This part revolves around topic modelling techniques with the adoption of deep learning models.

• Pre-training is a Hot Topic: Contextualized Document Embeddings Improve Topic Coherence [Paper] [Github]
• Topic Modeling with Contextualized Word Representation Clusters [Paper]
• TopicBERT: Topic-aware BERT for Efficient Document Classification [Paper] [Github]
• BERTopic [Paper] [Github]

🌱 Semi-supervised topic models (SSTM)

SSTM allows users to inject prior knowledge about topics into topic models

• SeededLDA [Paper] [Github]
• Anchored CorEX [Paper] [Github]

⚡ Dynamic Topic Models (DTM)

DTMs are set of topic models that take into account the evolution of topic through time

• Dynamic Topic Model, David Blei [Paper] [Github]
• Dynamic Non-negative Matrix Factorization (Dynamic NMF) [Paper] [Github]

Topic Model Libraries

• OCTIS: Comparing Topic Models is Simple! A python package to optimize and evaluate topic models (accepted at EACL2021 demo track) [Github]
• Topic modeling using Gensim [Article]

Surveys:

• Topic Modelling Meets Deep Neural Networks: A Survey [Paper]