AI & ML
Modeling Document and Section Hierarchies in Embedding Space
arXiv stat.ML0 views
arXiv:2611.21980v1 Announce Type: new Abstract: Embedded topic models integrate word embeddings into the topic generation process, producing coherent topics and robust performance on large vocabularies. However, each document is represented by a single topic distribution, ignoring its internal structure. Many real world collections, including legal documents, company and government reports, and online discussion forums, exhibit hierarchical organization in which documents consist of multiple sections that often address distinct subtopics. We introduce the Hierarchical Embedded Topic Model (hETM), a structured topic model that couples the topic distribution of a document and those of its sections through a hierarchical prior in the embedding space. We derive an amortized variational inference procedure for hETM and evaluate it on structured collections from each of these domains. Experimental results show that hETM achieves the lowest or near-lowest held-out perplexity among flat and hierarchical topic model baselines, along with comparable or higher topic quality. It also performs strongly on a downstream clustering task, achieving the best clustering performance on most of the collections.
Source:https://arxiv.org/abs/2611.21980PreviousMissing Security-Fix Backporting for GitHub Actions WorkflowsNextBandit Online Submodular Minimisation Using Zeroth-Order Oracles
Related posts
Advancing chip design with OpenAI models
0OpenAI Blog
Vulcan failure caused by inadequate fuel supply to engines, ULA says
0Ars Technica
Sam Altman warns against a future where AI, not humans, calls the shots
0TechCrunch
Kumo Tabular: How NVIDIA Built a Foundation Model That Predicts on Any Table Without Training
0dev.to

Asus TUF Gaming F16 (2026), A15 & T700 Mini: Specs, Pricing & The ₹25,000 GPU Catch
0dev.to
I Stopped Chasing Every New AI Tool — And My Blog Finally Started Growing
0dev.to