CobwebTM: Probabilistic Concept Formation for Lifelong and Hierarchical Topic Modeling
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908973106462720 |
|---|---|
| author | Singaravadivelan, Karthik Gupta, Anant Wang, Zekun MacLellan, Christopher J. |
| author_facet | Singaravadivelan, Karthik Gupta, Anant Wang, Zekun MacLellan, Christopher J. |
| contents | Topic modeling seeks to uncover latent semantic structure in text corpora with minimal supervision. Neural approaches achieve strong performance but require extensive tuning and struggle with lifelong learning due to catastrophic forgetting and fixed capacity, while classical probabilistic models lack flexibility and adaptability to streaming data. We introduce CobwebTM, a low-parameter lifelong hierarchical topic model based on incremental probabilistic concept formation. By adapting the Cobweb algorithm to continuous document embeddings, CobwebTM constructs semantic hierarchies online, enabling unsupervised topic discovery, dynamic topic creation, and hierarchical organization without predefining the number of topics. Across diverse datasets, CobwebTM achieves strong topic coherence, stable topics over time, and high-quality hierarchies, demonstrating that incremental symbolic concept formation combined with pretrained representations is an efficient approach to topic modeling. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_14489 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | CobwebTM: Probabilistic Concept Formation for Lifelong and Hierarchical Topic Modeling Singaravadivelan, Karthik Gupta, Anant Wang, Zekun MacLellan, Christopher J. Computation and Language Topic modeling seeks to uncover latent semantic structure in text corpora with minimal supervision. Neural approaches achieve strong performance but require extensive tuning and struggle with lifelong learning due to catastrophic forgetting and fixed capacity, while classical probabilistic models lack flexibility and adaptability to streaming data. We introduce CobwebTM, a low-parameter lifelong hierarchical topic model based on incremental probabilistic concept formation. By adapting the Cobweb algorithm to continuous document embeddings, CobwebTM constructs semantic hierarchies online, enabling unsupervised topic discovery, dynamic topic creation, and hierarchical organization without predefining the number of topics. Across diverse datasets, CobwebTM achieves strong topic coherence, stable topics over time, and high-quality hierarchies, demonstrating that incremental symbolic concept formation combined with pretrained representations is an efficient approach to topic modeling. |
| title | CobwebTM: Probabilistic Concept Formation for Lifelong and Hierarchical Topic Modeling |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2604.14489 |