Featured Project
Research in Motion
Ferguson Twitter Networks
Twitter user interaction networks from the Ferguson protests and ensuing events. From archive.org archived tweets.
Northwestern Libraries · Collaborative language research
Model of Models
A free research platform for text mining and topic modeling across millions of scholarly documents.
Explore language patterns in PubMed, JSTOR, HathiTrust, social media archives, and public datasets through machine learning, BERT projections, network analysis, and interactive visualizations.
Initially designed and developed at the University of Cincinnati with funding from the Mellon Foundation, MoM is now maintained at Northwestern University Libraries.
Research at corpus scale
Topic modeling
Build LDA, pyLDAvis, and multilevel topic models from large scholarly and public text collections.
BERT projection
Explore document embeddings, semantic clusters, and timeline views for curated research datasets.
Social network analysis
Study public discourse through Twitter/X interaction networks, timelines, and named-entity views.
Public research maps
Connect data to place through interactive projects such as CO2 emissions, race, income, and metropolitan change.
Frequently asked questions
What is Model of Models?
Model of Models is a free research platform from Northwestern University Libraries for large-scale text mining, topic modeling, network analysis, and machine-learning exploration of public and scholarly text corpora. Researchers submit a query, the platform trains an interactive machine-learning visualization (topic model, embedding space, or network graph) against a chosen corpus, and returns the result in the browser. It was previously known as Learning Machines.
Is Model of Models free to use?
Yes. The platform is free, ad-free, and publicly accessible. Browsing the pre-built research projects requires no account; a free account is needed only to launch new model runs.
Who maintains Model of Models?
Model of Models was initially designed and developed at the University of Cincinnati with funding from the Mellon Foundation. It is now maintained by Northwestern University Libraries.
What corpora can I analyze?
The platform indexes research corpora in AWS OpenSearch, including PubMed and medical-literature abstracts, JSTOR, news archives, curated Twitter/X network datasets, and literary and historical text collections. A few corpora are restricted to approved research teams. There is no user-uploaded data.
What machine-learning methods does Model of Models support?
The platform supports LDA and Multilevel LDA topic modeling (gensim), Word2Vec and Doc2Vec embeddings, and BERTopic with sentence-transformer embeddings plus UMAP and HDBSCAN clustering on NVIDIA GPUs. Results render as interactive pyLDAvis visualizations, scatter plots, and network graphs.
Do I need to install anything?
No. Model of Models runs entirely in a modern web browser; there is nothing to download or install.