# Network Backboning with Noisy Data

arXiv: https://arxiv.org/abs/1701.07336 · HTML: https://mlsanity.com/paper/1701.07336/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 25 Jan 2017.

## Built on by (most cited first, 11 of 11)

- [Homophily-enhanced Structure Learning for Graph Clustering](https://mlsanity.com/paper/2308.05309/) (2023)
- [Sentiment and structure in word co-occurrence networks on Twitter](https://mlsanity.com/paper/2110.00587/) (2021)
- [Spectral Neural Graph Sparsification](https://mlsanity.com/paper/2510.27474/) (2025)
- [Support Thresholds, Not Algorithms, Limit Rare-Association Recovery in Co-Purchase Networks](https://mlsanity.com/paper/2609.20171/) (2026)
- [Sparsistent filtering of comovement networks from high-dimensional data](https://mlsanity.com/paper/2101.09174/) (2021)
- [The Healthy States of America: Creating a Health Taxonomy with Social Media](https://mlsanity.com/paper/2103.01169/) (2021)
- [A Hierarchical Network-Oriented Analysis of User Participation in Misinformation Spread on WhatsApp](https://mlsanity.com/paper/2109.10462/) (2021)
- [Reconstructing signed relations from interaction data](https://mlsanity.com/paper/2209.03219/) (2022)
- [Dream Content Discovery from Reddit with an Unsupervised Mixed-Method Approach](https://mlsanity.com/paper/2307.04167/) (2023)
- [Unsupervised Learning via Network-Aware Embeddings](https://mlsanity.com/paper/2309.10408/) (2023)
- [Evaluating GNNs for Success Prediction in Artist Collaboration Networks](https://mlsanity.com/paper/2609.02920/) (2026)

## Abstract (arXiv)

Networks are powerful instruments to study complex phenomena, but they become hard to analyze in data that contain noise. Network backbones provide a tool to extract the latent structure from noisy networks by pruning non-salient edges. We describe a new approach to extract such backbones. We assume that edge weights are drawn from a binomial distribution, and estimate the error-variance in edge weights using a Bayesian framework. Our approach uses a more realistic null model for the edge weight creation process than prior work. In particular, it simultaneously considers the propensity of nodes to send and receive connections, whereas previous approaches only considered nodes as emitters of edges. We test our model with real world networks of different types (flows, stocks, co-occurrences, directed, undirected) and show that our Noise-Corrected approach returns backbones that outperform other approaches on a number of criteria. Our approach is scalable, able to deal with networks with millions of edges.