Story Clustering

When multiple feeds cover the same story, NewsBlur groups them together so you only see it once.

Story titles view showing clustered stories with sources listed underneath

Clustered stories in the titles view

Feed options popover showing clustering controls

Clustering controls in the feed options popover

Story detail view showing cluster cards for alternative sources

Rich cards for alternative sources in the detail view

How it works

When a feed updates, NewsBlur checks for duplicate stories across all your subscriptions
Title matching catches exact and fuzzy duplicates using significant-word overlap
Semantic matching uses Elasticsearch to find stories about the same topic with different headlines
Duplicates fold underneath the highest-scoring version in your river view
Click any source to read that version instead, or mark all duplicates as read at once

How does clustering detect duplicate stories?

Two layers work together. Title matching normalizes headlines and groups exact and fuzzy matches. Semantic matching sends titles to Elasticsearch's more_like_this query to catch stories about the same event written with completely different headlines.

Can I automatically mark duplicates as read?

Yes. Enable "Mark all as read" in the feed options popover under Story Clustering, or in Manage > Preferences > Stories. When you read the representative story, all other stories in the cluster are marked as read.

Can I turn clustering off?

Yes. Toggle "Keep stories separate" in the feed options popover or in Preferences. Every story will appear individually as before.

What plan do I need?

All users see clustered stories on popular feeds. Premium Archive subscribers ($99/year) get full control: toggle clustering on or off, choose display styles, and automatically mark duplicates as read.