Forums

Articles
Create
cancel
Showing results for 
Search instead for 
Did you mean: 

Does Confluence need a native model for measuring knowledge entropy rather than simply tracking stal

AGASTYA SHARMA
Rising Star
Rising Star
Rising Stars are recognized for providing high-quality answers to other users. Rising Stars receive a certificate of achievement and are on the path to becoming Community Champions.
July 12, 2026
Most organizations identify outdated documentation based on the last modified date.
 
However, documentation can become less trustworthy even when it is frequently edited due to contradictions, duplicated information, fragmented ownership, and inconsistent terminology.
 
Should Confluence move beyond "stale page detection" and instead introduce a measurable Knowledge Entropy Score that evaluates structural degradation across an entire knowledge base?
 
Has anyone developed similar governance models for large enterprise deployments?

2 answers

1 vote
Joshua Brock _ Seibert Group_ GmbH
Community Champion
August 14, 2026

Hi @AGASTYA SHARMA ... great question!

I think you’re identifying an important distinction: content freshness and knowledge quality aren’t the same thing.

A recently edited page can still conflict with another page, use outdated terminology, duplicate an existing source, or have no clear owner. Conversely, an older page may still be accurate and authoritative. A useful Knowledge Entropy Score would therefore need to combine several signals rather than simply penalizing age—for example:

  • conflicting or overlapping sources covering the same subject

  • missing ownership or review accountability

  • inconsistent terminology and classification

  • time since the content was last substantively reviewed

  • whether the current page still matches an approved version

The difficult part would be deciding how those signals should be weighted across different types of knowledge. A six-month-old policy and a six-month-old historical decision record shouldn’t necessarily receive the same treatment.

If you’re open to a Marketplace app, Aura Workflow & Approval can help with the governance and review portion of that model. It supports:

  • custom lifecycle states and scheduled transitions for content review

  • reviewer approvals and a history of workflow activity

  • notifications and actions when content reaches a review deadline

  • official versions that distinguish reviewed content from subsequent edits

It wouldn’t calculate the broader entropy score you’re proposing or automatically identify contradictions and duplicated knowledge. However, it can provide structured review, ownership, and approval signals that could form part of such a model.

I’d be interested to know whether you envision the score primarily as an administrative reporting tool or as something visible to readers and Rovo when selecting sources. Those could lead to quite different implementations.

And just in full disclosure, I work at Seibert Group, the team behind the Aura Apps.

Hope this helps and best of luck!

Joshua
Content Writer and US Representative
Aura Apps & Agile Hive (products of Seibert Group GmbH)

1 vote
Ajay _view26_
Community Champion
July 12, 2026

Suggest an answer

Log in or Sign up to answer
DEPLOYMENT TYPE
CLOUD
PRODUCT PLAN
FREE
TAGS
AUG Leaders

Atlassian Community Events