Narrative Map FAQ

Are the Narrative Map tags the same as my project tags?

No. Your project tags are set up when a project is activated and are used to build your reports and dashboards. The Narrative Map instead shows tags discovered by Pipeline AI across your subscribed coverage — the themes, topics, products, organisations and spokespeople it finds in the articles. It’s a separate, exploratory view.

Are the tags generated by AI, or can I add my own?

Every tag on the Narrative Map is generated by AI (a smart tag), across all of your coverage — they aren’t added by hand. Manually added tags belong to a specific project or organisation, whereas the Narrative Map works across your whole coverage universe.

Can I line up coded coverage, such as press releases, with the AI tags?

Yes. Restrict the map to the time period of your coded coverage and apply the coded-coverage filter, so the data is narrowed to what’s relevant. You’ll then see the AI tags that come through for that slice — a useful way to see the themes emerging around a particular PR or campaign.

Why does the same name appear under more than one category?

The AI decides how to classify each mention from the context it sees in the coverage, using a general model that works across all clients rather than rules written for one brand. Occasionally that means a name is categorised in more than one way — for example a product also appearing as an organisation, or a brand showing up as a topic. If you spot a classification that looks wrong, flag it to your CARMA representative so it can be tightened.

Why is the number of mentions higher than the number of articles?

A single article can carry several tags, so you’ll almost always see more mentions than articles. The mentions figure counts how often tags occur, not how many articles there are.

Why don’t I see every tag on the map?

Your coverage contains far more tags than can be shown at once, so the map samples the top 100. Narrowing your filters is the best way to bring the tags you care about into view.

What do the colours mean?

In cluster view, each colour represents a category that Pipeline AI groups the content into — for example people, products and themes. See Choose a visualisation in the Narrative Map.

What’s the difference between Auto and Noor AI clustering?

Both are AI. Auto groups the tags themselves — it looks at all the tags on the map and groups the ones that belong together, using any brief you’ve provided. Noor AI groups by reading the coverage — it samples the top articles behind the tags, works out the themes from their content, then sorts the tags into those clusters. There’s no hard rule; choose whichever suits what you’re exploring. Full detail is in Reshape the map with AI clustering.

Does creating an AI cluster create new tags?

Yes — when you create a new AI cluster, it generates smart tags.

Can I use my existing filters to narrow what the map explores?

Yes. Any filter in the filter menu — date range, tier, media type, coded coverage, and smart filters by tag type — narrows the coverage first. The map then shows the AI and smart tags that come through from that narrowed set. See How to use filters.

Is the Narrative Map a replacement for search?

No. Use search when you already know what you’re looking for. Use the Narrative Map when you want to explore and understand what’s in your coverage — emerging themes, the narratives around a topic, and how conversations connect.

Need more help? If your question isn’t covered here, contact your CARMA representative or support team, and include the project and what you were trying to explore.