How to classify news items by their actual topic
Topic classification helps organize a selected news feed for a digest or research folder. Consistent categories make the collection easier to browse while retaining the original references.
Start with the right material.
- What you provide
- a visible news list plus the user's category labels.
- The intended result
- Classify displayed items with evidence and flag cases needing full-article context.
How the workflow works.
Use the supplied title and available text to choose a supported topic. Allow multiple topics or an uncertain label when your classification scheme requires that distinction.
What to check in the result.
Inspect ambiguous items and avoid drawing a detailed conclusion from a headline alone. The classification describes the selected content and does not establish the truth or completeness of the reporting.
Import the CSV with a consistent delimiter and inspect text identifiers, dates and amounts before using spreadsheet totals. Keep the original file available while checking any automatic type conversion.
Learn from the worked example.
The downloadable CSV example uses these columns: Headline, Category, Evidence. Compare a complete record with the source before sorting, combining or importing your own output.
The supplied example is illustrative. It is not a recorded run on your data.
Download the example ↓Use it with your own material.
- Open the source you want to work with and activate WebAct on that tab.
- Copy the task from the hero, add any required fields or preferences, and review the proposed actions.
- Compare the first result with your source. Ask for a specific correction or refinement where needed.