跳至主要內容
前往資訊主頁
不確定該從哪裡著手嗎?歡迎進行簡短測驗,以便取得個人化建議。
Inquirer

Inquirer turbo-charges tagging with AI

Major newsroom saves hours and standardizes article tags with Gemini-powered tool.

The Challenge

Inquirer's manual article-tagging process was inconsistent and inefficient. With over 300 different election-related tags in their database—many of them duplicates or slight variations—it was difficult to track coverage and analyze content performance accurately. It took the team 32 minutes to recategorize 100 articles, creating a significant workflow bottleneck, particularly for important and extensive election coverage.

The Results

Using Gemini, the Inquirer team analyzed their list of duplicated and wrong tags, and created a standardized "holy list" of just 19 categories for election coverage. They then developed an AI-powered tool that could automatically categorize 1,500 articles in just 30 minutes—a task that would have taken over 12 hours to complete manually. The AI tagger achieved 90% accuracy, far surpassing the consistency of the previous human-led process. This tool promises to make both the data and editorial teams more efficient, allowing for a more robust and streamlined analysis of all content.

  • 90% accuracy of AI-tagged articles vs. 46% for human-tagged articles
  • 300+ inconsistent election tags reduced to one standardized list of 19
  • 12+ hours cut to just 30 minutes for 1,500 articles
Copy of Monica Baylon-Solideo headshot
“We started with one specific AI tool that was non-intrusive to the editorial team. But now they have seen the benefit of the tagger, they are open to other AI tools in the newsroom as long as they are used in a safe and thoughtful way."
Monica Baylon-Solideo
Data Analytics Manager, Inquirer
Copy of monica-ralph-presenting
Monica Baylon-Solideo and Ralph Gurango present their findings at the GNI AI Workshops graduation event.
要離開並失去進度嗎?
如果離開這個頁面,你將失去當前所完成的所有課程進度。確定要繼續並失去已完成的課程進度嗎?