Converting Legacy Unstructured Content to DITA Using Claude AI
- Presentation
- AI and New Technologies in TC
- 02. April
- 15:00 - 15:40 PM (CEST)
- 13:00 - 13:40 PM (UTC) | 09:00 - 09:40 AM (EDT)
- Find your local time here: Time Converter
- finished
-
Keith Schengili-Roberts
- DitaWriter.com
Contents
Presenter Keith Schengili-Roberts talks about his experiences converting legacy, unstructured text (from a 1919 Model T car manual) using Claude AI to structured DITA content. While the process is not without its faults, AI's efficiency and utility at converting and transforming legacy content into structured DITA are undeniable. Keith explores the pros and cons of using AI to assist content conversion, examining the pitfalls and how they can best be avoided, and answers why you will want to move your unstructured content to a structured format for future AI-based applications. The presentation presents practical steps for converting unstructured content using an LLM, while also providing the rationale and ROI for converting legacy content to structured content.
Takeaways
Attendees will learn how to use Large Language Models to process unstructured legacy content to structured DITA XML, and how content benefits human audiences and AI-based applications like chatbots.
Prior knowledge
The audience ought to be familiar with the concept of structured content, and ideally with DITA tagging rules. AI-related concepts will be presented at a beginner level, as this field is still new for more technical writers.