Aidan Randle-Conde of Hanzo has started a new blog series on navigating AI success metrics. Part I, on precision and recall in email and document analysis, is out!

Aidan’s post (Part I: Navigating AI Success Metrics – Precision and Recall in Email and Document Analysis, available here) dives into the world of email datasets, exploring how precision and recall can be applied to document-based datasets to evaluate the effectiveness of AI in handling email communications.

Aidan defines recall, rejection, and precision and illustrates the concepts with discussion and a couple of graphics. He also discusses the challenges associated with human review, including level of effort and “the interpersonal and intrapersonal variability humans exhibit when making decisions.” While the terms and concepts may be elementary to some of you, they are necessary to evaluate the effectiveness of any approach to classifying documents, including AI-based approaches.

Aidan also briefly discusses the potential of Large Language Models (LLMs) and the potential future role of traditional approaches like Technology-Assisted Review (TAR) and Continuous Active Learning (CAL).

Read Aidan’s post here to learn more about precision and recall in email and document analysis! It’s just one click! If you read it, the precision and recall will both be 100%! 😀

You can also check out this post which previews the entire navigating AI success metrics series.

So, what do you think? How do you evaluate the effectiveness of AI in document analysis? Please share any comments you might have or if you’d like to know more about a particular topic.

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by my employer, my partners or my clients. eDiscovery Today is made available solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscovery Today should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.