BarbriSFCourseDetails

Course Details

This CLE webinar will assist class action lawyers and other litigators in understanding sentiment analysis, its value in litigation, and how the latest advances in machine learning and AI can overcome challenges presented by more traditional analytical approaches. The program will discuss what sentiment or emotional analysis is and how it works, the various approaches used to obtain this data, the categories of cases in which it may be the most useful, and evidentiary challenges, as well as share success stories.

Faculty

Description

To urge or oppose certification or prove damages, class action lawyers on both sides routinely need to pinpoint and quantify changes (or lack of) in attitude or conduct, i.e., "sentiment," towards a particular subject (person, product, company, geography, value, etc.) before and after some pivotal event. That event might be a marketing campaign, a label, an allegedly defamatory statement, an environmental accident, or market information.

Machine learning and other variations of AI have the potential to overcome many of the challenges presented by traditional sentiment analysis approaches because they offer highly efficient conceptual analysis with limited influence from human subjectivity or strict and incomplete rules. This same type of AI can also provide "sentiment analysis" results using real time facial or vocal expressions and with or without the subject's knowledge or consent, creating significant privacy and "intimate" knowledge concerns.

Listen as our expert panel offers litigation insights about the potential uses of sentiment analysis and explores some of the ethical minefields to be avoided.

Outline

  1. Overview of sentiment analysis and its uses in litigation
  2. Sentiment analysis approaches
  3. Applying machine learning or other AI to conduct sentiment analysis
  4. Evidentiary issues

Benefits

The panel will discuss these and other key issues:

  • What are the various approaches or forms of sentiment analysis?
  • How can machine learning overcome some of the challenges to various forms of sentiment analysis?
  • What are the evidentiary challenges to sentiment analysis derived from machine learning?