Sentiment and emotional analysis detect positivity, negativity, and other signals such as anger, and desire across communications, providing law firms and organisations with contextual visibility into the expression of emotions buried deep in data. 

One of the key challenges in using sentiment analysis is the potential for harmful biases in the underlying models. If a sentiment analysis model is capable of discriminating bias, it may carry that bias over to its conclusions.

As AI usage exponentially expands, managing the risk of bias in models is becoming increasingly urgent. Especially in a legal setting, we all have a part to play in actively measuring and managing bias so we can utilise AI that is ethical, trusted, and morally principled.

In partnership with the ACi, Relativity’s feature webinar A Principled Approach to using Sentiment Analysis in Investigations, led by Aaron Ahmadia, Senior Director Applied Sciences, Relativity, will join legal luminaries in exploring how to prevent biases permeating throughout investigative datasets.

Key discussion points include:

• Why the European Union has drawn up a new legal framework to regulate the development and use of Artificial Intelligence
• The biggest challenges and advantages of using sentiment analysis in investigations
• How legal teams can mitigate the risks of biased data and break systemic prejudices
• How investigators and legal teams use advanced AI effectively whilst acting responsibly and within the parameters of regulatory guidelines