AI for Personalization & Recommendation

Overview

Content Personalization and Recommendation are essential for Public Service Media to stay relevant, cut through the digital noise and to ensure that the quality content they produce reaches an audience as broad as possible online.

The two sessions that compose this course have been designed so participants may approach Personalization and Recommendation from different angles, be it strategic, content-related, technical, or ethical. To make it more relatable to trainees, the course will present multiple uses cases from EBU Members.

While session 1 offers an opportunity to understand what Personalisation and Recommendation are and discuss systems from a strategic perspective, session 2 approaches Recommendations from a more technical perspective.

Please note that the two sessions have been designed to be accessible to technologists and non-technologists alike. We recommend that you take part in teams of 2-3 colleagues with complimentary professional profiles to make the most of the course and to get the conversation going with your colleagues and management in your organization.

Who is it for

• Teams of 2-3 colleagues
• Product managers, product owners, strategists, content managers
• Software engineers, developers, UX designers

What you will learn

Session 1: Personalization & recommendation - Strategic Overview

• What personalization and recommendation mean in the context of a media organization.
• Why you should consider Personalization and recommendations.
• Do you need a login? Why is a login useful for?
• Introduction to recommendations system: content to content recommendation and personalised recommendations.
• What is needed to deploy recommendations in your digital products?
• Make or buy? Alternatives?
• Strategic discussions on Large Language Models and conversational interfaces as the ultimate personalization

Session 2:  AI for personalization & recommendation – technology overview

This session will address personalization and recommendation from a technology angle and give trainees a chance to “look under the bonnet”:

• What are the prerequisites to deploy recommendations in your digital products.
• What are the different approaches: collaborative filtering, embeddings, Language Models, LLMs
• Use cases: recommendations for audios and videos, recommendations for news articles, recommendations to the journalists in the newsrooms. How the EBU can support you in your journey: technical infrastructure, advice, communities, surveys.