PANEL « BUILDING TRUST » EBU BIG DATA CONFERENCE, 21 MARCH 2017 Pierre-Nicolas Schwab Head of Big Data, RTBF (Belgium) [email protected]
PANEL « BUILDING TRUST »EBU BIG DATA CONFERENCE, 21 MARCH 2017
Pierre-Nicolas Schwab
Head of Big Data, RTBF (Belgium)
WHY TRUST MATTERS
• Trust in medias has decreased further: 26% trust in online media *
• Online media lesstrusted of all. Yet, main source of news among younger audiences(source: Reuter digital news report 2016)
* Source: Kantar Sofres (2016, 2017)
ALGORITHMS TOO CAN BUILD (DIS)TRUST
Personalization represent a challenge with 3 key problems identified :
– Too much personalization key information may be missing
– Alternative viewpoints may be absent
– Privacy may be threatened
Source : Reuters Digital Report 2016
5 RULES TO MAKE ALGORITHMS MORE ETHICAL
1. Avoid discrimination
2. Promote gender equity
3. Open up customers’ minds (exploration) rather than trapping them (exploitation)
4. Respect right of not being tracked
5. Educate users on algorithms
EDUCATION IS KEY
• Empower users: give control back
• Build knowledge
• « Show » your algorithms (open the black boxes !)
“An [algorithmic] system that you can't audit is a system that you can't use" Marc Rotenberg, CPDP conf. 2017
4 MEASURES TAKEN AT RTBF (1/4)CONFIDENTIALITY CHARTER
• New privacy and confidentiality charter (effective 01 Dec 16)
• Simplified yet complete
• Clear promises
• Special focus on children
• Dedicated video
4 MEASURES TAKEN AT RTBF (2/4)SINGLE-SIGN-ON
• Single-Sign-On (SSO) launched 01 Dec 2016
• Personal information centralized
• Easier for user to control, modify and erase information
• Essential tool for reconciliation of online usages and respect of GDPR
4 MEASURES TAKEN AT RTBF (3/4)RECOMMENDATION SYSTEMS
• Design of recommendation systems follow ethical rules
• Personalization will follow the concept of deliberative democracy
• Focus on exploration and discovery (≠ filter bubble)
4 MEASURES TAKEN AT RTBF (4/4)EDUCATION TO ARTIFICIAL INTELLIGENCE
• 1-hour TV program with real users (shooting on 29 March)
• Communication on corporate website
• Badge on website to signal algorithmic curation
• Systems designed to support visualization of algorithms’ effects