15 Creating secure personalized experiences

 

This chapter covers

  • Building on user data to convey trust users can believe in
  • Understanding privacy, security, authentication and identification needs
  • Exploring the difference between personalization and customization
  • Implementing authentication and identification
  • Making the right audio choices for your system persona

This chapter builds on what you learned about context in chapter 14 as you now explore creating individualized voice interfaces that users will trust and want to use. Even while writing this chapter, there were weekly reports of data breaches at big corporations exposing access to health records, financial information and passwords. In the voice industry, reports[35] that workers listen to your private conversations and that smart speaker apps are vulnerable to phishing contribute to growing concerns around intrusion on privacy. And let’s not even get into any international political or legal topics.

Headlines feed on everyone’s fears. After all, fear sells! The concerns are not based on lies per se—home voice assistants do trigger the mic by mistake when they shouldn't, and there’s no excuse for security breaches. The only way to improve VUIs is to collect user data, analyze it, and iterate to improve performance. What’s missing is an understanding about what’s worth worrying about and how to address it appropriately.

15.1  The importance of knowing who’s talking

15.2  Individualized targeted behaviors

15.2.1    Concepts in personalization and customization

15.2.2    Implementing individualized experiences

15.3  Authorized secure access

15.3.1    Approaches to identification and authentication

15.3.2    Implementing secure gated access

15.4  Privacy and security concerns

15.5  System persona

15.5.1    Defining and implementing a system persona

15.5.2    How persona affects dialogs

15.6  System voice audio

15.6.1    TTS or voice talent, generated or recorded

15.6.2    Finding and working with voice talents

15.6.3    One or several voices

15.6.4    Prompt management

15.7  Emotion and style

15.8  Voice for specific user groups

15.9  What’s next?

15.10   Summary