Experiments with cognitive systems to deepen the news experience

I am experimenting with how we can use cognitive systems — technology that uses natural language processing and machine learning — to provide users with the opportunity to interact more naturally with the news. To test this, I am developing a prototype of a user interface called “Walter” that will use voice technology to bridge the gap between the user and a newsroom. By engaging with “Walter,” users will have the opportunity to interact with the news of the day as if talking to their own personal journalist who provides valuable information to them.

How will your approach answer the journalism challenge you are trying to solve?

I see “Walter” as a bridge for the user to interact with the news, while also providing a holistic news experience. With users more mobile than ever before, it’s valuable to experiment with different ways to provide them with an experience that allows for the news to follow them.

Users want to feel as if they are at the center of a breaking news story. Users will be able to talk to “Walter” through voice technology, and ask questions like “What is happening in Syria? Tell me about President Obama’s State of the Union address?”

By having them engage with an intelligent system like “Walter,” I can actually create an interactive experience for them, provide them with information they want to know and, I hope, entice them to want to learn more later.

Users will be able to control how and when they get news from the interface. “Walter” will have the acumen to match the pace of a user’s day in order to provide the correct medium for news delivery. If you are “on the go,” for example, news would be delivered in audio format, while in the evening when users may have more time, they would get longer-reads and podcasts.

How is your approach different from what already exists?

IBM’s Watson, Apple’s Siri and Google Now are all using cognitive systems to help anticipate what information users may need, or to help answer specific questions. But such innovations have yet to be applied to journalism. It would be revolutionary to be able to speak into a device and have the news delivered to you. It would give the user more power to ask for detailed information. Cue up what stories you want to read or listen to in the evening. Give users full control of how and when their news is delivered — all through the sound of their own voice.

What elements of your proposal have you researched, tested or created?

Preliminary work is in progress right now, but I have done some initial interviews with users to learn about their relationships to their mobile phones, their daily phone habits, and how they get their news on phones. I’ve also begun initial design work on the interface.  I am blogging about my journalism challenge at www.donnaborak.com.

What are your immediate next steps?

  • Build and experiment with a low-fi prototype of the user-interface of “Walter,” conduct user-testing, iterate on ideas.
  • Start building my design and developer team. Gather a team of advisers and partners to help me gain more feedback and refine my prototype.
  • Begin pursuing funding options to build a next iteration prototype.

What resources or advice do you need to complete those steps?

I need collaborators and advisers with product design and technical expertise who want to work with me to build a unique news delivery experience utilizing big data and cognitive systems.

What specific aspects of your project would you most like to receive feedback about?

  • Do you know of any other organizations that are experimenting with the use of cognitive systems?
  • Do you know of anyone who may be interested in collaborating with me to build a prototype?
  • Is there any research on users’ experience with voice automated systems that would help me to better inform my design process?

This is the refinement phase of Borak’s effort to address a challenge in journalism. Learn more about her initial exploration phase of the process. Have questions or suggestions about this challenge? Email dborak@stanford.edu or follow her on Twitter @donnaborak