Microsoft Office 365

Using Machine Learning to Help Process Email – Microsoft Adds Clutter to Declutter the Inbox

The Office 365 team announced (and released) Clutter for Office 365 today, a machine learning method of decluttering a user’s inbox by automagically moving lower priority items out of the Inbox and into a secondary folder.

Clutter learns from your actions to determine the messages you are likely to ignore. As less important messages arrive, they are automatically moved to the Clutter folder. Clutter does this by leveraging Office Graph’s sophisticated machine learning techniques to determine which messages are Clutter. It gets smarter over time, learning from your prior actions with similar messages, and assessing things like the type of content and even how you are addressed in the message. The Clutter experience is personalized to each individual and reflects an email experience that adapts to your actions and preferences without you having to do anything. The information Clutter learns from each user’s actions are only applied to that user’s experience and are not shared with anyone else.

Clutter is powered by Microsoft Office Graph, which is also the power behind the recently released Delve capability for Office 365. A few points in terms of how it works:
– the Office 365 administrator needs to enable Clutter for their Office 365 network.
– each user needs to activate Clutter processing in the Outlook Web App.
– once enabled in Outlook Web App, a secondary folder is created (“Clutter”). This is shown across all devices and clients accessing Office 365, including the Outlook client on a PC / Mac and on mobile devices.
– Clutter doesn’t override existing Rules that a user has created to sort messages.

A few reflections about the idea:

1. The proof is in the pudding. It’s a great idea, and is sorely needed to help with processing email. Whether it works effectively enough will take time to figure out.

2. The nuances in how it deals with particular messages is important. If a user always ignores messages from Bob (and has done for the last 156 messages), what happens if the 157th message is directly applicable to their work and of high priority? Will the Office Graph be smart enough to figure that out?

3. When Microsoft announced the Office Graph earlier this year, it said that Delve (codenamed Oslo) was the first application of that idea. Here’s another one. I look forward to seeing what else is coming.

4. Microsoft is getting very Apple-esque in its approach to releasing things. No endless pre-announcements about things coming. Just cool stuff done and dusted. E.g., last week it was the partnership with Dropbox for Office integration. What’s coming next week?

5. In thinking about #4 above, one of the beautiful things about cloud based services is the opportunity for the vendor to recreate a more direct relationship with each end user, and hopefully by doing so create increased loyalty among a much larger customer base. In pre-cloud days, everything was vetted through the IT Department, and this didn’t always end well.

6. Darrell Webster tried it out and shares his experiences.