I found the Day 3 general session of Insight 2014 to be quite interesting for a number of reasons. Captain Richard Phillip's first-hand account of his ordeal and rescue was fascinating, and Kevin Spacey's keynote was one of the more inspiring messages at a conference like this. On the business side, the challenges faced by Singapore in identifying illegal or dangerous activities in and around their ports was eye opening, and then of course there was the little matter of the landmark partnership between IBM and Twitter announced mid-way through the session.
Here are my notes from the session:
- NORA - Non-obvious relationship awareness. Now called identity insight.
- G2 - Sensemaking
- G2 Vision
1. Evaluate each new observation against previous observations
2. Determine if what is being observed is relevant
3. Delivering this actionable insight to its consumer... fast enough to do something about it while it is still happening.
4. Doing it at substantial scale
1/2 of the world's supply of oil and 1/3 of global trade passes through Malacca Straights and Singapore Straights. 2000 ships pass through every day, of which 1000 will drop anchor in port. (That works out to 1 every 2-3 minutes.)
What they have discovered with G2: Which ships are doing illegal activities right in the sea. (Trading oil, etc.)
- e.g., which ships turn off their GPS's to hide location and sometimes identity. Stays there for 30 hours.
- Using satellites from above, can site that a ship is 3 meters higher off water than before the meeting, which likely means a transfer of freight without going to port. No transmissions (causing danger for other ships).
- Needed help to focus limited resources. Jonas referred to their starting point as "Fantasy analytics" - where tried to start with the limited data that they had at the time. The answer: widening the observation spaces
- How they perform their analysis:
- Divide the region into space-time boxes. e.g., 20km box by 1 day. Can see where vessels hanging out in places that haven't before, or places they haven't been before, or places with little activity.
- Looking for patterns:
- Two ships come from opposite directions, meet up, then go separate ways, then meet back up somewhere else 22 hours later.
- Or someone who has the same passport being used in two separate places at same time.
- Vessels changing identities over and over again (like car changing license plate)
- Identity Insight technology is now available in SPSS Modeler
- G2 Sensemaking for AML (anti-money laundering) will be available soon for the banking industry
Captain Richard Phillips, Maersk Alabama, interviewed by Mark Jeffries
[Fascinating first-hand recounting of the hostage situation and subsequent rescue.]
Bob Griffin, GM, IBM i2 threat and counterfraud
- i2 analyst notebook. (i2 Enterprise Insight Analysis)
- Created 20 years ago for intelligence and law enforcement community.
- Do visual hypothesis, relationship association, link analysis
- Advanced 3d visualizations, geospatial analysis with partners like Esri
How was software used by teams
- Enables the Navy and rescue teams to understand how hostage takers move, patterns in when they move (e.g., when moon is full?), what saying on social media
- Package up and deliver intelligence to the point of the spear - folks on ground providing operational support, making decision on best way to intercede or disrupt and defeat the adversary
- Allows to get complete picture on social awareness, understand who's on adversary group - their relationships within the group, who's a rising star, who has power.
Described what happens when someone logs into their website and indicates that they want to cancel their life insurance policy. Within millisceonds, that information is flagged and routed out to the member services rep organization so that the rep has the opportunity to evaluate the member's account status and pick up phone and give them a call. All of it is built all on the IBM stack.
Alistair Rennie, Paul Papas, IBM Interactive Experience
Twitter partnership - "As of today, business decisions will never be the same."
- Retail systems know what's on the shelf, but now can know what products people care about and are going across the street to buy at a competitor. Do further - understand what drove the person to that specific store. Companies can optimize supply chain based on what inventory will be needed before even there.
- Imagine ability to identify the birth of a trend. Move with speed to act on that new information.
Integrating Twitter with IBM technology in the cloud. (One centralized place in the cloud where it will be easy to get access to.)
- For business people, use it in sophisticated ways. Why integrating it with Watson Analytics.
- Important for data professionals, and will integrate with Big Insights platform, for people doing sophisticated analysis
- For Developers - will integrate into Bluemix. Untapped creativity, and want to see what happens.
On the difference between working with Netflix vs. a traditional network:
When pitched the idea for "House of Cards" to the networks, they wanted to see a pilot. When pitched it to Netflix, they asked how many episodes they wanted to do. Why the difference? Data
- Netflix knew their audience would do well with the show
- They knew audience liked the director
- They knew audience liked the British version
Netflix has data: they know when you pause the show, when you return, when you rewind, what zip code you're in.
Netflix will make 10 different versions of a trailer and tailor that trailer to you based on what you watch and where you are. (If you watch a lot of Kevin Spacey movies, you got the trailer with a lot of Kevin Spacey in it. If you watch a lot of David Fincher movies, you got a different trailer.)
It was estimated that Netflix would need to add around 500k more members to break even on their investment in the show, and have brought in 17 million members since the show launched.
Data allowed Netflix to agitate for change. They started a war, and now HBO is responding with HBO Go as standalone offering, without requiring cable package. CBS will make their shows available via streaming. And the winner will be you.
Entertainment is not the only industry going through this change. Uber is disrupting their industry. Uber uses data to send cars to area where they expect to be demand, before the demand shows up.