Saturday, December 14, 2013

Data Portraits - Visual Data Mining

There is no Wikipedia entry for the term "Data Portrait" ---yet.  Somewhat surprising.  All data representations are portraits but obviously there is an emerging conceptual sub-set view.

Images of various data portraits here at this link.

A fascinating data portrait has stuck in the back of my mind since seeing it a year ago.  It displayed one day of geo cell phone tracking movement in a European (Helsinki?) city.  The city had both a major bridge in the center and also a ferry system across a lake or river.  It was  time lapse video that showed the influx of cell phones to the city from outskirts and then the return to the suburbs in the late afternoon.  Concentrated inner city movement during the day.  It looped at the end of the video and the impression was one of a city breathing in and out with the cyclical movement of cell phones.

Speaking of breathing, this link examines the environmental quality of air we breathe related to monitoring traffic using cell phone geo traffic movement information.

Each cell phone of course representing a very high correlation to a single person.  Tracking was certainly done by recording real time geo cell phone signal to tower communications and the passing of the signal from tower to tower.

This type of macro geo tracking of cell phones in a specific selected map area has some fascinating information applications.  The first obvious one is traffic flow for traffic management purposes in planning and maintenance of traffic patterns.  Another data map that is analogous to cell phone tracking is geo tracking airline flights.  I have seen them represented as a video time lapse.  A video of the all real airline traffic for all aircraft similar to the map showing all flight connections of an airline that we can see in the back of airline magazines that are the last passenger refuge of boredom on a flight.

For example:  Couple cell phone tracking to airports, flights across a geographic area and then cell phones from airports to destinations and the Data Map would reveal something interesting.  Commercially valuable?  The cell phone data filter would be the key to value of macro cell phone movement data which is the same filter problem faced by the NSA.  All data has to be recorded but a means to pick a needle out of the haystack has to be devised.  All needles have to be tagged somehow for later filter application once a target or target has been selected.

If target selection is based on all cell phones entering/leaving an airport then there has to be a track back in time to the path from an to the airport.  At some point in that path the generation of data input for subsequent search comes up against privacy barriers that must be overcome by anonymizing the info, authorizing the info or disregarding any privacy barriers.

This link presents a cell phone tracking data portrait model and discusses privacy and anonymous identity.

Data Portraits paint fascinating visuals.

This link maps the web as artwork.

Yesterday I saw a graphic showing cell tower transmissions in Chicago.  It applied coloring to the the various bandwidth hopping transmissions to create a rainbow effect. 

This link creates personal portraits of people related somehow to their quantified self.  I would have to dig to see how it is done but it uses the term. " Albiac asked people to submit photos of themselves to his site, where they were turned into triptychs using images from the Hubble Space Telescope."  It is a higher level of abstraction to manipulate a data information representation into an somewhat abstract but recognizable painting portrait of a specific person.  How would Mona look as a result of this view?

A technical discussion of data portraits at this link.  It is one of the most interesting links that I have examined recently.  There are some golden nuggets in the link:


"Using metaphor: Metaphor is a powerful, but sometimes tricky, way to introduce meaning. Metaphors help us understand one thing in terms of another – it has been argued that all of our ability for abstract thought is by using metaphor to building up understanding from a physical foundation."


"The metaphorical poetics of PeopleGarden [9], which portrays the participants in a discussion group as flowers and the group as a garden, makes interpreting meaning intuitive. As the height of a real flower indicates its age, the height of a PeopleGarden flower indicates how long someone has been posting to the group. The number of petals represents posting frequency: a lush blossom indicates an engaged contributor. A petal’s color indicates whether it is an initial posting or response and the color fades over time: it is easy to remember that a faded flower is an inactive participant. The flower metaphor makes the portraits easily legible. It gives them visual appeal and a sense of vitality: rather than a dry statistical graph, here the data appears as an enticing garden. The problem is that the metaphor overwhelms the content it depicts. A person is portrayed as a pretty flower no matter what he or she is saying."

"In the world of information visualization, the goal is to depict the data as objectively as possible. This is the opposite of art, where the artist’s subjective vision is central. Data portraits sit between these extremes: their techniques come from the world of statistical analysis, but their purpose is artistic. Some may be closer to one extreme or the other – neither is “right”, but understanding where a particular portrait falls in this subjectivity continuum is a key element in understanding its function."
A data Selfie?????   We are still the nominal owners of our own privacy so no privacy rules violated in using our own private info for any expression we choose.   
Referring again to the prior link examining Data Portraits


"Privacy and control are closely intertwined: having a relatively innocuous piece of information revealed without your consent may feel like a violation, while freely displaying personal information of your own volition can be empowering. Privacy depends on audience." 


"Privacy and control are central issues in portraiture. When the subject controls the portrayal, and chooses to reveal little other than the public image, the result portrait is often dull and lifeless: e.g., many official portraits of politicians and executives. Portraits that reveal the subject’s vulnerability, such as Nan Goldin’s photographs, are more interesting to the audience, but they may be very discomforting to the subject. How much is revealed by a portrait and who controls it is the center of the tension between artist, subject and audience."


Perhaps the best way to preempt privacy is to make "non-privacy" the default mode on a web connected device or the price of entry to interaction on the web.  This is an startling example of pin pointing location using photo related location info publicly revealing privacy information.

This is a fascinating data portrait: Facebook's migration station: what our location updates say about human movement.

This link is a selfie data portrait of personal travel tracked by a cell phone for 3.5 yearsThe author is a co-founder of software that conserves battery life and provides device tracking info.  

Search travels around web revealed the term Visual Data Mining relates to Data Portrait.  

Visual Data mining image search results here at this link.

This link presents a visual representation of traffic noised mined from some data base of information related to traffic in Helsinki produce by HeatMiner software.   Helsinki is probably the location of daily cell phone traffic patterns that I mentioned earlier in this post although I still cannot find the specific link.   HeatMiner is a fascinating product I found at Cloud'n'Sci.fi   An aggregator of news related to Algorithms as a Service.  I like the logo at the site!  From the site that has a model of building a business on Open Data.
"The Cloud'N'Sci.fi marketplace enables a new way of making Business from Open Data. Instead of making a complete business solution of your own from scratch, commercialize your data sources and algorithms as data refining solutions and/or utilize existing solutions to build lightweight end-user applications!"

This branches from the discussion of Data Portraits but connects to Bitcoin.  The developers of Bitcoin were more crypto algorithm scientists/Engineers/Builders than just programmers.  Cloud'nSci.fi has an interesting presentation at this link on how Algorithm specialists can build a startup.  Like Bitcoin?  Algorithms.  The new gold.
 




No comments: