Posts tagged Algorithms


“Random Selection in Random Image” is new website by Jan Robert Leegte. It does exactly that: it loads a random image from Flickr and it makes a random selection. Reload the website to get another random selection in a random image.

(via Random Selection in Random Image - today and tomorrow

Random Selection in Random Image” is new website by Jan Robert Leegte. It does exactly that: it loads a random image from Flickr and it makes a random selection. Reload the website to get another random selection in a random image.

(via Random Selection in Random Image - today and tomorrow

Essay-Grading Software, as Teacher’s Aide - Digital Domain - NYTimes.com

This spring, the William and Flora Hewlett Foundation sponsored a competition to see how well algorithms submitted by professional data scientists and amateur statistics wizards could predict the scores assigned by human graders. The winners were announced last month — and the predictive algorithms were eerily accurate.


The Web offers advertisers a slew of creepily effective targeting mechanisms, but they only work for some stuff, some of the time. An ad on the Web may do a better job of reaching its audience than, say, a magazine ad. But that doesn’t mean it does a good job.
Example: Here’s data from Nielsen, via Bernstein analyst Carlos Kirjner, which tracks the accuracy of a recent ad campaign by “a manufacturer of women’s personal care products.” It was supposed to target women between the ages of 25 and 54.
But most often it didn’t — the most accurate publisher got the ads in front of the right people 40 percent of the time. Overall, the campaign only hit the target 25 percent of the time. And nearly half the time — 47 percent — the ads got served to men.

(via Web Ads Need to Get More Accurate - Peter Kafka - Media - AllThingsD)

The Web offers advertisers a slew of creepily effective targeting mechanisms, but they only work for some stuff, some of the time. An ad on the Web may do a better job of reaching its audience than, say, a magazine ad. But that doesn’t mean it does a good job.

Example: Here’s data from Nielsen, via Bernstein analyst Carlos Kirjner, which tracks the accuracy of a recent ad campaign by “a manufacturer of women’s personal care products.” It was supposed to target women between the ages of 25 and 54.

But most often it didn’t — the most accurate publisher got the ads in front of the right people 40 percent of the time. Overall, the campaign only hit the target 25 percent of the time. And nearly half the time — 47 percent — the ads got served to men.

(via Web Ads Need to Get More Accurate - Peter Kafka - Media - AllThingsD)

Stocks: How Tweet It Is - WSJ.com

Money managers long have used various tools to gauge market sentiment, such as investor surveys or tallies of positive and negative news stories.

Twitter-based trading strategies gained momentum last year, after two papers were published by a team of Indiana University researchers. The first, which focused on Twitter and was published in the Journal of Computational Science in March, used a psychological rating scale to rate tweets according to six moods: calm, alert, sure, vital, kind, and happy.

For example, a tweet that read, “I’m so relaxed right now,” or “I’m in a bath listening to jazz,” would have a high “calm” rating, says Johan Bollen, one of the Indiana researchers.

UI Press | Stephen Ramsay | Reading Machines: Toward an Algorithmic Criticism

Besides familiar and now-commonplace tasks that computers do all the time, what else are they capable of?

Stephen Ramsay’s intriguing study of computational text analysis examines how computers can be used as “reading machines” to open up entirely new possibilities for literary critics. Computer-based text analysis has been employed for the past several decades as a way of searching, collating, and indexing texts. Despite this, the digital revolution has not penetrated the core activity of literary studies: interpretive analysis of written texts.

Computers can handle vast amounts of data, allowing for the comparison of texts in ways that were previously too overwhelming for individuals, but they may also assist in enhancing the entirely necessary role of subjectivity in critical interpretation.

Reading Machines discusses the importance of this new form of text analysis conducted with the assistance of computers. Ramsay suggests that the rigidity of computation can be enlisted by intuition, subjectivity, and play.


Eve Sussman’s experimental cinema project whiteonwhite:algorithmicnoir uses a computer to build a movie out of 3,000 video clips, 80 voiceovers and 150 pieces of music.
Though it makes choices in largely random fashion, the computer is surprisingly adept at editing together a pretty good movie. Sometimes.
whiteonwhite played three shows at the Sundance Film Festival (it has also been performed in New York and will soon travel to Berlin and Santa Fe, New Mexico). Of the three Sundance screenings, Sussman and co-writer Jeff Wood described one as so-so, one somewhat painful to watch, and one as outstanding.
“Sometimes it works like we expect it would…. Dramatic tension builds and then gets resolved,” Wood said in an interview with Wired.com. “Sometimes it doesn’t work at all.”
The program that drives whiteonwhite operates on similar principles to Pandora. Each clip has a specific tag that triggers the selection of the next clip. The tag “white” might pull up 80 “white” clips, from which the computer chooses one. Music and voiceover are assembled in similar ways. The process of selection is logged on a separate monitor for the audience to watch.

(via Algorithm-Powered Movie whiteonwhite:algorithmicnoir Computes at Sundance | Underwire | Wired.com)

Eve Sussman’s experimental cinema project whiteonwhite:algorithmicnoir uses a computer to build a movie out of 3,000 video clips, 80 voiceovers and 150 pieces of music.

Though it makes choices in largely random fashion, the computer is surprisingly adept at editing together a pretty good movie. Sometimes.

whiteonwhite played three shows at the Sundance Film Festival (it has also been performed in New York and will soon travel to Berlin and Santa Fe, New Mexico). Of the three Sundance screenings, Sussman and co-writer Jeff Wood described one as so-so, one somewhat painful to watch, and one as outstanding.

“Sometimes it works like we expect it would…. Dramatic tension builds and then gets resolved,” Wood said in an interview with Wired.com. “Sometimes it doesn’t work at all.”

The program that drives whiteonwhite operates on similar principles to Pandora. Each clip has a specific tag that triggers the selection of the next clip. The tag “white” might pull up 80 “white” clips, from which the computer chooses one. Music and voiceover are assembled in similar ways. The process of selection is logged on a separate monitor for the audience to watch.

(via Algorithm-Powered Movie whiteonwhite:algorithmicnoir Computes at Sundance | Underwire | Wired.com)

lettersfromhere:

This is emoji writ large. The Fühl-o-meter/Public Face is an interactive art installation that calibrates the mood of the city in which it has been erected with a monumental illuminated Smiley. The work of artists Richard Wilhelmer, Julius von Bismarck, and Benjamin Maus, the urban emoticon accurately communicates its host city’s gefühlszustand according to “mood data” obtained using integrated software which analyzes photos of the faces of passing pedestrians and processes emotions out of them. Mechanical armatures modulate the face’s expression in real-time, making it appear by turns happy, sad, or apathetic with corresponding gestures (smiley, frown, and blank).

(via Architizer Blog » An Urban Emoticon that Measures the Happiness of Cities)

Gifts and algorithms

Marginal Utility (Rob Horning) posts about “algorithmic cures for gift anxiety.” Pointing to this story about a Walmart tool called Shopycat, which I gather scrapes social media from your friends to tell you what they want for Christmas (or whatever), he observes: 

I’m sure this should probably be hailed as an economistic victory against the deadweight loss of Christmas. More people will get what they want, and less time and money will be “wasted” figuring it out. But in an algorithmically airless world of perfect emotional efficiency, where every gift given is the right one and the risk of social faux pas are eliminated, I’m not sure what will be left of the holiday spirit, which seems to hinge ultimately on a generous amount of familial forgiveness.

Well said … 

… although, my assumption is that the sequel service would be one that not only deduces what your friends want, but actually buys it and sends it to them. Right?

How do you engineer a system that, while not literally random, produces the feeling of serenidipitous discovery, meaning emerging from what seems like meaninglessness?

KS: The point is we’re using algos to analyze the world, but it goes beyond that: we’ve weaponized some of them, so they’re not just thinking, they’re acting. For me, it’s not about the finance piece. That’s just the visible apex of the thinking and practice. I really do believe that as surely as we had a flash crash on Wall St, we can have them in culture if we’re not careful. And the first way to be careful is simply to understand that there’s something to be careful about.
RW: And that “something to be careful about” involves the sort of creeping prevalence of algorithm-made decisions?
KS: That’s part of it. I’ve lived my whole life in New York City, so I’ve always been obsessed with why things are the way they are in New York. What’s interesting is that when you trace them, most of the reasons were generally around the needs of people and the stuff they did, like move crates from ships, or take trolleys. But if you look at how a lot of things are being shaped now, the criteria have nothing to do with humans. That’s weird.

(via A conversation between Rob Walker and co-founder of Area/Code, Kevin Slavin : Observatory: Design Observer)

KS: The point is we’re using algos to analyze the world, but it goes beyond that: we’ve weaponized some of them, so they’re not just thinking, they’re acting. For me, it’s not about the finance piece. That’s just the visible apex of the thinking and practice. I really do believe that as surely as we had a flash crash on Wall St, we can have them in culture if we’re not careful. And the first way to be careful is simply to understand that there’s something to be careful about.

RW: And that “something to be careful about” involves the sort of creeping prevalence of algorithm-made decisions?

KS: That’s part of it. I’ve lived my whole life in New York City, so I’ve always been obsessed with why things are the way they are in New York. What’s interesting is that when you trace them, most of the reasons were generally around the needs of people and the stuff they did, like move crates from ships, or take trolleys. But if you look at how a lot of things are being shaped now, the criteria have nothing to do with humans. That’s weird.

(via A conversation between Rob Walker and co-founder of Area/Code, Kevin Slavin : Observatory: Design Observer)

Army Tracking Plan: Drones That Never Forget a Face | Danger Room | Wired.com:
“Long Range,  Non-cooperative, Biometric Tagging, Tracking  and Location” system, and “Adversary Behavior Acquisition, Collection, Understanding, and Summarization (ABACUS)” tool, explained.

Army Tracking Plan: Drones That Never Forget a Face | Danger Room | Wired.com:

“Long Range, Non-cooperative, Biometric Tagging, Tracking and Location” system, and “Adversary Behavior Acquisition, Collection, Understanding, and Summarization (ABACUS)” tool, explained.

The Meaning Machine - Alexis Madrigal - Technology - The Atlantic

The Meaning Machine takes all of your inputs at one end — photographs, status updates, game plays, song listens — and transforms them into meaning that’s organized and designed.

The Meaning Machine is what happens when we apply statistical methods to human lives. Run regressions on your experience of the world and this is what you get. Right now we call it Facebook Timeline, but it will have many forms over the coming decades.

The Meaning Machine relieves you of the struggle to examine your experience of the world. You only need to post status updates and photos. Just live life and record it in social media. The Meaning Machine takes it from there. Feel the algorithm!

Read the rest here.

As Slavin correctly and instantly deduced, this is where I got the previous algorithm quote. In fact, I should have reblogged that quote from Slavin! But I was reading the Internet backwards, by accident.

Anyway read Alexis’ piece, please: It’s short, sharp, and totally excellent.

Much like our memories, Facebook Timeline understands that some moments have resonance that lasts through the years. It’s a marvel of computer programming: An algorithm that comes eerily close to emulating human memory; perhaps the first algorithm to spark such a deep emotional response.
A blueprint for the inevitable future of warfare: when time is critical and running decisions up the chain isn’t feasible, software will make key decisions about what constitutes a target, what falls within the bounds of the “rules of war,” and whether or not it’s safe to commence firing. If a program can satisfy whatever requirements have been seeded in its coding, then it’s bombs away.

feltron:

Onformative Actelion Imagery Wizard - made in Processing.

A bit more:

Actelion is a biopharmaceutical company …

[The job:] Create a new identity for our brand Actelion….

Based on the idea »From medical industry to medical magic,« a new visual identity was developed that uncovers the invisible magic moment of medicine.

The new imagery is based on the smallest possible unit: digital molecules. These capture the magical moment when new medicine arises from molecules; innovative design for a company that supplies the world with innovations. We worked closely together with Interbrand in creating the graphic imagery and developed a tool for automatic image generation that enables the generation of a unique, in-itself homogeneous graphic image world out of heterogeneous visual material. The new visual identity is featured in the 2010 annual report and continues to evolve in its updated website.
This looks cool, but also strikes me as sort of gimmicky. And design issues aside, the idea of “medical magic” gives me the creeps.