The Secret Ingredient of the World’s Best Apple Pie

Algorythms can predict the future, but unlike the Delphi oracle, they do it based on hard scientific methods, instead of intestines and animal bones. Finding old friends on Facebook, recomending books you might like on Amazon, or predicting the outcome of the 2012 presidential election – you name it, an algorythm does it.

So what about food? Could these same math whizzes help us bake a better pumpkin pie or mix up a tastier batch of sweet potatoes this Christmas? Lada Adamic, a computer scientist at the University of Michigan and Facebook, thinks it just might be possible. she and her team have come up with an algorithm to guess how successful a recipe will turn out. And the math works surprisingly well. It predicts with nearly 80 percent accuracy how many stars your mother's cranberry recipe will receive on allrecipes.com. Plus, it can recommend ingredient replacements to make your pie crust and potatoes more healthful.

She and her team took nearly 50,000 recipes and 2 million reviews from allrecipes.com and then hacked up an algorithm to extract out all the ingredients, cooking methods and nutritional profiles. With just these items, her algorithm could predict the recipe's rating with an accuracy of about 70 percent. But the magic happened when Adamic built a "social network" for the ingredients. She looked at how often two ingredients appear in the same recipes. Those that frequently show up together — milk and butter, nutmeg and cinnamon, basil and rosemary — sit close to each other in the network, but those that rarely appear in the same dish, such as coconut and parsley, are far from each other.



Physicists at Harvard University performed a similar network analysis on ingredients' flavors, but Adamic took it a step further and integrated the data into a recipe prediction program.

Adamic's network analysis boosted the accuracy of her recipe recommendations by about 10 percent. But it also revealed a treasure-trove of information about the way Americans mix and match ingredients, which ones we like to leave out or throw in extra.
Her algorithm analyzed reviewers' recommendations for customizing recipes, such as "I replaced the butter in the frosting by sour cream, just to soothe my conscience about all the fatty calories" and "This is a great recipe, but using fresh tomatoes only adds a few minutes to the prep time." Then the mathematics stitched together little clusters or communities of interchangeable foods and spices.

The result is a list of recipe replacements more comprehensive and scientifically accurate than anything you'll find in the Joy of Cooking or online.

Read the full article on arXiv.com, or a lighter verison on the npr blog.

Network Science of the Game of Go

You can make networks from pretty much anything. Connect music based on taste or phone calls, companies based on their ownership, spread routes of abstract movements, and lots more. It is high time to start using networks to understand games. But what of the structure of games themselves? In a paper that was recently published in EurophysicsLetters, two French scientists decided to apply network science to the game of Go.

They constructed their networks in a simple way: If one board position can lead to another, they are connected. Using a dataset of about 1,000 professional games and 4,000 amateur games, they began to construct these networks.

In a Game of Go players put black and white stones on a grid board.

Of course, the Go board is very large and so you can’t compare entire board layouts. Instead, they decided to make it much more tractable and look at the board composition surrounding a newly placed piece (a move in Go consists of putting a stone on an intersection of the grid lines of the board). In this case, they looked at the pieces immediately surrounding a newly placed piece (for a 3×3 grid). They calculated that this creates 1107 possible moves, which can be connected if the moves occur one after another, and are in the same region of the board. They also examined the frequency of moves, which obeys a heavy-tailed distribution (whether or not it is a power-law as they claim seems a bit weaker).

The network analyses in the paper are a bit odd, though they find many classic graph structures, such as a heavy-tailed link distribution and high amounts of clustering. Gratifyingly, the networks constructed from amateur and professional games are distinct, though in somewhat subtle ways. 

Read the article here, or the short version on Wired!

SNA – The Secret Weapon against Terrorism

In his forthcoming book Network Science, Albert-László Barabási has already reported about the role of social network analysis in the capturing on Saddam Hussein. Our readers know, that the blog itself is no stranger to the subject. A new American paper sums up how and why this approach can be useful in fighting political violence.

The academic community studying terrorism has changed dramatically in the past decade, and the descriptive and explanatory potentials have grown strongly. On of the reasons for its popularity is the increasing acknowledgment within the academic community of the important association between the group’s dynamic and (social)structure, and its members’ motivations and behaviors.

The Network of the Terrorist Group responsible for the Attack against WTC.

Understanding the motives and the processes that led the group to engage in political violence requires a look beyond the apparent causal relations between the causes of the violence and the violent activities. Since September 11th, growing numbers of media outlets have increased their coverage of terrorist incidents and groups. This, combined with the striking increase in the efforts and resources invested in data collection about these groups by academic and governmental agencies in recent years.

An IRA statement.

Violence or political action is a result of collective action, i.e., an output of a process, which is an action of a group of actors who interact with each other on some level, so SNA seems like an obvious analyzing tool. The sizes of these gropus varies widely (from 2 man groups to milites like the IRA), so an instrumental approach in bigger newtorks focus on command and information channels and the roles of leaders.

After drawing the networks of terrorist organizations, measuring the influence and power of individual actors becomes relatively easy, with the help of centrality and betweenness measures. Unveiling the hierarchy could also help authorities in dismantling them, making targeting a lot less complicated.

You can read the whole article at the academia homepage.

Ckeck out the original article for more.

Research is the New Discovery

A software called Livaplasma helps you discover new music you might like, with music you already do. 

This is what a music search looks like, if you look for music like Led Zeppelin, bands from the are and the genre become part of the network. Cream, The Who and Pink Floyd are the top recommendations.

The crator of the site Frédéric Vavrille, made the site back in 2004. But music recommendations is not all Liveplasma can do. It also works for books and movies.

With Inception as a key word, the network consist of other Cristopher Nolan movies, and genre specific works like The Girl With the Dragon Tattoo and The Matrix.

Jane Eyre's networks include Jane Austen's other works, books by Dickens and the Bronte sisters.

Curious? Check out the site for yourself at liveplasma.com !

The Global Super-Entity - The Economic Ruling Class of The World

A small, tightly woven network of companies, mostly banks, wields disproportionate control over the global economy, according to a new study. The findings shed some light on the intimate ways 21st century capitalism works — and how those functions can undermine the entire system.

A trio of systems theorists at ETH Zurich examined the world’s 43,060 transnational corporations and studied their share ownerships, searching for commonalities that tie the companies together. They worked with techniques used to study complex systems in nature to construct a model of which companies controlled which other companies, and through which networks.

Ultimately, Stefania Vitali, James Glattfelder and Stefano Battiston identified a core of 1,318 companies with interwoven ownerships, each with ties to two or more other companies. They were connected to an average of 20 each, the researchers found. The network forms a “giant bow-tie structure,” with a small, tight knot in the middle and connections spanning outward in an increasingly nebulous pattern. The knot is very small and dense compared to the other sections, and the researchers dubbed it an economic “super-entity.” It is also very closely held — about three-quarters of the ownership remains in the hands of the core itself.

While the authors note that there’s no example of this core intentionally acting as a bloc — in other words, there’s no vast economic conspiracy — that doesn’t mean it can’t act that way. “Globally, top holders are at least in the position to exert considerable control, either formally (e.g., voting in shareholder and board meetings) or via informal negotiations,” they write.

“Nearly [40 percent] of the control over the economic value of TNCs in the world is held, via a complicated web of ownership relations, by a group of 147 TNCs in the core, which has almost full control over itself,” the authors explain. Unsurprisingly, three-quarters of these companies are banks.

The Core of the Network

They add that domestic anti-trade strictures prevent the core from acting as some kind of cash cartel.

Concentrated power in the hands of a few has clear implications for global financial stability — which everyone already knows, given what the world went through starting in 2008. But this study puts it in empirical terms. Further studies that build upon the assumptions made in this paper could potentially help policymakers and economists studying ways to stabilize financial markets.

The Top 20 Corporation in the Core of the Network:

1. Barclays plc2. Capital Group Companies Inc
3. FMR Corporation

4. AXA

5. State Street Corporation

6. JP Morgan Chase & Co 

7. Legal & General Group plc 

8. Vanguard Group Inc

9. UBS AG

10. Merrill Lynch & Co Inc 

11. Wellington Management Co LLP

12. Deutsche Bank AG

13. Franklin Resources Inc

14. Credit Suisse Group

15. Walton Enterprises LLC

16. Bank of New York Mellon Corp

17. Natixis

18. Goldman Sachs Group Inc

19. T Rowe Price Group Inc

20. Legg Mason Inc

Source: New Scientist

The Dark Side of Hospitals

After virtual threats and food poisoning, a new study takes a closer look at viruses in hospitals.

Hospitals shouldn’t make you sicker. But plenty of people acquire illnesses while hospitalized—in some countries, such so-called nosocomial infections afflict more than 10 percent of patients.

Jack Nicholson's life might not be the only one threatened by a nurse.

To investigate transmission pathways, European researchers of the SocioPatterns collaboration fitted 119 people in a ward of the Bambino Gesù Children's Hospital with radio-frequency identification (RFID) badges. The tags registered face-to-face interactions—and the potential spreading of airborne pathogens.

The map.

Nurses interacted with the widest variety of individuals across the ward—patients, doctors, other nurses, and so on. The study indicates that nurses should take priority in strategies for preventing or controlling hospital outbreaks.

Different groups of the hospital.

The scientific method used in the analysis was developed in the MIT Media Lab. The sociometric badges aim to eliminate behavioral changes that occure because they are participating in an experiment. The devices are capable of  capturing face-to-face interactions, extracting social signals from speech and body movement and can also measure proximity and location of the users. The invention was listed as on of the top 10 innovations by the Harvard Business Review.

Check out the interactive map on Scientific American! For more about the method, we recommend the company's page.

Can Social Media Become the Saviour of Democracy ?

An article in Nature claims to have proven the  direct impact of  social media on political activity. Researchers at the University of Carolina along with people from Facebook run a gigantic experiment.

On Nov. 2, 2010, the day of the nationwide Congressional elections, nearly every Facebook member who signed on — 61 million in all — received a nonpartisan “get out the vote” message at the top of the site’s news feed. It included a reminder that “today is Election Day”; a link to local polling places; an option to click an “I Voted” button, with a counter displaying the total number of Facebook users who had reported voting; and as many as six pictures of the member’s friends who had reported voting. The results: 340,000 additional votes nationwide! Pretty amazing, but how can we be sure these people would not have voted by themselves?

Two randomly chosen control groups, of 600,000 Facebook members each, did not receive the pictures. One group received just the “get out the vote” message; the other received no voting message at all.By examining public voter rolls, the researchers were able to compare actual turnout among the groups. They determined that the message showing friends who had voted was directly responsible for 60,000 more votes nationwide and indirectly responsible for 280,000 that were spurred by friends of friends — what they called “social contagion” effect.

Significantly if not surprisingly, the voting study showed that patterns of influence were much more likely to be demonstrated among close friends, suggesting that “strong ties” in cyberspace are more likely than “weak ties” to influence behavior. It also found an indirect impact from the messages: friends of friends were influenced as well.

Fun fact, they also discovered that about 4 percent of those who claimed they had voted were not telling the truth.Because only about 1 percent of Facebook users openly state their political orientation, the researchers said they could not determine whether political leanings had any influence on social networking and voting behavior.Past studies have shown that a variety of methods for mobilizing potential voters have a disappointing effect. Knocking on doors is the most effective technique; e-mail is one of the least.