Mapping bilateral information interests using Wikipedia edit activities.

We live in a global village where electronic communication has eliminated the geographical barriers of information exchange. With global information exchange, the road is open to worldwide convergence of opinions and interests. However, it remains unknown to what extent information interests actually have become global.

To address how interests differ between countries, we analyzed the information exchange in Wikipedia, the world’s largest online collaborative encyclopedia. From the editing activity in Wikipedia, we extracted the interest profiles of editors from different countries. Based on a statistical model for interest profiles, we created a network of significant links between countries with similar activity. Using clustering, we find that countries can be divided into 18 clusters with similar interest profiles, which suggests that language, geographical proximity, religion, and historical background diversify the interests.

Each color represents cluster of countries that share similar information interests. 

We quantify the effects of these factors using regression analysis and find that information exchange indeed is constrained by the impact of social and economic factors connected to shared interests.

Link to the paper in Nature Palgrave Communications.

Data for download.

Why some societies are more cooperative than others?

Imagine you are living in a remote village and because of a harsh winter the village is cut off from the main source of electricity. There is no electricity to use for light, heating and television. Luckily, one villager has a generator that can provide sufficient electricity to all other villagers. The only condition is that all villagers should follow rules so that no one overuses the electricity. Will the generator work just enough for everyone or will it collapse?

The true story of a village in the Netherlands in 1978 inspired my psychologist friend Anna Sircova, me and other collaborators to model how people cooperate with respect to their time perspective.

“We live in and with time. Due to various reasons, we can easily become overly oriented on the future, get stuck in the past, or live completely in the present moment. Therefore, how we perceive time, can determine how much we are willing to cooperate. The perspective on time, not only creates personal differences, but it determines our social behavior, something that has been largely neglected in modeling social behaviors. ”, says Anna Sircova.

In our recent paper published in PlOS ONE, we combined aspects of personal time perspective with social interactions to investigate to what extent people cooperate. We used time perspective profiles from 25 countries to compare which ‘villages’ could survive the winter longer. Based on this results we found that UK, New Zealand and Germany are among those countries with high cooperation index. In the lower ends there are Lithuania, Mexico and China.


These results are correlated with other socio-economic indices such as Human Development Index (HDI). There is a link between how our perception of time creates cooperation in the society and how society encourages us to perceive the time.

The generator in the small village in the Netherlands collapsed twice and situation could only be resolved by appointing patrols to monitor other’s misuses, but we hope that theses results can be used to improve psychological aspects of cooperation to prevent social tragedies.

Structural differences between public and personal online interactions.

In the online space, even though our social encounters are not tightly bounded to “offline” social norms, still we try to build a good image of ourselves to be able to maintain our online relationships. Therefore, it is crucial to understand how people create “online” social norms and to what extend such norms resemble our ancient social norms.

In our recent paper, Link, we analyzed 6 years of online communication in a movie based website. We analyzed individuals in the social network when they discuss movies in forums or send personal message to one another.

Such data enable us to study the formation of “online” social norms and it’s evolution.

In this post, which is a short summary of our paper, I am addressing these questions:

  1. To what extend do we create social ties based on the means of the communication in the online space.
  2. How much do we exchange communication with someone based on the public or personal interactions?
  3. Is there any limit on how many unique people one can interact with per day in online space?
  4. Does creating a new social tie demand social investment in online space?

Now let’s turn to some of our analysis and results:

  1. In our daily life we encounter many people in public places. However, people we choose to interact with in our personal space are selective and limited. Similarly, in online space we observe that people create more ties in public forums than in personal messaging space.
  2. 1In our daily life, we socialize with more people in public places; we say hello to our clients and have coffee with our colleagues. In our private life, we interact with fewer people, our family and beloved ones, but we exchange more intimacy and communication with them. We observe similar behavior in online space; the distribution of communicational exchange (number of communications going back and forth between two people) is larger in personal messages.4
  3. If you wonder if there is a limit to how many new people one can socialize with per day in the “online” space, the answer is yes! While majority of the community members socialize with one or two new people per week, the upper limit is about 20 new people per day. In other words, we don’t observe anyone in this community who socializes with more than 20 other members per day. That brings us to the idea of Dunbar’s number [Ref.] that suggests human brain has a certain capacity of creating and maintaining social ties. We dig deeper into the idea of maintaining social ties in the next section.6
  4. One can assume that in the online space, there is no limit on creating and maintaining social ties. We can create and maintain as much ties as we desire (time is the only limit!). But this is not true! Looking at the pattern of creating a new tie, we see that it takes longer to create the next tie in personal space than in public space. And the more ties one has, the longer it takes to create the next tie in personal space. We need to put energy and time to keep our friends, even in online space!


Our study attempts to shed lights in understanding how the means of communication, alter the structure of social networks, how do social norms emerge in online communities and how we communicate and organize our online life.

Technology appropriation

We are all consumers of technology. We adopt technology out of our curiosity or peer pressure and we consume as we wish.

However, the story doesn’t end here. We also assign our own subjective meaning to technology that we consume. Designers, design technology in a certain way and we consume it in a way to be more comfortable for us. This is what is called as technology appropriation. 

“Goods are neutral, their uses are social, they can be used as fences or bridges.” Douglas and Baron(1996)

In our recent paper, we briefly touch upon this concept. [Read Here]

Recently something funny happened to me. My mom got a smart phone but she is not very good at using touch screen. In general her approach to the technology is rather unusual. For example when she writes a message to me in Skype, she doesn’t know that there is a BIG button called “space”. So she always puts “.” between every word! Something like this:

Going back to the story. We made a Viber group with my sisters and mom to make our communication easier. Since my mom doesn’t know how to use the touch screen, she came up with her own solution to communicate with us. Guess what!



She wrote: Dear S. I love you

She draws by her finger and post the image! I am not sure how easy it is to find the drawing app but apparently for her it is easier. I think this is a good example of the technology appropriation!

Do you have any of these examples?

P.S. Special thanks to Ann Samoilenko for introducing me the concept.

What we don’t know about Stanley Milgram.

Most people recognize Milgram by his research on the obedience. This review attempts to highlight Milgram research on topics except small-world and obedience experiments. The review is based on the book  “The man who shocked the world”  by Thomas Blass.


Followed by the work of his advisor Asch on Conformity, Milgram conducted social conformity setting in another setting in Norway. The subjects were asked to answer some questions where they were told that there are other subjects in the experiments (bogus).


Illustration 1: level of conformity in each sets of experiments. The blue color shows Norwegian subjects and the gray color shows French subjects.

The subjects, follow bogus majority about 62% of the time. However, if the answer have serious consequences, the conformity dropped to 50 % (aircraft condition). In private condition, where subjects write down the answer instead of announcing it, the level of conformity was 50 %. In bell condition, the subjects could ring bell and ask for repetition of the words. That leads to 69% conformity. In Censure condition, the confederates, gave contradicted opinion right after the subject respond and that jumped the conformity to 75%. Milgram repeated the experiment with Norwegian from different social classes and there was no statistically significant differences. Milgram repeated the experiment in France. What he discovered is that social conformity in general was lower among Parisian than Norwegians.

In another example, Milgram showed that people can behave differently within a group than as an individuals. The subjects increased the level of shock as a learner gave incorrect answer (like in obedience setting). However, the subjects either act in group – where the other two confederates called for on-step shock increase each time the learner made mistake – or they act individually.


Illustration 2: The subjects increase the level of shock as a result of being in group pressure.

lost letter

Milgram not only examined six degree of separation, but he also experimented the diversification of the subject of the targeted letter and the subject who send the mail. The technique also used to predict elections in one county. More letter had been mailed from either party, could determine what party people are voting for. In terms of ethnicity, they also found differences between black and white subjects. His research on distributing letters addressed to pro-Peking and pro-Taiwan, was unfinished.


Illustration 3: results of lost-letter technique. Depends on the location of the letter and the address, the number of letters that had been returned varied.

 City psychology

Help stranger –  Milgram conducted various experiments to explain behavioural differences between urban and rural residents. He introduced an idea of overload – in which a system  holds more input than it can be understood. In such experiments, he addressed the readiness to help stranger. People went house to house and asked if they can use the telephone for emergency calling. What he found was that people in cities were less helpful than the small town residents.

In another experiments, with his students, they conducted series of experiments in the New York subways. The students asked the passengers to give their seats to them. They wanted to know in what extend people are willing to give their seat to a stranger.

Mental maps –  How the objective geographical layout of a city was represented subjectively in the minds of its residents.  He explained “… people make many important decisions based on their conception of a city, rather than the reality of it. …”. E.g. the Parisians were asked to draw the map of Paris. Milgram recognized certain order of entering a place in the map. That Shows the importance and the history of the city. The mental map of the city show how the city sits in the mind of its residence.

Vertical city versus flat city – How does living in skyscraper modify thinking and behavior? What effect does the vertical life have on human relationship?

Familiar stranger – The students gave a photo of crowds standing in the subway to the passengers and ask the subjects to recognize the people in the photo. 89% of the people recognized at least one familiar stranger. On average, they reported seeing 4 familiar stranger. 47% of people were curious about the familiar stranger.


Illustration 5: The familiar stranger experiments. The people in the experiments were asked to name which passengers they can recognize from the photo.

“… in order to handle all the possible inputs from the environment, we filter out inputs so that we allow only diluted form of interaction. In the case of familiar stranger, we permit a person to impinge on us but close off any further interactions. In part this is because perceptual processing of a person takes less time than social processing. We can see a person at a glance but it takes more time to sustain a social involvement.”

Power of crowds

Milgram designed an experiments to test the social contagion. Up to 15 stimulus crowds, gathered in a busy street in New York looking up to the sky. They realized, as the number of stimulus increase from 1 to 15, more passerby gaze up to the sky.


Illustration 6: Power of the crowd experiment. The size of the crowd who look up the sky increase as a function of the stimulus crowd.

Movies that were conducted by Milgram

Nonverbal communication

The city and the self

Conformity and independence

Human aggression

Invitation to social psychology

How nations mix ingredients in their food based on traditional Persian medicine.

January 2010, I was in Berlin with a good Turkish friend. She is vegetarian. At some point she expressed her stomachache. I turned to her and said “of course you have stomach ache because you ate this and that together”.  She looked at me strange that what I am talking about!

That was a shock to me. The ingredients mixing knowledge that back in Iran sounds completely obvious to everyone, is not that obvious even in Turkey! I started explaining her the rules on how to combine foods and my friend was amazed. Then after, I tried to observe how other nations combine food and how the knowledge of food mixing is widespread.

In 2011 I met Yong Yeol Ahn in a conference. He was researching on the networks of food pairing in various cuisines. Gratefully, he shared the dataset of recipes with me.  Gradually, as a small side project, I started looking at some aspects of that data.

Let me first start with sharing with you a background of the tradition knowledge of ingredient mixing. Traditional Persian medicine developed by Avicenna (981-1037). Avicenna in his well-known book, The canon of Medicine, integrated and studied the principle of various medicine such as Greeks, Europeans, East Indians, Persians, Arabs, Chinese and Tibetans. In this book, he aggregated variety of methods and makes them into one standard principle of medicine.

Avicenna methodology is based on the ancient theory of four temperaments, known as Unani medicine (Tibb). In the Unani medicine there is not a single cause for disease, rather a result of disorder in the body metabolism.

Unani Tibb, a humoral medical system, presents causes, explanations and treatments of disease based on the balance or imbalance of the four humors (akhlaat) in the body: blood (khun), mucus (bulghum), yellow bile (safra), and blackbile (sauda). These combine with four basic qualities (quwaat): heat (garmi), cold (sardi), moisture (rutubaat) and dryness (yabis). Dominance of one of the humors in the body gives each person his or her individual temperament (mizaj) : sanguine (dumwi), phlegmatic (balghumi), choleric (safravi), and melancholic (saudawi) ( reference)

For Avicenna and his contemporary followers, there is no such thing as a ‘single’ cause of disease; rather an outcome of various factors, such as the food and body metabolism. Avicenna believed most illnesses arise solely from long-continued errors of diet and regimen.

Based on Unani system, foods are said to be either ‘hot’ (garmi) or ‘cool’ (sardi). This classification is consistent with other systems that seek to obtain an overview of the metabolic process, such as the East Indian Ayurvedic medicine and traditionsl chinese medicine.

In this post, I do not intend to discuss how this medicine developed and spread out over the world. Rather, I want to take a look on how other nations mix their food ingredients based on traditional Persian food culture.

Persian cuisines adapted the Unani system in the way of combining different ingredients. In Persian culture, foods are categorized into two big categories, hot and cold. As a rule of thumb, basic components of a food are metabolically hot. Cool ingredients are added seasonally such that the whole cuisine has good hot-cold balance.

Foods that are sardi, or cold, place several burdens on the body. They are harder to digest. They are harder to assimilate (absorption of micro-nutrients into the blood stream and cells; moreover consequently, they leave a greater residue of superfluous waste products. (reference)

Here is the hot-cold list of ingredients that I collected from different Iranian sources. I tried to keep it reliable but also considering crowd wisdom. Therefore it might not be completely accurate. When I was browsing through different Iranian websites, I noticed some of categorizations are contradictory. I tried to check more than two or three sources to make sure that the categorization is reliable.

hot (garmi) ingredient list: scallion, pea, roasted_pea, soybean, radish, sesame_oil, nutmeg, walnut peanut, peanut_butter, fenugreek, caraway, cayenne, basil, vanilla, parsley, bell_pepper, green_bell_pepper, chickpea, leek, broccoli, peanut, berry, carrot, ginger, red_bean, nutshell, coconut, kakao, cabbage, honey, dates, sesame, mango, morus, banana, peach, orange, eggplant, potato, mushroom, shiitake, savory, mint, saffron, pepper, grape, green_bean, pistachio, wheat, pasta, noodle, bread, egg, cumin, almond, coconut, hazelnut, celery, shrimp, camel_meet, liver, butter, lamb, chicken_broth, bird, chickpea, cilantro, cinnamon, raisin, lobster, sesame_seed, cayenne, marjoram, oat, tarragon, thyme, cauliflower, fennel, oregano, lavage, nutmeg.

cold (sardi) ingredient list: citrus, oyster, cheese, jerusalem_artichoke, bitter_orange, snap_bean, green_bean, cod, cilantro, starch, pork, tomato, vinegar, coriander, genus, watermelon, cucumber, pumpkin, Pomegranate, cheese, watermelon, pear, apricot, okra, spinach, corn, grain, rhubarb, squash, lime, lettuce, rice, lentil, strawberry, dill, coriander, yogurt, milk, fish, veal, apricot, rabbit, goat, beef_broth, roasted_beef, margarine, butter_milk, celery, sprouts, zucchini, spinach, cabbage, okra, cauliflower, broccoli, white_potato, sweet_potato, carrot, cucumber, soybean, turnip pea, lima_bean, kidney_bean, coconut, brown_rice, cilantro, rheum tuna, salmon, veal, bacon, crab, pork_sausage, catfish, tamarind, mussel.

I used the data from this study to check ingredient combinations in different geographical regions. Here is a sample of how the data looks like:

North America: green_bell_pepper, wheat, meat, onion, chicken_broth, cayenne, scallion, celery, smoked_sausage, bell_pepper

For various geographical regions I plotted number of hot and cold ingredients per recipes. Each dot in the plots is one recipe. The darker the blue, shows more recipes are lying there.




The figures show the similar relation between using number of hot and cold ingredients across different geographical regions. Now lets take an average of each plot and compare them together.s


So what we have learned from this analysis; There is a difference in the number of ingredients across regions. In some regions food are simpler, consists of few amount of ingredients. That are points in the bottom left of the plot. Such regions are north America, Europe and Middle East. Whereas in Asian, African and Latin American cuisines number of ingredients per recipe on average is higher.

There is a relationship between number of hot ingredients versus cold ingredients. Out of 10 ingredients in one recipes, the fraction of hot to cold ingredient is approximately 6.5 to 3.5. Why is that?

Of course there are limitations in this analysis that one needs to consider: the hot/cold categorization is based on Persian traditional food culture. Many ingredients are not listed. Many ingredients change their effect when they are cooked. Ingredients have different portion in one dish, here we ignore that. With more globalization, the culture of food is changing and maybe categorizing recipes based on geographical space is not 100% accurate.

There are open questions here: What is the scientific reason behind the relationship between hot and cold ingredients? Is it related to the chemical reactions or temperature in different regions? Can this knowledge be used to diagnose some disease related to food or stomach ache? Can we use this knowledge to combine ingredients and come up with new recipes? Can we make recipes with right combination of ingredients to give us the best metabolism?