The authors of inventions that have ever had to implement their development with partners who could take care of all the commercial and financial problems, are well aware how hard it is to find a partner

With the advent of social networking social circle has expanded beyond recognition and it helps to more quickly, if necessary, to find a partner

But as always happens in such cases, increased with the number of social networks of potential candidates or candidates for the partners, it complicates the process of selecting candidates who fit the criteria for all partners such as

Moreover, the selection criteria do not always include only the financial resources and technological preparation of the candidate, but many other vital issues, including basic decency and a real professional reputation

This issue is part of a more general problem of finding business partners, investors, strategic partners and potential distributors, as well as subcontractors for the manufacture of innovative product

Very often with a significant complication of the design of innovative products, as well as the tightening of requirements for environmental safety and ergonomic parameters, at the stage of development of technical documentation required for the project professionals in these fields, which can be trusted with sensitive or confidential information about the innovative product

Because social networks are already functioning successfully for a long time, there is a lot of experts in these technologies, which tend to respond quickly to emerging problems of information

Here is an example of such an attractive design:

American scientists from the New York University Sinan Aral and Dylan Walker created the search technology of influential people in social networks.

As a result, researchers were able to reaffirm long-standing hypothesis of the existence of social scientists a "mighty heaps" - influential people form a relatively small and fairly dense subnet.

However, scholars interested in the applied aspect of the work - finding ways to conduct the most effective advertising campaigns in social networks. That is understandable, because the study was a place of testing the world's most popular social network Facebook.

The hypothesis that the spread of ideas in a society controlled by a small number of people, was first formulated in the 50th year of the last century.

However, many guessed this before - all the world's conspiracy theory is based on this simple idea. Until recently, however, research in this area stood still - analysis of the effect of a real person to another is a complex matter, and often quite bad.

The best that could count sociologists, it is common popular argument in the spirit, such as "tipping point" Malcolm Gladwell.

When this book came out, The Daily Telegraph called it "a surprisingly unusual study of the unknown phenomenon as a social epidemic."

With the advent of social networks have become commonplace unusual research - at those wishing to study social phenomena (such wish at this time joined by scientists of different areas, such as physics), there was a vast field for research.

Mathematics influence

June 21, 2012 in the journal Science published an article, the authors suggested that search technology influential people in social networks.

The method is quite simple: scientists are building a mathematical model of social networks, but because it is calibrated using experimental data.

"Lenta.ru" a lot of time writing about the basic principles of such a simulation. In short, the social network in the framework of this method is represented by a graph in which users are represented by vertices, and edges connect those users who are friends with each other.

In such a graph with a probability model is given by the speaker that reproduces certain social process.

In this case, the speaker has been arranged quite simple: each fixed time t let the vertices along the edges of the signal. The signal is not indulged in all the edges, and on several randomly selected from the members of this summit.

The number of signals received by a particular vertex function is determined by a random Poisson process. This is a random process, which at each moment of time t gives a random variable like this: the probability of dropping out of a non-negative integer k is equal to lk e-l / k!

The response to a signal from the user is determined not only by the number of received signals - it also depends on a few "social" parameters.

We considered a number of fairly obvious parameters - such as age, gender, status of relationships. Using the prepared array of experimental data (described below), the researchers applied to it the method of logistic regression.

The essence of this method lies in the generation of a test, which provides the input data of the two values 1 (the event occurred) or 0 (event did not happen).

Total peak could in some sense to obey a signal or not to obey the signal. In the case of the very top of subordination became his source.

The experimental data

As experimental data, the researchers used information provided by an unnamed company in the advertising campaign on Facebook.

The action was as follows: users sent out an invitation to the installation of some applications.

This application can be assessed on a scale films and recommend them. After evaluation of the program chose the Friends of the few people at random and on behalf of the person writing the chronicle of the message.

The report said that, they say, so and so is your friend A has put such and such an assessment here is this film. Try it and you have our application.

Similar reports have appeared, and after each subsequent assessment of the film by the user A. According to scientists, the principle of their applications, like the well-known principle of the application for Facebook called Flixster.

Advertising campaign lasted 44 days. During this time, its members were 7730 people with a total number of friends more than 1.3 million users. A total of 41,686 sent out automated messages, thanks to which the application currently installed 976 people (this is referred to in excess of 7.7 thousand, which become parties to the promotion of other channels).

Before analyzing the data collected, the scientists were convinced that their sample is devoid of statistical distortions.

To do this, they compared the demographic and age-based indicators of people who took part in an experiment with data about users of Facebook.

Here it should be noted that the company is Mark Zuckerberg does not disclose the personal information of its users, so scientists had to use third-party data istrategylabs.com.

As a consequence, in their own words, they obtained data on demographics registered, but not necessarily active users.

Using the experimental data, the researchers first calibrated its own mathematical model.

After that, they calculated the two user-defined function - the function of the impact and function of suggestibility.

The first describes how willing friends for members responding to its signals. The second determines how the user is willing to respond to signals from friends.

After experimenting with these features, the scientists concluded that the dynamics of the network is defined in equal measure suggestibility and influence.

The researchers also found that influential characters form a fairly small subnet (the dream of any fan of conspiracy theories).

It was found that people older than 30 years are less influenced by friends. In addition, on average, men were more influential than women, while some women have a greater impact on men than on other women.

Speaking on the distribution associated with the status of the relationship, the least affected by those who in this column on Facebook is "married" ("Single"). The chief, however, the discovery was that the influence and suggestibility almost never occur in one and the same person.

Roughly speaking, all people on Facebook can be divided into those who listen and those who speak.

Instead of

Experts embraced a new job with enthusiasm. "The work was carried out in a phenomenally grand scale" - leads ScienceNOW words of sociologist Brian Uzzi.

The scientist also said that such a clear stratification of the leading and trailing can lead to a revolution in advertising on social networks.

It is clear that companies are interested in how to lead a campaign by influential people, and the new method can detect them.

Group of influential people - this is precisely those groups of the composition of which can be selected by candidates for partnership and cooperation agreement for the commercialization of innovative technical solutions

It is also necessary to understand how in modern integrative inventions and innovations shared the role and influence in the process of commercialization and adaptation to the conditions prevailing in the market sector which may include implementing innovative projects

A variety of innovative projects and products, and defines the requirements for a variety of potential partners who are able to meet such requirements

At the same time Uzzi stresses that additional experiments to test how advertising migrates from the Internet in real life.

On the other hand, the prospects of the new method may be overestimated.
In particular, for its practical application requires significant computation, which hold for, say, a million users is extremely difficult.

So it is unlikely in the near future we should expect a revolution in marketing.

However, research proved the separation of members of the society on the leading and trailing leads once again to reflect on the nature of human relationships, especially in the innovation process.