Social:Project networks

Project network — is the technological framework, an online service or the web site intended for provision of a possibility of self-organization to the participants possessing key competences in project team for execution of actions with initially set purposes, achievement of which defines project completion.

Difference from a social network

The fundamental difference of the offered network structure is establishment of contacts between the specialists interested in participation in the specific project, project implementation and its commercialization whereas the structure of the organization of the various existing social networks is oriented only to exchange of information. At the same time, the financing of projects can be the most various, from work on the specific order, to work in an initiative order, with subsequent participation in the tender or other type of commercialization of project deliverables.

Theoretical aspects

Studying specialized networks to which the offered design networks belong in theoretical aspect is far from end - it is connected with an advancing of practical application of network structures over their theoretical research. At the same time use of modern opportunities of information technologies allows to take a new look at the processes of formation of the scientific, technical and technological design groups and all aspects connected with the process of initiation and performance of work on the projects. Many, at first sight, diverse, researches in technology, sociology, telecommunication and other sciences are based on a postulate of not reducibility of structural descriptions to uniform models of specialized networks. However any researches and developments which are carried out with use of information technologies in many areas, it is possible to consider through a prism of classical project management. In this regard the theory developed is applicable for the social networks, but obviously needs a certain modification.

Approaches in research of project networks

At the moment, theories of social networks research, generally the following questions:

• statistical properties of networks;
• models of networks;
• forecasting of the processes happening in networks.

In applied researches are used typical characteristics, such as: network size, network density, degree and density of centrality and equivalence.

The general approaches applied to the analysis of social networks can be applied also to the analysis of project networks, it is natural taking into account certain specifics of initiation and maintaining projects. First of all, these specifics are reflected in elements which form a project network.

For further research of project networks it is reasonable to enter some concepts. At the initial stage the project network is in the "sleep mode", so there is a regular, for a social network, information exchange between potential participants of project teams and, therefore, in a network "traditional agents" or other actors. After that, within processes of initiation of projects, there are project teams (the description of mechanisms of initiation of project teams is beyond this article) that are expressed in the emergence of nodes with the increased concentration of connections, the project network leaves "sleep mode" and begins to carry out the functions on ensuring implementation of the initiated projects. For designation of the working teams, that the nodes made in a project network with the increased concentration of connections between actors, the concept of the actor of 1 level is entered. This basic concept for a project network as it designates availability in a project network of the operating command executing the project. If actors of 1 level establish connection among themselves, for example, within a matrix organizational structure, the concept of the actor 2 levels is entered. The graphical representation of actors of 1 and 2 levels is given in figure 1.

Figure 1. The graphical representation of actors of 1 and 2 levels

One of important elements of the research of project networks is creation of models which reflect specifics of their functioning. Applying the classification given in for the creation of one of the models of project networks, it is possible to suggest to use statistical models of social networks and, in particular, model of weak connections.

In the modern society these specialized networks of the informal relations allow to find a job, through "on-line exchange", to carry out information exchange, to resolve problems, passing the government and other traditional institutions, in some cases, they allow to receive orders on accomplishment small on amount of works (freelancers). There are bases to believe that increase of the status of a professional group (including project), leads to increase in a flow of information in networks of informal social and professional contacts. And, so-called weak information connections, i.e. connections with a little famous colleagues or project teams, can be more effective, than "strong connections" — with permanent employees, at the same time manifestation of effect of a synergy is not excluded.

In case of the creation of models of project networks, by analogy with social networks, it is necessary to enter the concept of clusterization. For example, if in the network count are available connection between tops 1 and 2, and between 2 and 3 it inevitably leads to connection between 1 and 3. In such models concepts of elasticity and coefficient of correlation of a network have an important role.

If to apply concept of an accidental network to the description of a specific social or project network, then from mathematics line items it will not be correct. In work it is specified that it is possible to pass to the concept of an accidental network through creation of statistical ensemble of networks (a set of networks) in which everyone the specific network has the probability of implementation, that is each network of ensemble has own statistical weight. After creation of such ensemble it is possible to calculate average value for some size in an accidental network, by averaging of this size on all implementations, accepting in attention their statistical weight. This, to a certain extent simplified approach is realized in accidental networks which are usually represented by accidental columns (Erdos-Renyi model). At this model in which statistical ensemble counts with a quantity of nodes X and a quantity of connections Y are provided all columns (networks) have the identical statistical weight of implementation. From this a conclusion follows that the probability of existence of connection between any two nodes is identical to such networks.

One of key characteristics of accidental networks, which is important for understanding of properties and processes which in them happen is such statistical characteristic of an accidental network as distribution of nodes on number of connections (DD, degree distribution).

The characteristic of DD, distribution of knots on number of connections of P(q) is probability that incidentally chosen knot in a casual network has the power of q:

[math]\displaystyle{ P(q)=\{N(q)\}/N }[/math]

Here {N(q)} — average of knots of degree of q in networks, at the same time averaging undertakes on the whole ensemble. The total number of knots at all members of this ensemble is identical and can be expressed as

[math]\displaystyle{ N=\sum^{n}_{i=1} {q}\{N(q)\} }[/math]

Researches have shown that distribution of nodes, in the considered accidental networks, on number suitable connections can be described according to the distribution law of Poisson. From this it is possible to draw a conclusion that in classical accidental networks approximately identical number of connections approaches nodes and there are no dominating nodes with a large number of connections (hubs). From the point of view of such approach the processes happening on small social networks and some types of specialized networks can be studied.

On number of connections on large social networks it is reasonable to apply sedate or exponential distribution to the description of probability of distribution of nodes. The conducted pilot studies have shown that real large networks have slowly falling down distribution of nodes on number of connections, and nodes, with the dominating number of connections, constitute noticeable part from connections of all network structure. Sedate distribution [math]\displaystyle{ p(q)=e^{-\lambda*q} }[/math] in case of great values of q - a regular example of slowly falling down distribution of nodes on number of connections. In fig. 2.a distribution of accidental process under Poisson's law and an approximate graphic representation of a network for q=4, and is given in fig. 2b for normal, exponential and sedate laws for which the approximate graphic representation of a network is given.

Figure 2.a Distribution of accidental process under Poisson's law and an approximate graphic representation of a network for q=4

Figure 2.b Distribution of accidental process for normal, exponential and sedate laws and an approximate graphic representation of a network (Sedate distribution of connections in nodes)

The important work allowing to understand one of approaches to creation of project networks is research P. Albert and L. Barabasha, on topology of computer networks which within pilot studies has found and actors concentrators (hubs) in different types of networks having the dominating number of communications in comparison with "regular" actors have theoretically proved. They have entered concept of the scale free networks and have revealed two conditions in case of accomplishment of which this type of networks arises:

• growth condition. After formation of a network with some small number of actors of n1, on each discrete temporary step the new actor with n (n ≤ n1) communications is added, and n condition ≤ is satisfied by n1 which connect the educated actor to n various already existing actors;
• condition of preference of accession. In case of the choice of actors with whom the new actor establishes connection, it is considered that the probability of joining of the new actor to existing, depends on former number of communications of the last.

The term "free scale network" means that in a network there are no nodes with some typical number of communications. The free scale network is a characteristic distinctive feature their increased resistance to damages. The such model authentically interprets project networks as actors of 1 level poorly interact with each other and the project, being the one-time entity possesses final time of existence, but in case of offices of projects in a network actors concentrators (hubs) begin to be formed. According to the theory R. Albert and L. Barabasha concentrators are often surrounded with smaller concentrators, and those, in turn smaller, etc. It also provides the increased stability of this kind of network structures. Loss of one of concentrators not crucially for a network as general communications will remain due to existence of other concentrators. Availability in the free scale networks Albert-Barabasha of concentrators of different "amount" does not contradict the fact that in project networks, by determination, teams of various number will be present and function. The larger the project is – the bigger number of actors unites in the actor of 1 level, that is in project team. However questions of interaction between actors of various levels need additional research. The internal infrastructure of project networks will be determined their properties and to develop by the principles or self-organization, or under external impact (influence) on a network.

Based on the given material it is possible to make the assumption that according to the characteristic of the DD network can be revolutionizing. At the formation stage, for example, of some social or project network, distribution of nodes on number of communications will submit to Poisson's law, and with growth of its popularity users will have expressed nodes concentrators and the characteristic of DD shall be described by the sedate law. It is not excluded that in case of recession of popularity users on a social network will have the return process, that is the network will "breathe". Thus, a network, social or project, it is possible to research as the dynamic system possessing a certain initial condition. This approach allows to study dynamics of the processes happening in network structures in case of process of transition of system of one condition in another. Set of all admissible conditions of dynamic system usually is represented through its phase space. Questions on the modeling of project networks through their representation as dynamic systems with specific initial conditions and research of their phase spaces, are of a certain scientific and practical interest, but do not enter a task of this work.

Universality the scale free networks specifies a way of further development of idea of creation and enhancement of project networks. So availability of larger, than from formation of numerous project teams, or even offices of the projects, actors concentrators in a network having a large number of communications can be treated as emergence in project networks of virtual associations on an industry sign, for example on nanotechnology, biology, the software, etc. Integration of such project network into the Unified information system of the Russian Academy of Sciences or into a scientific and innovative network of Russia can be the following level (emergence of super concentrators). Within international cooperation the Canadian "Network of the centers of perfection (NCE)", the German program "Network Management of the East (NEMO)", the French network of the scientific researches CNRS or such programs of the EU as "Eureka" and the European Technological Platforms can act as super concentrators of a project network, for example.

The analysis of the main approaches to the research and the modeling of project networks is at the initial stage. It is required to execute a large volume of work for creation of mathematical models of project networks of various degree of complexity and to determine a technique for studying of the processes happening in these structures. In case of the description of some properties of project networks (correlation, transitivity, structure of consolidation), at the moment, we have to rely on the factors possessing high degree of uncertainty.

The creation of theoretical bases of the analysis and synthesis of project networks will be important for practical implementation of this perspective view of the network structure.

See also

• Social network
• Project management

Literature

• Timofeev K.N. Project networks //Innovation management: from theory to practice Collection of papers of VII annual (II international) research and practice conference of management faculty National Research University: Higher School of Economics - Saint Petersburg, 2012
• Gubanov D.A., Novikov D.A., Chkhartishvili A.G. Social networks: the models of informational influence, management and confrontation. - Moscow.: FIZMATLIT Edition, 2010, P. 228
• Evin I.A The introduction to the theory of complex networks. //Computer research and modulling. Vol 2, N2, 2010
• Dorogovtsev S. N. Lectures on Complex Networks, Oxford University Press, Oxford, 2010
• Albert, R. and Barabasi, A.L., 2002, Statistical mechanics of complex networks, Rev. Mod. Phys. 74. Voronina L.A., Ratner S.V. The scientific innovation networks of Russia: experience, problems, opportunities.- Moscow.: INFRA-M, 2010.- P 254.