Analysis of Webspaces of the Siberian Branch of the Russian Academy of Sciences and the Fraunhofer-Gesellschaft

IT in Industry, vol. 6, 2018 Published online 09-Feb-2018
Copyright © Dehmer, Klimenko, Shokin, Rychkova, 1 ISSN (Print): 2204-0595
Dobrynin, Konstantinova, Medvedev 2018 ISSN (Online): 2203-1731
Analysis of Webspaces of the Siberian Branch of the
Russian Academy of Sciences and the
Fraunhofer-Gesellschaft
Matthias Dehmer
UMIT
Hall in Tyrol, Austria
Andrey A. Dobrynin, Elena V. Konstantinova,
Andrei Yu. Vesnin
Sobolev Institute of Mathematics SB RAS
Novosibirsk, Russia
Olga A. Klimenko, Yuri I. Shokin, Elena V. Rychkova
Institute of Computational Technologies SB RAS
Novosibirsk, Russia
Alexey N. Medvedev
Central European University
Budapest, Hungary
Abstract—In this paper, two webspaces of academic
institutions of the Siberian Branch of Russian Academy of
Sciences (SB RAS) and of the Fraunhofer-Gesellschaft (FG),
Germany, will be investigated. The webspaces are represented by
directed graphs possessing vertices corresponding to websites. An
arc connects two vertices if there exists at least one hyperlink
between the corresponding websites. Webometrics is used for
ranking the websites of SB RAS and FG. We discuss numerical
results when studying the websites structurally. In particular, we
examine scientific communities of the underlying websites
representing directed graphs and draw important conclusions.
Keywords—network; webometrics; quantitative measure;
communities
I. INTRODUCTION
In this paper, a webspace is a structural object (graph)
formed by a set of websites and hyperlinks between them [1]. To
investigate a webspace structurally, we use methods from
webometrics, i.e., the contemporary method for studying
information resources, structure and technology features of the
web. The development of webometrics has started in 1997 after
the seminal paper of Almind and Ingwersen [2]. Methods from
webometrics possess statistical nature and do not serve as full
description of diverse information processes that occur in the
webspace.Therefore to analyze the structure of webspaces, we
are goingto use graph-theoretical methods [1, 3].
There have been a large number of contributions in the
literature for studying webspaces resembling university websites
and academic institutions [4-12]. Since the number of webspaces
to be studied is infinite, there is still space left for performing
research on this topic. In this paper, we tackle this problem by
studying websites from Russia to investigate the underlying
institutions specifically.
It is well-known that the structure of real-life networks is not
random [13]. When dealing with non-random topologies, the
structural heterogeneity (or complexity) may be captured by
calculating various measures [14, 15]. Another problem in this
context is to determine the community structure of networks.
Community detection has been one of the hot topics in network
sciences and, hence, the problem received considerable
attention in the last decade [16]. The concept of the community
in a network is usually derived from common understanding of
communities in social networks [17].Graph-theoretically, the
problem has been defined by identifying the set of vertices
which are more tightly connected compared to the rest of the
network. Note that in order to solve the community problem, a
precise mathematical quantity Q has been introduced based on
the following description: to partition the vertex set of a network
into a union of subsets that maximizes Q [18].However, this
problem has been proven to be NP-hard and only heuristics
algorithms are available to determine Q [19].
Here we consider webspaces generated by websites of
academic institutions of the Siberian Branch of the Russian
Academy of Sciences (SB RAS)and academic institutions of the
Fraunhofer-Gesellschaft (FG), Germany. The structure of these
webspaces is formedby websites of the scientific institutions and
hyperlinks between them. Websites of SB RAS and FG will be
ranked by using methods from webometrics, and numerical
scores for examining community structure of webspaces
aredetermined. Since Russian webspaces have only been little
investigated, we believe that our work will have an impact for
the webscience community.
II. REPRESENTATION OF WEBSPACES
A simple model for representing the structure of webspaces
is a directed weighted graph G = (V, E) with vertex set V and arc
set E. We assume that webgraphs [14] do not possess any
self-loops and multi-arcs. In this paper, vertices of V correspond
to websites. Suppose that the vertices v and u of G correspond to
sites X and Y; then an arc (v,u) connects the vertices v and u if

there exists at least one hyperlink in X, referring to site Y. The
number of hyperlinks from X to Y is represented by the weight w
of the arc (v,u). The distance d(v,u) between vertices v and u in a
graph is the number of arcs in the shortest directed path
connecting them.
Fig. 1. Webgraph R of institutes of Siberian Branch of RAS.
Fig. 2. Webgraph F of institutes of Fraunhofer-Gesellschaft.
Let R and F be webgraphs of SB RAS and FG, respectively.
Their structures are shown in Fig. 1 and Fig.2.Graph R has95
vertices and 949 arcs while G consists of 72 vertices and 321
arcs. Additional information on these graphs can be found in
[20,21]. The number of arcs going from (to) a vertex v is
denoted by deg+
(v) (in-degreedeg-
(v)). A pair (deg+
(v), deg-
(v))
gives information on vicinity size of v.The weighted out-degree
(in-degree) wdeg+
(v) (wdeg-
(v)) is the sum of weights of arcs
coming from (reaching) a vertex v. Total degrees are defined as
deg(v) = deg+
(v) + deg-
(v) and wdeg(v) = wdeg+
(v) + wdeg-
(v).
A vertex v is called isolated if deg(v) =0. The degree
distributions of the graphs R and F are shown inFig. 3 to Fig. 6.
III. METHODS
A. Ranking academic institutions by using webometrics
The “Ranking Web of World Research Centers” is an
initiative of the Cybermetrics Lab, a research group belonging to
theConsejo Superior de Investigaciones Cientificas (CSIC),
Spain. Quantitative methods have been designed to measure the
scientific activity on the Web to determine ratings of universities
and research centers of various countries [22].The cybermetric
indicators have been useful to evaluate science and technology
and they serve as a proper complement to the results obtained by
using bibliometric methods connected to scientometric studies.
Fig. 3. Degree distribution of graph R.
Fig. 4. Degree distribution of graph F.
Fig. 5. Weighted degree distribution of graph R.
Starting from 2008, the Institute of Computational
Technologies of Siberian branch of Russian Academy of
Sciences (SB RAS) generates ratings of websites of scientific
institutions of SB RAS [6, 8, 21]. The ranking method is
presented in [22]. In this paper, we are going to extract statistics
from three major search engines: Yandex [23], Google [24], and
Bing [25]. To evaluate websites, the method uses the following
parameters:

• V – visibility. The parameter equals the number of
external links from other websites to the considered one.
Since the data from different engines is distinct, the
average value is taken: V = (VYandex + VGoogle + VBing)/3.
• S – size. The parameter equals the number of webpages of
the website determined by the search engines. Again, we
use the average value: S = (SYandex + SGoogle+ SBing)/3.
• R – richness value. The parameter equals to the number of
documents the website has with file extensions of Adobe
Acrobat (.pdf), Microsoft Word (.doc) and PowerPoint
(.ppt). The quantity is determined by search engines'
query, therefore we use averaging: R= (RYandex+ RGoogle)/2.
• Sc – citation index obtained from citation system Google
Scholar [26]. This parameter reflects the academic
importance of the website.
The overall rating evaluation includes the following steps.
1. Evaluation of the visibility V, size S and richness R
parameters for all websites in the network.
2. Ranking the values of V, S and R. The parameter array,
say V, is ranked in decreasing order. The website with maximal
V receives rank Vr = 1. The websites with identical values of V
get equal ranks. Similarly, we compute the ranks Sr and Rr by
using the parameters S and R for each website in the network.
3. Evaluation of the rank of the citation index Sc. We
compute the values Sc. The rank Scr is obtained by ordering
these values. The website with the minimal value receives the
rank-value Scr = 1.
4. Computing the sum of the obtained ranks for each
website: W = Vr + Sr + Rr + Scr.
5. The final rating is obtained by sorting the list of W scores
in increasing order. Therefore, the lower the value of W is, the
higher is the rank (rating position) of the website.
B. Quantitative measures for webgraphs
One of the common approaches when studying web
structures is based on quantifying structural information by
using various quantitative measures [14, 15]. Usually, a
quantitative graph measure is a graph invariant that maps a set of
graphs to a set of numbers such that invariant values coincide for
isomorphic graphs [27]. Such invariants can quantify either local
or global properties of graphs. Local measures, as a rule,
describe a graph structure near particular vertices. In contrast,
global measures encode structural information of the entire
graph. Some global invariants may be regarded as a complexity
measure of a graph [28, 29]. We consider the following graph
invariants.
The average degree, adeg(G), of a n-vertex graph G is the
average value:
=
1
=
∈
1
.
∈
The weighted analogue, awdeg(G), is given by the formula:
=
1
=
∈
1
.
∈
The diameter, diam(G), of a graph G is the largest distance
between two vertices:
= max , | , ∈ }.
It says how far one can travel in a webspace without any
repetitions of websites.
The vertex index, cv(G), of a graph G. This invariant
indicates which part of a websiteis involved into information
relationships(every website of this part has at least one arc).Let
G be a n-vertex graph with k isolated vertices. Then
! = 1 −
#
.
The quantity cv(G) reflectstages of webspace growth.
Namely, cv(G) is close to 0 in the initial stagewhen forming the
webspace; the value cv(G) = 1 indicates that allwebsites are
contained in the network.
The arc index, ca(G), of a graph G.The maximal number of
arcs in a directed n-vertex graphis equal to n(n–1), n> 1.Let
Ghas t arcs. Then the arc index is defined as
!$ =
%
− 1
.
This graph invariant is also referred to as network density
[30].The quantity ca(G) shows which part of arcsparticipate in
changes between websites. The maximal value ca(G) = 1
expresses that one can reach any other website by one click
starting from an arbitrary website.
The betweenness centrality, betw(v), of a vertex v shows the
importance of a vertex in terms of routing and connectivity.
This quantity [31] is a local graph invariant defined as:
& % =
'()
'()
,
(* *)
where σst is the total number of directed shortest paths from
vertex sto vertex t and σst(v) is the number of those paths that
pass through v.
The clustering coefficient, cc(G), of a graph G.By writing
neighborhood of a vertex v, we refer toall vertices that are
adjacent to v (without orientation of arcs).Let V2 be the set of all
vertices of a directed graph G with deg(v) = 2.Let Gv be the
directed subgraph induced by the neighborhood of v.The
clustering coefficient for a vertex v is defined by ca(Gv), i.e. it is
the arc index of Gv[17, 32].Then the clustering coefficient of G
is the averagevalue of the clustering coefficients for all vertices
regarding V2, namely:
!! =
1
| +|
!$ .
∈ ,
The introduced numerical graph invariants are applied tothe
web graphs R and F.

C. Communities in graphs
We study the community structure in terms of splitting the
vertex set V into non-intersecting subsets (or communities) that
maximize the directed and weighted modification of modularity
coefficient [18]. Denote by wij the weight of an arc (i,j) of a graph
Gwith vertex set V = {1, 2,…, n}. For weighted degrees of
vertices and the total degree w, we getwdeg+
(i) = Σ(i,j) wij,
wdeg-
(i) = Σ(j,i) wjiand w=Σiwdeg+
(i) = Σiwdeg-
(i). Then the
modularity Q(G) can be defined as
- =
1
. /0 −
1
2
/,0 ∈3
456/, 607,
where Ci is the cluster of vertex iand δ(Ci,Cj) is the Kronecker
symbol. It equals1 if the vertices i and j are in the same
community; otherwise it equals 0. The unweighted version of
modularity, Qun(G), is obtained from Q(G) by omitting the
weight from every arc. That is for every arc (i,j) we assign a new
weight wij’ = 1 if wij ≠ 0 and wij’ = 0 otherwise. If a graph G has
q arcs, then Qun(G) can be written as follows [18, 33]:
-89 =
1
:
.1 −
1
:
2
/,0 ∈3
456/, 607.
The quantities Q and Qun are applied to the graphs R and F.
IV. RESULTS AND DISCUSSION
A. Ranking academic institutions
The results from final ranking the websites academic
institutions of Siberian Branch of RAS and Fraunhofer-
Gesellschaft are presented in [20, 21]. First parts of the rankings
are shown in Table I and TableII. The comparison of four
webometrics quantities for these websites is presented in Fig. 7
and Fig. 8. From these rankings and involved computations (V,
S, R and Sc) we are able to perform several observations:
TABLE I. RATING SCORES P FOR INSTITUTIONS OF SB RAS
P Name of organization Website address W
1
Portal of Siberian Branch of Russian
Academy of Sciences
www.sbras.ru
7
2 Institute of Computational Technologies www.ict.nsc.ru 22
3 Institute of Cytology and Genetics www.bionet.nsc.ru 22
4 Budker Institute of Nuclear Physics www.inp.nsk.su 27
5 Sobolev Institute of Mathematics www.inp.nsk.su 35
6 Institute Computing Simulation icm.krasn.ru 35
7
State Pubic Scientific
Technological Library
www.spsl.nsc.ru
42
8
A.P. Ershov Institute
of Informatics Systems
www.iis.nsk.su
44
9 Branch of SPSL SB RAS www.prometeus.nsc.ru 51
10
Institute of Automation and
Electrometry
www.iae.nsk.su
53
11
Institute of Problems
of Developmentof the North
www.ipdn.ru
56
12
Novosibirsk Institute of Organic
Chemistry
www.nioch.nsc.ru
60
13 Boreskov Institute of Catatysis www.catatysis.ru 61
14 Presidium of SB RAS www.sbras.nsc.ru 68
15 Kirensky Institute of Physics www.kirensky.ru 70
TABLE II. RATING SCORES P FOR INSTITUTIONS OF FG
P Name of organization Website address W
1 Fraunhofer Headquarters www.fraunhofer.de 6
2
Institute for Systems
and Innovation Research
www.isi.fraunhofer.de
26
3
Institute for Open Communication
Systems
www.fokus.fraunhofer.de
30
4
Institute for Manufacturing
Engineeringand Automation
www.ipa.fraunhofer.de
34
5
Institute for Industrial
Mathematics
www.itwm.fraunhofer.de
37
6 Institute for Solar Energy Systems www.ise.fraunhofer.de 42
7 Institute for Industrial Engineering www.iao.fraunhofer.de 43
8 Institute for Laser Technology www.ilt.fraunhofer.de 43
9 Institute for Integrated Circuits www.iis.fraunhofer.de 46
10
Institute for Information Center
for Planningand Building
www.irb.fraunhofer.de
59
11
Institute for Factory Operation
and Automation
www.iff.fraunhofer.de
62
12
Institute for Algorithms and
ScientificComputing
www.scai.fraunhofer.de
72
13 Institute for Building Physics www.ibp.fraunhofer.de 75
14
Institute for Intelligent Analysis
and Information Systems
www.iais.fraunhofer.de
78
15
Institute for Wind Energy and
Energy System Technology
www.iwes.fraunhofer.de
82
Fig. 7. Final rating scores of the first 15 institutions of SB RAS.
Fig. 8. Final rating scores of the first 15 institutions of FG.
• 27 websites of the SB RAS network and 18 websites of the
FG network have more than 1000 external links.

Therefore, 28% of websites of the SB RAS network and
25% of websites of the FG network have sufficiently many
external links;
• 88% of the websites of the SB RAS network and 95% of
the websites of the FG network have more than 100
webpages. The composition of the websites of the SB RAS
and FG networks is similar: the number of websites for SB
RAS with R> 100 is 47 (45%); for FG we obtain 48 (35%);
• the Google Scholar citation index for FG websites is
greater than for websites of the SB RAS network: the
number of websites with parameter Sc exceeding 10 is 42
(44%); for SB RAS and 66 (92%) for FG.
B. Quantitative graph measures
Global quantitative properties of the considered graphs are
presented in Table III. The average vertex degrees of the graphs
decrease twice while the weighted average degrees decrease ten
times after deleting the administrative hubs. The diameter of the
graphs ranges from 2 in R and for F to 7, respectively.
The vertex index cv indicates that three graphs contain
isolated vertices, i.e., the corresponding websites aren't involved
in any communications. That means nobody can neither visit
these websites, nor leave them. Namely, graph R has 2 isolated
vertices. The other invariants of Table III show that almost all
global and local arc saturations of webgraphs are very small.
TABLE III. QUANTITATIVE INVARIANTS OF WEBGRAPHS
Invariant R (SB RAS) F (FG)
average degree, adeg(G) 9.99 4.46
average degree, awdeg(G) 743.21 763.88
graph diameter, diam(G) 2 2
vertex index, cv(G) 0.98 1.00
arc index, ca(G) 0.11 0.06
transitivity coefficient, cc(G) 0.07 0.09
A local invariant, the betweenness centrality, shows the
webgraphs are very centralized. Approximately 8% of vertices
possess a significantly higher betweenness centrality score and
degrees comparing to the rest. The betweenness centrality
scores are shown in Fig. 9 and Fig. 10. Some structural
properties of webgraphs R and F have been studied in [9].
V. DISCUSSION AND CONCLUSION
The search of modularity maxima has been performed by
using a combination of heuristic algorithms, mainly based on the
tabu search algorithm [33]. The observed heterogeneity
expressed by vertex degrees and the betweenness centrality
score makes it difficult to reveal communities in the network.
The best obtained modularity score was ≈ 0.15 for R and ≈ 0.13
for F (see Table IV). Whenever the algorithm assigns a
community to one of the most influential vertex, its
neighborhoods showed a tendency to fall into this community.
The numbers of the corresponding communities and their sizes
are presented in Table V.
Fig 9. Betweenness centrality distribution of graph R.
Fig 10. Betweenness centrality distribution of graph F.
TABLE IV. MODULARITY RANKS
R (SB RAS) F (FG)
modularity Q 0.153 0.131
modularity Qun 0.155 0.252
TABLEV. NUMBER OF COMMUNITIES AND THEIR SIZES
R (SB RAS) F (FG)
Q 5 (46,41,6,1,1) 5 (62,4,2,2,2)
Qun 8 (30,18,14,11,10,10,1,1) 7 (21,10,10, 9,9,9,4)
Along with the weighted modularity, we also computed the
unweighted version. The unweighted graph showed a weaker
community structure when considering R compared to F. We
emphasize that by omitting the weights, we got better partitions
(see Table IV and Table V).
It is evident assuming that the communities in these
academic networks should reflect scientific collaborations
between the corresponding institutes. This hypothesis has been
checked for the SB RAS graph R, where we composed the found
partitions into communities based on the subject areas of
institutes. The resulting subject partition has modularity rank
Q = 0.115 for R which is far from the optimally obtained
partitions.
ACKNOWLEDMENT
This work was supported in part by the Austrian Science
Funds for supporting this work (project P26142), the Russian
Foundation for Basic Research (grant 16-01-00499), Presidium
RAS (grant 0314-2015-0011), the Leading Scientific Schools of
the Russian Federation (grant 7214.2016.9).

REFERENCES
[1] S. Chakrabarti, Mining the Web: Discovering Knowledge from Hypertext
Data, Francisco: Morgan Kaufmann, 2002.
[2] T. Almind andP. Ingwersen,“Infometric analyses on the world wide web:
methodological approaches to “webometrics”, J. Document., 1997, vol.
53(4), pp. 404-426.
[3] M. Dehmer, Strukturelle Analyse web-basierter Dokumente. Multimedia
und Telekooperation, Wiesbaden: DeutscherUniversitätsVerlag, 2006.
[4] M.L. Arslan and S.E.Seker,“Web based reputation index of Turkish
universities”,2014. http://arxiv.org/ftp/arxiv/papers/1401/1401.7547.pdf
[5] A.A. Pechnikov and A.M. Nwohiri,“Webometric analysis of Nigerian
university websites”. Webology,vol. 9(1), 2012.
[6] Y.I. Shokin, O.A. Klimenko, E.V. Rychkova, and I.V.
Shabalnikov,“Website rating for scientific and research organizations of
the Siberian Branch of Russian Academy of Sciences”, Computational
Technologies. vol. 13(3), 2008, pp. 128-135, (in Russian).
[7] Y.I. Shokin, A.Y. Vesnin, A.A. Dobrynin, O.A. Klimenko,
E.V. Konstantinova, I.S. Petrov,and E.V. Rychkova, “Investigation of the
academic webspace of the Republic of Serbia”, In:
Zbornikradovakonferencije MIT2013,Belgrad, Serbia, 2014, pp. 601-607,
(in Russian)
[8] http://www.mit.rs/2013/zbornik-2013.pdf.
[9] Y.I. Shokin, A.Y.Vesnin, A.A.Dobrynin, O.A.Klimenko, E.V. Rychkova,
and I.S. Petrov,“Investigation of the academic webspace of the Siberian
Branch of the Russian Academy of Sciences”, Computational
Technologies, vol. 17(6), 2012, pp. 86-98, (in Russian).
[10] Y.I. Shokin, A.Y. Vesnin, A.A. Dobrynin, O.A. Klimenko, and
E.V. Rychkova,“Analysis of a web-space of academic communities by
method of webometrics and graph theory”, Information Technologies,vol.
12, 2014, 31-40, (in Russian).
[11] D. Stuart, M. Thelwall, and G. Harries,“UK academic web links and
collaboration – an exploratory study”,J.Inf Sci., vol. 33(2), 2007, pp.
231-246.
[12] M. Thelwall andD. Wilkinson,“Graph structure in three national academic
webs: power laws with anomalies”, Am. Soc. Inf. Sci. Technol.,vol. 54(8),
2003, pp. 706-712.
[13] A.Y. Vesnin, E.V. Konstantinova, and M.Y. Savin,“On scenarios of
attaching new websites to the webspace of SB RAS”, Vestnik NSU:
Information Technologies,vol. 11(4), 2013, pp. 28-37,(in Russian).
[14] M.E.J. Newman, A.L. Barabási, and D.J. Watts, The Structure and
Dynamics of Networks, Princeton Studies in Complexity, Princeton
University Press, 2006.
[15] M. Dehmer and F. Emmert-Streib,“Mining graph patterns in web-based
systems: aconceptual view”, In: A. Mehler,S. Sharoff, G.Rehm, and
M. Santini, Eds., Genres on the Web: Computational Models and
Empirical Studies. Berlin/New York: Springer, 2010. pp. 237-253.
[16] M. Dehmer and F. Emmert-Streib, Eds, Quantitative Graph Theory:
Mathematical Foundations and Applications, Discrete Mathematics and
Its Applications, Chapman and Hall/CRC, 2014.
[17] S. Fortunato,“Community detection in graphs”, Physics Reports,vol.
486(3), 2010, pp. 75-174.
[18] S. Wasserman and K. Faust, Network Analysis:Methods and
Applications, Cambridge: UK, Cambridge University Press, 1994.
[19] M.E.J. Newman,“Analysis of weighted networks”, Phys. Rev. E. vol.
70(5), 2004, 056131.
[20] U. Brandes, D. Delling, M. Gaertler, R. Goerke, M. Hoefer, Z. Nikoloski,
D. Wagner,“On modularity clustering”, IEEE Transactions on Knowledge
and Data Engineering,vol. 20(2), 2008, pp. 172-188.
[21] Hyperlinks between websites of institutes of scientific society
Fraunhofer-Gesellschaft, http://w.ict.nsc.ru/sitepage.php?PageID=1000.
[22] Rating of websites of scientific organizations of the Siberian Branch of
RAS,http://w.ict.nsc.ru/ranking/indexen.php?s_InfoID=15
[23] Project Ranking Web of World Research Centers,
http://research.webometrics.info
[24] Web search engine Yandex, http://www.yandex.ru
[25] Web search engine Google, http://www.google.ru
[26] Web search engine Bing, http://www.bing.com
[27] Indexing service of Google Scholar, http://scholar.google.com
[28] F. Harary, Graph Theory, Addison Wesley Publishing Company, 1969.
[29] F. Emmert-Streib andM. Dehmer,“Networks for systems biology:
conceptual connection of data and function”, IET Systems Biology,vol. 5,
2011, pp. 185-207.
[30] A. Mowshowitz andM. Dehmer,“Entropy and the complexity of graphs
revisited, Entropy, vol. 14(3), 2012, pp. 559-570.
[31] P. Hage and F.Harary, Structural Models in Anthropology,
CambridgeUK: Cambridge University Press, 1983.
[32] S. Gago, J.C. Hurajová, and T. Madaras,“Betweenness centrality of
graphs”, In: M. Dehmer and F.Emmert-Streib, Eds., Quantitative Graph
Theory: Mathematical Foundations and Applications, Discrete
Mathematics and Its Applications, Chapman and Hall/CRC,
2014,pp. 233-257.
[33] D. Watts andS. Strogatz,“Collective dynamics of small world networks”,
Nature, vol. 393, 1998, pp. 440-442.
[34] A. Arenas, A. Fernández, and S. Gómez,“Analysis of the structure of
complex networks at different resolution levels”, New Journal of
Physics,vol. 10(5), 2008, 053039.

Analysis of Webspaces of the Siberian Branch of the Russian Academy of Sciences and the Fraunhofer-Gesellschaft

More Related Content

What's hot (13)

Similar to Analysis of Webspaces of the Siberian Branch of the Russian Academy of Sciences and the Fraunhofer-Gesellschaft (20)

More from ITIIIndustries (20)

Recently uploaded (20)

Analysis of Webspaces of the Siberian Branch of the Russian Academy of Sciences and the Fraunhofer-Gesellschaft