What have we learned by applying social network analysis to the study of university industry relations?

Full text

3rd International Conference on Project Evaluation
ICOPEV 2016, Guimarães, Portugal
181
WHAT HAVE WE LEARNED BY APPLYING SOCIAL NETWORK
ANALYSIS TO THE STUDY OF UNIVERSITY INDUSTRY RELATIONS?
Fernando Romero1,2*, Eric Costa1
1Department of Production and Systems, University of Minho, Guimarães, Portugal
2ALGORITMI Research Centre, University of Minho, Guimarães, Portugal
*Corresponding author: fromero@dps.uminho.pt, University of Minho, Portugal
KEYWORDS
Innovation networks, university-industry relations,
social network analysis, review
ABSTRACT
The aim of this work is to give an overview on the
development of theoretical concepts and
methodological approaches to investigate innovation
networks, in particular the use of social network analysis
in the study of university industry relations. The
structure of networks can be analysed through the lens
of Social Network Analysis. This methodological
approach is described and its fundamental concepts are
presented. The paper then reviews the applications of
this approach on the study of university industry
relations. These relations can be considered as an
innovation network, in the sense that the interactions
established by its participants have more or less defined
innovation goals. Different structures in the relations
may result in different innovation outcomes, and the use
of SNA may be particularly useful to understand
differential outcomes. It is thus important to take stock
of the knowledge concerning the efforts that have been
made to probe the complex phenomena of university
industry relations and, in particular, how approaches
based on social network analysis have been used to
understand it. This work is based on a review of
available literature on the topics. The paper aims at
systematizing the information and knowledge related to
the application of SNA on university industry networks,
highlighting the main research pathways, the main
conclusions and pointing possible future research
questions.
INTRODUCTION
Social network analysis can bring many benefits for the
study of the relations between the university and the
industry. Relations between university and industry can
be considered as an innovation network, in the sense that
the interactions established by its participants have more
or less defined innovation goals (Mansfield and Lee
1996). Social network analysis is the study of social
structure (Wellman and Berkowitz 1988). Social
network analysis describes a group of quantitative
methods for analyzing the ties among social entities and
their implications (Wasserman and Faust 2007). An
important aspect in social network analysis is to identify
key players in a network (Borgatti 2003). Social
network analysis allows calculating measures and
drawing graphs that describe and illustrate the
individual and collective structure of a network.
The main measures calculated in SNA are cohesion
measures, centrality measures and subgroup measures.
Cohesion describes the interconnectedness of actors in a
network (Hawe et al. 2004). The main measure of
cohesion is the density of the network, which
corresponds to the total number of ties divided by the
total possible number of ties. Centrality measures
identify the most prominent actors, i.e. those extensively
involved in relationships with other network members
(Freeman 1979). The subgroup measures show how a
network can be partitioned in more or less independent
subsets.
With the use of social network analysis it is possible to
understand the different innovation outcomes in
university industry relations by analyzing the different
SNA measures and the structure of the social network.
SNA can be conducted to find the key elements in the
network that exhibit a wide range of connections
strength. The key elements can influence the network
structure and they play a significant role for affecting the
innovation networks developed between university and
industry (D'Este and Patel 2007). This paper makes a
review of the literature that has used social network
analysis to study university industry relations. The paper
aims at systematizing the information and knowledge
related to the application of SNA on university industry
networks, highlighting the main research approaches,
the main conclusions and pointing possible future
research questions. The following section presents the
research methodology. Sections 3 and 4 are the core of
the article, where the results of the literature review are
presented.
METHODOLOGY
The most important databases on scientific literature
were accessed and searched using a combination of

3rd International Conference on Project Evaluation
ICOPEV 2016, Guimarães, Portugal
182
relevant search strings. The accessed databases were
Web of Science, Scopus, JSTOR, Emerald, IEEE,
ABI/INFORM, EconLit, Academic Search, NBER, and
others. A selection of approximately 150 papers was
retrieved based on relevance, quality, non-redundancy
and impact criteria. A set of approximately 30 papers
was selected to write this review, based on the same
criteria and on subjective appreciations of their
contributions to knowledge and to the academic debate.
MOST COMMON NETWORK ANALYSIS
CONCEPTS USED IN STUDIES OF
UNIVERSITY-INDUSTRY RELATIONS
Social network analysis use concepts that are related to
the structural properties of the network and indicators
that are related to relational properties of the network.
The most used social network analysis concepts related
to structural properties of the network in studies on
university-industry relations are the concepts of density,
component, and subgroups. The most used social
network analysis concepts related to relational
properties of the network are the concepts of degree,
geodesic distance, centrality and betweenness centrality.
The concepts related to structural properties of the
network are basic and important concepts that
characterize the overall structure of the network, namely
in terms of its global cohesion (trough the concept of
density), in terms of its internal structure concerning the
existence of large groups inside the network (through
the concept of component) and in terms of smaller,
cohesive and more specifically defined subgroups (also
through the concept of component and, more rarely,
through the concept of clique). The combination of these
three indicators and an adequate interpretation of their
meaning provides useful descriptions and
characterizations of the network, in terms of the position
of their nodes and constituents. The characterization is
frequently complemented with visual aids, namely
through sociograms. The sociograms by themselves are
very useful in the overall characterization process of the
network. Several studies use exclusively the sociograms
to analyse the structure of the network, without
performing, or at least presenting, a formal numerical
analysis using the formal concepts of social network
analysis.
The concepts related to relational properties of the
network are often at the centre of the analytic procedure,
and are used in several ways according to specific
research objectives. The concepts of degree and
centrality is used to detect to what extent actors are
connected to other actors, and that of betweenness
centrality is used to characterize the intermediary
position of the actors in the network. Besides the main
concepts referred above, other concepts related to these
ones are also used, but less often. These include the
directional variants of degree centrality, the concepts of
in-degree or out-degree, other centrality measures such
as the closeness centrality, the eigenvector centrality
(which is an indicator of closeness centrality that
minimizes local conditions), and the concepts of direct
ties, indirect ties and valued ties. These more specific
and detailed concepts/indicators are rarer in the
literature that analyses university-industry relations.
MAIN THEMATIC APPROACHES
There is not a great number of articles that addresses
specifically the problem of university–industry relations
using SNA techniques. There is a variety of perspectives
that reflect specific and idiosyncratic concerns of the
authors. Apparently there are few papers that follow the
same guidelines or share identical perspectives.
However, there are small groups of authors that build on
past works or use identical databases, such as patent
databases.
The articles were classified in three main themes, in
terms of the main study object or main research
preoccupation or framework: 1) the study of the
characteristics of personnel/institutional networks that
are prominent in university-industry relations; these
studies generally rely on the use of patents that are co-
produced jointly by university and non-university
members, and the patterns of collaboration are analysed;
2) the study of university-industry relations in the
context of specific industrial settings or in the context of
specific institutional conditions; these studies may rely
also on patent databases but other types of data may be
used, either primary data, obtained through
questionnaires, or secondary data, obtained through
diverse documental sources; 3) the contribution of the
study of university-industry relations to the validation of
theories; these studies also rely on a mix of patent,
primary and secondary data.
In addition to these themes there are other themes that
are addressed in these studies, either in a parallel way or
as a theme that frames the former or the research
approach. These may include the search for an optimal
structure for innovation production and diffusion, the
validation of theories, the consideration of structural
properties of networks as independent or dependent
variables, the use of different methodological
perspectives and data sets or just the description of a
certain phenomenon or process.
The combination of these themes and subthemes
increases the content variety of the set of papers that
were reviewed. As a consequence, and as stated above,
the themes that could be common to the papers are, in
broad terms, the three main themes above indicated, but,
within each one, the approach and main research
concerns and targets are quite different. As such, the
literature will be analysed not only through the lens of
the broad themes, but also through the details of the
specific papers. This methodology will permit to extract
from the papers the main academic debates and to
highlight the respective contributions to knowledge. The
next sections will perform that task. Table 1 synthesises
the results.

3rd International Conference on Project Evaluation
ICOPEV 2016, Guimarães, Portugal
183
CHARACTERISTICS OF
PERSONNEL/INSTITUTIONAL UNIVERSITY-
INDUSTRY NETWORKS
Databases on scientific literature have been extensively
used to analyse the patterns of collaboration between
scientists. Patent databases are also being explored to
analyse the patterns of collaboration between academia
and industry.
The impact on fundamental research of an orientation to
patenting and commercialization has been researched
trough the relationship between patenting activity and
publication record of university researchers, and in
general the results point to a positive correlation
between patenting and publication activity (Czarnitzki
et al. 2009). This theme is revived with a social network
approach (Balconi and Laboranti 2006) and the results
support the positive relationship between publication
record and patenting activity. The author argues, in line
with other similar arguments (Rosenberg and Nelson
1994), that industry feeds on academic research but that
academic research also needs inputs from high
technology industries in order to find direction to its
research. So, academics that are close and collaborate
with industry producing patents are also the ones that are
more productive in purely scientific terms.
An exploratory analysis of the simultaneous
embeddedness of researchers in scientific and
technological networks (Breschi and Catalini 2010),
which compares networks of authors, inventors and
authors-inventors, and the overlap between them, argues
that author-inventors play a crucial role in connecting
the other two networks (only authors and only inventors)
and occupy important positions in each community, in
spite of the fact that maintaining a central position in one
community comes at the expense of being able to
occupy a similar position in the other community. The
role of academics as fundamental intermediaries
between public and private research is explored in a
study (Lissoni 2010) that founds that academic
inventors tend to be more central actors in broker and
gatekeeping positions, although strong brokerage
positions are very few and held by scientist with many
patents and publications. De Stefano and Zaccarin
(2013) reach similar conclusions regarding the larger
relational activity of academic authors-inventors vis-a-
vis industrial authors-inventors.
Two important differences were also apparent in
Balconi et al. (2004): academic inventors were more
connected than non-academic inventors (higher degree
values), and had a more central position (higher values
of betweenness). The central position of academics or of
the university is a characteristic that often shows up in
analysis of networks where public research
organisations are involved (Owen-Smith and Powell
2004, Balconi and Laboranti 2006, Breschi and Catalini
2010, Protogerou et al. 2013).
The main objective of Leydesdorff’ study (Leydesdorff
2004) is to reveal the knowledge base of patents and to
see how much innovation is really based on science.
This question is important because theories about
university-industry relations are historically influenced
by the biotechnology sector. The biotechnology sector
is a science-based one whose inventive activities tend to
be performed in close collaboration with public research
organizations and whose output is patented through co-
authorships or co-assignments between academic and
industrial inventors. The access to and the analysis of
patents databases have become easier and many studies
have thus relied on these data to infer general
conclusions to other fields of science, that are not so
formalized as the biotechnology sector in terms of
literature relations. The study analysis two sets of
patents, extracted from the USPTO, one based on
patents that have a university as a co-assignee, and
another that has a Dutch address as an assignee. The
structure of the co-words networks linking patents and
their citations to other patents and scientific literature is
analysed. The analysis is entirely based on the
visualization of sociograms, with nodes as (co)words.
The two networks are quite different. In the set of
university patents (which represents university-industry
relations) the fields of biotechnology and molecular
biology dominate the set and the knowledge base of the
patents, and the visualisation shows a neat organization
around the intellectual organization of the disciplines. In
the set of Dutch patents (representing the knowledge
base of the internationalized Dutch economy) the
visualization shows a recognizable representation of the
Dutch industrial structure with a dominance of electro-
technical and chemical applications and large
multinational corporations. Although biomedical
application integrates the patents they are not central to
the whole set. These results strongly suggest that
inferences of university-industry relations based on
literature and patent analyses are heavily conditioned by
the specificity of the biotechnology sector.
The question of the influence of the nature of the
relations on the performance of the network is a debated
issue addressed with social network analysis. The
concepts of strong and weak ties were introduced by
Granovetter (1973) and represent different forms of
social capital. Strong ties represents strong and regular
interactions and weak ties represent sporadic and
temporary interactions.

3rd International Conference on Project Evaluation
ICOPEV 2016, Guimarães, Portugal
Table 1: University-industry relations and social network analysis: main debates and conceptual propositions arising
from the literature review
Main concepts
References (authors, year)
Conceptual propositions proposed by
the literature
Patterns of university-
industry relations
(Leydesdorff 2004, Gilsing and
Duysters 2008, Krätke and
Brandt 2008)
Biotechnology has a specific pattern of
university-industry interaction, not
generalizable to other fields; patterns of
university-industry relations are
influenced by regional industrial
structures
Influence of commercial
orientation on fundamental
science production
(Balconi and Laboranti 2006)
Academics more connected to industry
are more productive in scientific terms
Strong and weak ties,
structural holes, social
capital
(Gilsing and Duysters 2008, Rost
2011, van der Valk et al. 2011,
Villanueva-Felez et al. 2013)
Balanced social structures (strong ties
with some weak ties) seem to be more
innovative; differential outcomes on the
nature of knowledge contingent on the
specific balance of the structure of
social capital
“Small worlds” networks
(Balconi et al. 2004, van der Valk
et al. 2011, Guan and Zhao 2013,
Protogerou et al. 2013)
Networks with high clustering and
short average geodesic paths are more
conducive to inventive or innovative
activity
Open-science and
proprietary technology
(Balconi et al. 2004, Owen-Smith
and Powell 2004)
The institutional attributes of open
science and proprietary technology
influences network structure; open
science networks are more connected
and dense than proprietary networks
that are more fragmented and disperse
Knowledge base or
environment as a relational
factor
(Owen-Smith et al. 2002,
Leydesdorff 2004, Gilsing and
Duysters 2008, Krätke and
Brandt 2008, Plum and Hassink
2011)
Different knowledge bases affect
network structural properties, the
position of individual entities in the
network and their capacity to access
knowledge
Public research
organizations as central
actors in innovation
networks
(Breschi and Catalini 2010,
Lissoni 2010, Minguillo and
Thelwall 2012, De Stefano and
Zaccarin 2013, Protogerou et al.
2013)
Academic authors-inventors assume
more brokerage positions; public
research organization are at the centre
of innovation programmes
New methodological
approaches
(Heimeriks et al. 2003, Kim
2012, Minguillo and Thelwall
2012)
Asides from patents indicators, other
indicators and data unmask
fundamental structural or relational
properties
Triple-helix theory
(Heimeriks et al. 2003, Khan and
Park 2013)
Triple helix assumptions on
institutional role intersections are
supported; multiple communication
channels with differential roles in the
Triple Helix relation
Industrial districts
(Morrison 2008, Capo-Vicedo et
al. 2013)
Public research organization as main
intermediaries of knowledge flows to
the district; weak knowledge exchanges
but strong information exchanges inside
the district actors

3rd International Conference on Project Evaluation
ICOPEV 2016, Guimarães, Portugal
Coleman (1988) claimed that cohesive groups and
strong ties were effective ways to coordinate an
exchange of knowledge flows, while Burt (1992) argued
that strong ties resulted in redundant information and
that innovation required new knowledge inflows and
perspectives coming from weak ties. Villanueva-Felez
et al. (2013) apply these concepts to assess in which way
the structure of researchers’ social capital affects
academic performance. The authors distinguish between
academics that are completely embedded in a network
that has no weak ties (establishing links with members
of his or her own department, without ties with
government, industrial, or other societal actors),
academics which are in a network that is formed
predominantly by weak ties and academics that are in an
integrated network that contains both strong and weak
ties. The results show that the academics in the network
with no weak ties are the less productive. On a study of
a network of inventors and on the assessment of the
impact of patents (based on forward citations) and
integration of knowledge (based on backward citations),
Rost (2011) concludes that inventors with balanced
social capital (strong ties but also some weak ties) come
up with the most innovative solutions, or integrate the
most knowledge or have the highest impact on future
knowledge. He concludes that Coleman’s and Burt’s
perspectives are complementary and that in the presence
of strong ties, weak network structures (structural holes
or a peripheral position) leverage the strength of strong
ties in the creation of innovation. Similar arguments are
advanced in a visual network analysis of two
government sponsored programmes that aimed to foster
innovation through public-private partnerships (van der
Valk et al. 2011) and also by other studies of university-
industry relations or industry networks (Ahuja 2000,
Gilsing and Duysters 2008)
UNIVERSITY-INDUSTRY RELATIONS AND
INSTITUTIONAL OR INDUSTRY CONDITIONS
The analyses of patent databases provides the basis for
the exploration of another important concept, which is
debated in multiple forms and in its multiple
consequences in studies of university-industry relations,
which is the distinction between the characteristics of
open science and proprietary technology (Merton 1957,
Cowan and Jonard 2003). The debate can be inserted in
a larger debate concerning the influence of diverse
institutional conditions on processes of relations
between organizational entities. Balconi et al. (2004)
conduct a study of Italian academic and industrial
inventors whereby, departing from assumptions on the
behaviour or characteristics of “open science networks”
and “proprietary networks”, expect to find differences
between the networks of academic and non-academic
inventors. In fact, the study found that networks of
industrial inventors are much more fragmented than
networks of academic inventors, except in the chemistry
field (defined in a broad sense, i.e. including
biotechnology). The chemistry sector, a science-based
field, was different because it was influenced by the
institutional weight of scientific inputs in commercial
technology.
The open science characteristics of scientific
communities translate, in social network terms, into the
so-called networks with “small worlds” characteristics
(Albert and Barabási 2002). The small world properties,
in the context of scientific networks in a specific
discipline, are defined by the existence of a large
component connecting almost all nodes, and within the
large component all nodes (scientists) are close to each
other (Newman 2001, Albert and Barabási 2002). These
characteristics of academic networks are not found in
networks of inventors, except in science-based fields.
These results are coherent with the results of
Leydesdorff (2004). The influence of small world
properties on innovativeness is addressed in the study of
industry networks (Verspagen and Duysters 2004) and
in university-industry networks (Guan and Zhao 2013),
and generally considered to be positive, although there
are disagreements concerning this positive influence
(Fleming et al. 2007).
Other articles support the importance of environmental
factors in shaping specific properties of networks. A
study of the Boston biotechnology sector (Owen-Smith
and Powell 2004) found that the information flows
between the actors of the network, which included firms
and public research organizations, depended not only on
network participation and position or geographic
proximity, but also on the institutional characteristics of
the network, that is, if the network was dominated by
public organizations, with an open science culture, or by
private entities, with a proprietary culture. In public-
dominated networks firm performance depended only
on net participation, unlike in networks dominated by
private entities, where innovative performance
depended on position factors, i.e., their closeness to
central actors (although this characteristic was weak in
terms of statistical significance).
An important determinant of cooperation between
university and industry, and an important factor in terms
of innovative performance, seems to be related to the
position of the firm in the network. That position may
be related to geography, in the sense that a firm that is
located in a densely populated region is positively
affected by the geography (Balconi and Laboranti 2006)
or that position may be related to the knowledge base
that the firm possesses and that may confer the firm the
possibility to connect with more or less central actors of
the network. A study of an industrial network in
Germany (Cantner and Graf 2006) argues that a
prerequisite for future cooperation is not based on past
cooperation but rather on a shared knowledge base. As
such, it questions ideas that argue that persistent
cooperation, based on trust, is necessarily the basis for
collaboration. Additionally it argues, based on
regression analysis, that job mobility of scientist and
engineers is a better predictor of relational structure than
past collaboration.

3rd International Conference on Project Evaluation
ICOPEV 2016, Guimarães, Portugal
186
In a study of two industrial networks (biotechnology and
multimedia) in a period that was characterised by
breaking with an existing dominant design and a shift
away from rules, norms, routines or activities, Gilsing
and Duysters (2008) argue that structural as well as
relational network properties are influenced by
environmental conditions. Environmental conditions
related to the different knowledge bases and the
validation and selection mechanisms inherent to each of
the two fields explain the relational and structural
properties of the two networks. For instance, the
connections of public research organizations are (again)
centrally present in the biotechnology field but absent in
the multimedia field (Gilsing and Duysters 2008).
Differences in the knowledge base show up as an
important factor in the determination of collaboration
structures in another study involving biotechnology
firms within a regional context (Plum and Hassink
2011). Besides indicating again the central position of
public research organizations in industrial
biotechnology networks, it points to differences related
to internal competencies of the firms regarding
differential capabilities in terms of the nature of the
knowledge required to develop the differential products
of each firm, in which the knowledge of the market also
has a role.
Although In a quite different perspective, a study of the
differences between the structures of two networks
emphasises the importance of environment in shaping
the properties of the network (Capellari and De Stefano
2014). Patents that are owned by the university (the
university is the assignee) or invented by the university
(the university is not the assignee but at least one of the
inventors is a tenured academic), are analysed
separately, showing differences in terms of size of
components, number and size of subgroups and the
brokerage position of inventors. The institutional factors
are mediated by two universities that have different
policies related to patenting ownership.
UNIVERSITY-INDUSTRY RELATIONS AND
THEORIES OF INNOVATION AND ECONOMIC
DEVELOPMENT
There is a strand of research of university-industry
relations using social network analysis methods that
adopt a deductive approach and try to validate some
relatively entrenched conceptual implications of some
theories.
One of the studies looks at the implications of the
industrial district approach. Morrison (2008), in her
study of the furniture sector in Italy, argues that the
community of informal ties inside the district appears to
be rather small and that ‘know how’ sharing is also
rather limited, contrary to assumptions from industrial
district theorists that based their ideas on the
development of these concentrated regions on intense
knowledge exchange between the actors. It, however,
supports the argument that public research
organizations, more than large firms, play a central role
as intermediaries in the knowledge flows for innovation
that occur in the industrial district, and that knowledge
for innovation does not arise only from close
interactions between the firms of the district, an idea that
is also supported by a study of a Spanish textile
industrial district (Capo-Vicedo et al. 2013).
The implications of the triple-helix approach are also
examined. Using webometric indicators and semantic
analysis of the contents of the webpages Kim (2012)
found that university and industry websites were similar,
thus suggesting there is an intersection or
interchangeability of the roles and function of the two
types of organizations, as suggested by the triple-helix
theory (Etzkowitz and Leydesdorff 1998). Diverse
channels of communication and relations between the
diverse institutional actors (co-authorship, participation
in projects, information diffusion) is also explored in
Heimeriks et al. (2003) which argue that each
communication channel or media has different
functional purposes in the maintenance of the links of
the triple-helix relation.
DESCRIPTIVE AND METHODOLOGICAL
CONTRIBUTIONS
The central position of public research organizations
shows up in descriptive analyses of networks that
involve heterogeneous actors. Both a study of the
network structure of science parks (Minguillo and
Thelwall 2012), using web links as indicators of
connections, and a study of the collaborative networks
established during the seventh Framework Programme
on Research and Technological Development of
European Commission, show the central position of
public research organizations. In the study of science
parks, governmental agencies also play an important
role, and in the case of the Framework Programmes,
although firms are present in larger numbers, they are
not the central actors.
Finally, there is a search for alternative methodological
approaches and indicators in the studies of networks of
university-industry relations. Some authors propose the
use of webometric approaches (Kim 2012, Minguillo
and Thelwall 2012) and other authors propose the use of
simultaneous indicators of relational characteristics,
such as citations, project participation, questionnaires or
other data (Heimeriks et al. 2003, Furukawa et al. 2011,
Almodovar and Teixeira 2014), arguing that analysis
based on a single indicator underestimate the level and
may not capture all of the complexities of the
collaboration patterns.
CONCLUSIONS
The use of social network analysis in the study of
university-industry relations was reviewed in this study.
There are not many studies that combine the two
perspectives and the ones that exist follow different
research objectives and concerns and different
methodological proposals. It seems evident that this

3rd International Conference on Project Evaluation
ICOPEV 2016, Guimarães, Portugal
187
particular knowledge quest is in a highly exploratory
phase. Nevertheless, the contributions to knowledge
have been varied and important, ranging from purely
descriptive studies and methodological explorations to
deductive testing of established theories. Some possible
research paths are open. Eventually, the use of more
complex and elaborated concepts of network analysis
could improve the analysis of data, it may have the
potential to reach different or stronger evidence and
conclusions and it may be an aspect that must be
improved. The diversity and plurality of university-
industry relations has not been properly addressed in the
literature, which tends to use patents as indicators of
collaboration. Environmental and institutional
influences of diverse sorts are clearly very important
factors that condition and determine university-industry
relations, and research is open to greater exploratory
efforts. There is a considerable potential to test
theoretical and conceptual propositions which are
assumed but have scarce empirical support.
REFERENCES
Ahuja, G. 2000. "Collaboration Networks, Structural
Holes, and Innovation: A Longitudinal Study."
Administrative Science Quarterly, 45(3): 425-455.
Albert, R. and A.-L. Barabási. 2002. "Statistical
Mechanics of Complex Networks." Reviews of
Modern Physics, 74(1): 47-97.
Almodovar, J. and A. A. C. Teixeira. 2014. "Assessing
the Importance of Local Supporting Organizations in
the Automotive Industry: A Hybrid Dynamic
Framework of Innovation Networks." European
Planning Studies, 22(4): 841-865.
Balconi, M., S. Breschi and F. Lissoni. 2004. "Networks
of Inventors and the Role of Academia: An
Exploration of Italian Patent Data." Research Policy,
33(1): 127-145.
Balconi, M. and A. Laboranti. 2006. "University–
Industry Interactions in Applied Research: The Case
of Microelectronics." Research Policy, 35(10):
1616-1630.
Borgatti, S. P. 2003. "Identifying Sets of Key Players in
a Network". In International Conference on
Integration of Knowledge Intensive Multi-Agent
Systems., Cambridge, MA, USA.
Breschi, S. and C. Catalini. 2010. "Tracing the Links
between Science and Technology: An Exploratory
Analysis of Scientists’ and Inventors’ Networks."
Research Policy, 39(1): 14-26.
Burt, R. S. 1992. Structural Holes. Harvard University
Press, Cambridge.
Cantner, U. and H. Graf. 2006. "The Network of
Innovators in Jena: An Application of Social
Network Analysis." Research Policy, 35(4): 463-
480.
Capellari, S. and D. De Stefano. 2014. "University-
Owned and University-Invented Patents: A Network
Analysis on Two Italian Universities."
Scientometrics, 99(2): 313-329.
Capo-Vicedo, J., F. X. Molina-Morales and J. Capo.
2013. "The Role of Universities in Making Industrial
Districts More Dynamic. A Case Study in Spain."
Higher Education, 65(4): 417-435.
Coleman, J. S. 1988. "Social Capital in the Creation of
Human Capital." American Journal of Sociology,
94(Supplement): S95-S120.
Cowan, R. and N. Jonard. 2003. "The Dynamics of
Collective Invention." Journal of Economic
Behavior & Organization, 52(4): 513-532.
Czarnitzki, D., W. Glänzel and K. Hussinger. 2009.
"Heterogeneity of Patenting Activity and Its
Implications for Scientific Research." Research
Policy, 38(1): 26-34.
D'Este, P. and P. Patel. 2007. "University-Industry
Linkages in the Uk: What Are the Factors
Underlying the Variety of Interactions with
Industry?" Research Policy, 36(9): 1295-1313.
De Stefano, D. and S. Zaccarin. 2013. "Modelling
Multiple Interactions in Science and Technology
Networks." Industry and Innovation, 20(3): 221-
240.
Etzkowitz, H. and L. Leydesdorff. 1998. "The Endless
Transition: A "Triple Helix" of University-Industry-
Government Relations." Minerva, 36(3): 203-208.
Fleming, L., C. King Iii and A. I. Juda. 2007. Small
Worlds and Regional Innovation. 18.
Freeman, L. 1979. "Centrality in Social Networks
Conceptual Clarification." Social networks, 1(3):
215-239.
Furukawa, T., N. Shirakawa and K. Okuwada. 2011.
"Quantitative Analysis of Collaborative and
Mobility Networks." Scientometrics, 87(3): 451-
466.
Gilsing, V. A. and G. M. Duysters. 2008.
"Understanding Novelty Creation in Exploration
Networks—Structural and Relational
Embeddedness Jointly Considered." Technovation,
28(10): 693-708.
Granovetter, M. 1973. "The Strenght of Weak Ties."
American Journal of Sociology, 78: 1360-1380.
Guan, J. C. and Q. J. Zhao. 2013. "The Impact of
University-Industry Collaboration Networks on
Innovation in Nanobiopharmaceuticals."
Technological Forecasting and Social Change,
80(7): 1271-1286.
Hawe, P., C. Webster and A. Shiell. 2004. "A Glossary
of Terms for Navigating the Field of Social Network
Analysis." Journal of Epidemiology and Community
Health, 58(12): 971-975.
Heimeriks, G., M. Hörlesberger and P. Van Den
Besselaar. 2003. "Mapping Communication and
Collaboration in Heterogeneous Research
Networks." Scientometrics, 58(2): 391-413.
Khan, G. F. and H. W. Park. 2013. "The E-Government
Research Domain: A Triple Helix Network Analysis
of Collaboration at the Regional, Country, and
Institutional Levels." Government Information
Quarterly, 30(2): 182-193.

3rd International Conference on Project Evaluation
ICOPEV 2016, Guimarães, Portugal
188
Kim, J. H. 2012. "A Hyperlink and Semantic Network
Analysis of the Triple Helix (University-
Government-Industry): The Interorganizational
Communication Structure of Nanotechnology."
Journal of Computer-Mediated Communication,
17(2): 152-170.
Krätke, S. and A. Brandt. 2008. "Knowledge Networks
as a Regional Development Resource: A Network
Analysis of the Interlinks between Scientific
Institutions and Regional Firms in the Metropolitan
Region of Hanover, Germany." European Planning
Studies, 17(1): 43-63.
Leydesdorff, L. 2004. "The University-Industry
Knowledge Relationship: Analyzing Patents and the
Science Base of Technologies." Journal of the
American Society for Information Science &
Technology, 55(11): 991-1001.
Lissoni, F. 2010. "Academic Inventors as Brokers."
Research Policy, 39(7): 843-857.
Mansfield, E. and J. Lee. 1996. "The Modern
University: Contributor to Industrial Innovation and
Recipient of Industrial R&D Support." Research
Policy, 25(7): 1057-1058.
Merton, R. K. 1957. "Priorities in Scientific Discovery:
A Chapter in the Sociology of Science." American
Sociological Review, 22(6): 635-659.
Minguillo, D. and M. Thelwall. 2012. "Mapping the
Network Structure of Science Parks an Exploratory
Study of Cross-Sectoral Interactions Reflected on
the Web." Aslib Proceedings, 64(4): 332-357.
Morrison, A. 2008. "Gatekeepers of Knowledge within
Industrial Districts: Who They Are, How They
Interact." Regional Studies, 42(6): 817-835.
Newman, M. E. J. 2001. "The Structure of Scientific
Collaboration Networks." Proceedings of the
National Academy of Sciences of the United States
of America, 98(2): 404-409.
Owen-Smith, J. and W. W. Powell. 2004. "Knowledge
Networks as Channels and Conduits: The Effects of
Spillovers in the Boston Biotechnology
Community." Organization Science, 15(1): 5-21.
Owen-Smith, J., M. Riccaboni, F. Pammolli and W. W.
Powell. 2002. "A Comparison of Us and European
University-Industry Relations in the Life Sciences."
Management Science, 48(1): 24-43.
Plum, O. and R. Hassink. 2011. "On the Nature and
Geography of Innovation and Interactive Learning:
A Case Study of the Biotechnology Industry in the
Aachen Technology Region, Germany." European
Planning Studies, 19(7): 1141-1163.
Protogerou, A., Y. Caloghirou and E. Siokas. 2013.
"Twenty-Five Years of Science-Industry
Collaboration: The Emergence and Evolution of
Policy-Driven Research Networks across Europe."
Journal of Technology Transfer, 38(6): 873-895.
Rosenberg, N. and R. R. Nelson. 1994. "American
Universities and Technical Advance in Industry."
Research Policy, 23(3): 323-348.
Rost, K. 2011. "The Strength of Strong Ties in the
Creation of Innovation." Research Policy, 40(4):
588-604.
van der Valk, T., M. M. H. Chappin and G. W. Gijsbers.
2011. "Evaluating Innovation Networks in Emerging
Technologies." Technological Forecasting and
Social Change, 78(1): 25-39.
Verspagen, B. and G. Duysters. 2004. "The Small
Worlds of Strategic Technology Alliances."
Technovation, 24(7): 563-571.
Villanueva-Felez, A., J. Molas-Gallart and A. Escriba-
Esteve. 2013. "Measuring Personal Networks and
Their Relationship with Scientific Production."
Minerva, 51(4): 465-483.
Wasserman, S. and K. Faust. 2007. Social Network
Analysis: Methods and Applications. Cambridge
University Press, Cambridge.
Wellman, B. and S. Berkowitz. 1988. Social Structures:
A Network Approach. Cambridge University Press,
Cambridge.