J. Serv. Sci. & Management, 2008, 1: 266-271
Published Online December 2008 in SciRes (www.SciRP.org/journal/jssm)
Analysis of University Science Research Capability
Elements and Evaluation based on BP Neural Network
Weiwei Liu & Chunsheng Shi
School of Management, Harbin Institute of Technology
Received October 28
th
, 2008; received December 5
th
, 2008; accepted December 14
th
, 2008.
ABSTRACT
After analyzing effect elements of University Science Research Capability (USRC) based on dynamic capability theory,
combined the substance of university science research with the highly self-organized, self-adapted and self-learned
characteristics of Back Propagation (BP) Neural Network, the paper conducts a research on evaluation of USRC, in
which an evaluation index system of USRC is constructed and a 15-7-1-typed BP Neural Network with three layers is
presented to evaluate USRC, which provides a BP Neural Network-based methodology for evaluation of USRC with
multiple inputs.
Keywords: BP neural network, science research, capability evaluation
1. Introduction
The capability of a university decides its competitive
advantage and management performance in the essence [1].
Therefore, it does make sense for universities to accumulate,
develop, evaluate and utilize their capabilities.
Nowadays, science research capacity is the core of
universities and an important indication of any powerful
university, all the universities tend to pay attention to
cultivate and enhance their capabilities, especially sci-
ence research capability. It’s necessary to analyze and
evaluate science research capability elements of universi-
ties. A number of researches have already been con-
ducted on evaluating competitive capabilities [2,3,4,5,6].
However, few researches have been taken on evaluating
scientific research capacity in universities. To fill in this
gap, the paper, first, investigates into the elements of
university science research capability based on dynamic
capability, and then, constructs BP Neural Network
model of evaluation, in the end, introduces into a simula-
tion evaluation, aiming at providing both theoretical and
empirical perspectives in the cultivation of university
science research capability.
2. Analysis of USRC Based on the Theory of
Dynamic Capability
2.1. The Theory of Dynamic Capability
The theory of dynamic capability is proposed firstly by
Teece, Pisano and Shuen in “Firm Capability, Resource
and Strategic concept” [7]. The theory of dynamic capa-
bility develops and consummate gradually in “Dynamic
Capabilities and Strategic Management” by Teece, Pis-
ano and Shuen [8]. They defined dynamic capability as
the capability of conforming, constructing and reconfig-
uring inner and exterior capabilities to adapt environment
changed rapidly. The definition has two outline: “dy-
namic” namely, enterprises must renovate capabilities of
themselves to adapt changeful environment; “capability”
namely, strategic management has key function in reno-
vating capabilities of themselves to adapt changeful en-
vironment. Some scholars consider that capability can be
defined as the gather of enterprise knowledge and capa-
bility that can change capability is technology knowledge
[9].
2.2. Analysis of USRC Based on Characteristics
of Dynamic Capability
The theory of dynamic capability comes from the theory
of resource base and absorbs many viewpoints of the
theory of core capability, thus, its characteristic is similar
with the theory of core capability, for example, the nature
of value, the nature of unique. Nevertheless, dynamic
capability is the capability that changes capabilities; its
difference in nature from core capability is that it’s the
nature of deploitation [10]. Since science research activi-
ties of universities can’t depart from resource base, the
products of universities science research activities has the
value of applying and generalizing, science research char-
acteristics of each universities are different from others
and different from the advantage of science research
competition rest with the nature of unique. Universities
should renovate constantly their science research capa-

Analysis of University Science Research Capability Elements and Evaluation based on BP Neural Network 267
bilities for adapting the changes of environment and
keeping the advantage of competition.
The theory of dynamic capability is production of
conforming inside knowledge and absorbing outside
knowledge [11]. Absorbing knowledge has bridge function
in resource and capabilities between inside and outside,
thus, the theory of dynamic capability emphasizes to
construct special capabilities of absorbing knowledge
from outside. Universities must enhance learning inter-
course between internal and overseas to heighten science
research capabilities. Universities can absorb advanced
knowledge from outside and master international learning
development trends. Universities should increase the op-
portunity of studying abroad and visiting, besides, should
introduce into visiting professors. The communion of
experiences is good for importing advanced techniques
and methods.
2.3. Analysis of University Science Research Ca-
pability Effect Elements
Competitive advantages of universities come from sci-
ence research capability. The elements of university sci-
ence research capability can be elaborated from four as-
pects: science research input, transformation efficiency
and science research output and science research man-
agement. Combined with the features of university de-
velopment, science research capability can be classified
into four aspects, i.e., science research input capability,
transformation efficiency capability and science research
output capability and science research management, sci-
ence research input capability effect university science
research capability, science research output capability
reflect university science research capability and trans-
formation efficiency capability effect university science
research capability indirectly, a sound management
mechanism makes effect elements exert supreme effi-
ciency and makes the whole benefit exceed the summa-
tion of parts. Science research management is holistic
macroscopically elements of effecting USRC, as shown
in Figure 1.
Science research input capability and science research
output capability affect each other in science research
process. Science research input capability provides sub-
stance base and intellect sustainment for science research
output capability. Science research output capability act
on science research input capability and provides reliable
basis for science research input capability. Transforma-
tion efficiency capability is used for transforming be-
tween science research input and science research output
effectively.
3. Comparison and Selection of Evaluation
Methods
As to the evaluation of USRC, such methods as Analytic
Hierarchy Process (AHP), Fuzzy Comprehensive Evaluation
(FCE) and Data Envelopment Analysis (DEA) have been
adopted with certain effects [12,13,14].
3.1. Analytic Hierarchy Process (AHP)
AHP is proposed by American T. L. Satty in 1970s
[15]
, its
essential principle is that setting down evaluation scheme
on the basis of objects with gradual rank, sub-objects,
restriction condition and adopting the method of com-
paring between one and the other to make sure judgment
matrix, and then, making proportion vectors of character-
istic vector that maximum characteristic of judgment ma-
trix corresponding to as coefficient, in the end, presenting
the power weight of each schemes synthetically. The
method is an improvement of the method of adding
power average, however, when evaluation scale is very
large and there are too many evaluation indexes, the
method prefer appearing problems, for example, judg-
ment matrix is difficult with meeting the requirement of
conformity. USRC there are too many evaluation indexes,
therefore, it can result with the difficulty in meeting the
requirement of conformity by AHP.
3.2. Fuzzy Comprehensive Evaluation (FCE)
Fuzzy Comprehensive Evaluation (FCE) is proposed by
Peizhuang Wang [16]. It has two steps: individual
evaluation according to each factor; integrative evaluation
Figure 1. University scientific research capability elements system
Science
research
management

268 Weiwei Liu & Chunsheng Shi
according to all the factors. Its primary function is that
quantificational evaluation to qualitative indexes. However,
it has two shortages: firstly, it’s difficult to compartmentalize
the criterion of evaluation, for example: the limit between
“very good” and “good” is difficult to ascertain, there are
no specific criterions; secondly, the quality of expert
mark can’t be ensured. Because of intense subjectivity,
FCE can’t be used for evaluating USRC.
3.3. Data Envelopment Analysis (DEA)
DEA is a systems analysis evaluation method established
on the basis of relative efficiency evaluation concept and
given by A. Charnes and W. W. Cooper [17]. It supposes
there are n decision making units (DMU), every DMU
has m kinds of input and s kinds of output. When evalu-
ating USRC, determinate selection rule should be trans-
formed into data envelopment analysis model. Firstly,
compartmentalize selection rule as input variable and
output variable, and then, establish data envelopment
analysis model and calculate relative efficiency of every
university, in the end, the place of USRC can be obtained
on the basis of calculation results. However, the method
has relative validity because of being constructed on the
basis of comparing each other. The evaluation of USRC
isn’t only used for getting horizontal comparison of many
universities, but also getting vertical dynamic development
trends. Because of disadvantages of DEA in relativity,
the evaluation of USRC can’t adopt DEA.
3.4. BP Neural Network
Among these methods, DEA is limited to the evaluation
of ascertained performance indexes, AHP has restrict of
evaluation scales, while FCE shows subjectivity in such
aspects as data collection and index weight determination.
Meantime, all these methods are based on the assumption
of linear relations between indexes and can only be
adopted within limits.
Characterized as self-adaptability, self-learning and
large-scale parallel calculation ability, BP Neural Net-
work is the most commonly used network model [18] in
such aspects as identification, classification, evaluation,
forecast, nonlinear mapping and simulating complex sys-
tem [19]. BP Neural Network is a multiple-layer feedback
network featured in back propagation of errors and con-
stituted of input, hidden and output layers.
BP Neural Network is the simplification, abstraction
and simulation of brain function. And it is a highly so-
phisticated and nonlinear dynamic system, in which
nonlinear mapping is adopted to find out inner relations
according to the existing training-set and through learn-
ing and training. Without the process of subjective calcu-
lation of index weight, the evaluation model proposed in
the paper decreases the subjectivity and overcomes the
possible linear problem between indexes effectively,
which enhances the reliability and adaptability of evalua-
tion for USRC.
4. Establishment of USRC Evaluation Model
Based on BP Neural Network
4.1. Establishment of the Index System of USRC
Evaluation
Competitive advantages of universities come from sci-
ence research capability. A sound evaluation index sys-
tem is the basis of effective evaluation of USRC. Ac-
cording to the establishment principles of index system
[20] and based on many interrelated reference literatures
[21,22,23,24], this paper selects science research input
capability, transformation efficiency capability and sci-
ence research input capability as the First-order indexes
and further proposes the index system to evaluate syn-
thetically USRC, as shown in Table 1.
Table 1. Evaluation index system of USRC
Object Layer First-order indexes Second -order indexes
Science research expense (U
1
)
Science research procession (U
2
)
Subject and science research base (U
3
)
Science research input capability
Science research equipment and condition (U
4
)
Personnel opening degree (U
5
)
Learning intercommunion (U
6
)
The amount of project (U
7
)
Transformation efficiency capability
Cooperation capability(U
8
)
Science research production (U
9
)
Production application (U
10
)
Student cultivation (U
11
)
Science research output capability
Project finishing status (U
12
)
Science research input management capability(U
13
)
Transformation efficiency management capability(U
14
)
USRC
Science research management capability Science research output management capability(U
15
)

Analysis of University Science Research Capability Elements and Evaluation based on BP Neural Network 269
4.2. Pre-treatment of evaluation index data
Evaluation index system consists of quantitative and
qualitative indexes. Quantitative indexes can be further
divided into positive, reverse and moderate indexes [25].
Considering the difficulties in comprehensive evaluation
directly based on the data and the convergence problem
of BP Neural Network, normalized and trend-assimilated
pre-treatment is needed first. For qualitative indexes, we
can make sure them by the method evaluation grade sub-
jection degree, for example, 1, 0.75, 0.5, 0.25, 0 corre-
spond respectively very good, good, common, bad, and
very bad. For quantitative indexes, there are three steps:
(1) Calculate the average value
j
P of the jth test index
j
P ,
j
P = mkj
n
x
n
iij
,,2,1,
)(
1
=
=
∑
(1)
(2) Calculate middle variable of index
ij
M:
If bigger target values indicate better effects, i.e. bene-
fit index,
ij
M=
j
jij
P
Px −
(2)
If smaller target values indicate better effects, i.e. cost
index,
j
ijj
ij
P
xp
M
−
= (3)
If fixed scope values indicate better effects, i.e. scope
index,
when
,Ax
ij
≤A
Ax
M
ij
ij
−
= (4)
when
, Bx
ij
≥
B
xB
M
ij
ij
−
= (5)
when
, AxB
ij
≥≥
ij
M=6 (6)
In Function (4), (5) and (6), A and B are best top
boundary and best bottom boundary of scope indexes,
ij
M reflects the degree
ij
x deflects to average.
Calculate subjection degree function value
ij
Y of in-
dexes transforming to [-1, 1]
ij
ij
M
M
e
e
Y
−
−
+
−
=1
1
(7)
4.3. Design of Physical Structure of USRC
Evaluation Model
The selection of network model structure is very impor-
tant. Good selection can reduce the network training times
and raise the network learning precision. The selection
procedure includes:
(1) Ascertain the layer number of BP Neural Network.
Theorem A. H. Kolmogrov has proven that a nonlinear
neural network with three layers and S-typed action func-
tion will approach any continuous function with any pre-
cision as long as it possesses enough hidden nodes [26].
As a result, the paper adopts three-layer BP Neural Net-
work in constructing the evaluation model of USRC, as
shown in Figure 2. As to the function between nodes, S
(Sigmoid)-typed function is introduced, as shown in
Function (8).
x
e
xf
−
+
=1
1
)( (8)
(2) Ascertain the numbers of neural cell nodes both in
the input layer and output layer of BP Neural Network.
The numbers are dependent on outer description. The
node number in the input layer is the number of charac-
teristic factors (independent variable) while that in the
output layer is the number of targets in the system. As a
result, we can conclude that the input-layer node number
is 15 and the output-layer is one, according to the index
system of USRC evaluation as shown in Table 1.
(3) Ascertain the node number in the hidden layer of
BP Neural Network. The node number in the hidden
layer has a direct relation with the requirements and the
node numbers both in the input and output layers. Too
few nodes will result in the failure in network training or
in the identification of unseen set before, while, too many
ones will result in an excessively long learning period
and an unobvious reduction in errors. Therefore, an op-
timal node number must be existent. Up to the present, no
sound method has been found to ascertain the number.
On the basis of previous researches on BP Neural
Figure 2. BP Neural network model structure
I
1
I
2
I
3
I
4
I
5
O
1
Input layer
Hidden layer
Output layer

270 Weiwei Liu & Chunsheng Shi
Network, two empirical formulas are concluded as follows:
51
.035.054.277.043.012.0
2
+++++= nmmnm λ
(9)
Anm ++= λ (10)
In Function (9) and (10), m indicates the node number
in the input layer, while n that in the output layer. A indi-
cates any integer between one to ten. λ is the ascertained
node number in the hidden layer.
According to Function (9) and (10) and the effective-
ness of models, the node number in the hidden layer is
assumed seven.
4.4. The Selection of Logic Algorithm in the
Evaluation Model of USRC
The BP algorithm put forward by Rumelhart provides
neural network with a more practical and effective train-
ing method. However, the algorithm has limitations when
applied in complex systems for the high nonlinearity re-
sults in the low efficiency and slow convergence of the
initial BP algorithm. In order to overcome these obstacles,
the paper adopts a refined algorithm, a combination of
additional momentum and Adaptive Learning Rate
Method.
According to the BP Algorithm and the Minimum
Mean-square Error, we can conclude the connection weight
adjusting function of additional momentum factors:
)()1()1( kwxkw
ijjiij
∆+−=+∆ αηδα (11)
)()1()1( kbkb
iii
∆+−=+∆ αηδα (12)
In Function (11) and (12), k indicates the training number.
∆
ij
w and ∆
i
b indicate the increments of weight.
η
indi-
cates the learning rate.
i
δ is error.
j
x is network input.
α
is momentum factor and the average value is 0.9.
The value of
η
decides the success of algorithm to
some extent. Large value results in the oscillation of error
function and small value the slowness of convergence. In
order to solve this, the method of adapting learning rate is
adopted in the network training, i. e. Adaptive Learning
Rate Method and the adapted function is concluded as fol-
lows [27]:





>+
<+
=+
elsek
kerfkerfk
kerfkerfk
k
),(
)(04.1)1(),(7.0
)()1(),(05.1
)1(
η
η
η
η
(13)
∑
=
−=
N
iii
OTerf
1
2
)( (14)
In Function (13) and (14), erf is the error function. T
i
is the expected output value of learning set.
i
O is the real
output value of learning set. N is the number of learning
sets.
5. Empirical Research of Model
According to the evaluation index system above, the pa-
per selects 12 groups of samples and divides into two
parts. The former 8 groups are adopted in network train-
ing as learning sets and the latter 4 groups in network test.
Besides, the paper selects 15 second-order inputting in-
dexes and adopts the module of artificial neural network
in MATLAB software. After the learning process of the
evaluation of USRC, the BP Neural Network-based
evaluation model of USRC is concluded. This process
consists of the following steps:
(1) Select 12 groups of evaluation indexes as learning
sets. The input values are the attribute values of 15 in-
dexes after pre-treatment and the corresponding output
value expected is the evaluated value of each sample.
(2) Start up the learning process after the normalization
of index attributes and iterate the process till convergence.
In this process, the network toolbox of MATLAB7.0 is
used, with a learning rate of 0.01, a momentum factor of
0.9, a maximum training step of 2000 and an error preci-
sion of 0.0001. During the training, some obviously ab-
normal data have been eliminated through the displaying
function of variance chart in MATLAB and the learning
results have been achieved, as shown in Table 2.
(3) Restore the trained parameters of the BP Neural
Network with three layers into the corresponding knowl-
edge database and input respectively 4 groups of validat-
ing data to further examine the network model. The com-
prehensive examination result is comparatively ideal,
which validates the effectiveness of the model, as shown
in Table 3.
The simulation evaluation result with test samples is
the same as experts’ evaluation result basically. The ap-
plication example indicates that the network has com-
prehensive practicability. The model can master expert
knowledge by sample learning and be used evaluation
with many indexes.
Table 2. The results of learning
Sample
code Training value
Expected
value Relative error (%)
1 0.556 0.549 1.28
2 0.117 0.119 -1.68
3 0.547 0.551 -0.73
4 0.423 0.421 0.48
5 0.113 0.116 -2.59
6 0.368 0.365 0.82
7 0.649 0.651 -0.31
8 0.235 0.227 3.52
Table 3. The examination results
Sample
code Training value Expected value Relative error (%)
Simulation
place Expert place
9 0.225 0.227 -0.88 4 4
10 0.345 0.343 0.58 3 3
11 0.435 0.433 -0.46 2 2
12 0.556 0.555 0.18 1 1

Analysis of University Science Research Capability Elements and Evaluation based on BP Neural Network 271
6. Conclusions
This paper constructs an evaluation index system of
USRC combined with characteristics of universities
based on dynamic capability theory, after analyzing ef-
fect elements of USRC, the paper presents a 15 - 7
-1-typed BP Neural Network with three layers to evalu-
ate USRC on the basis of the highly self-organized,
self-adapted and self-learned BP Neural Network com-
prehensive evaluation method. The method constructs a
comprehensive evaluation model combined with both
quantitative and qualitative indexes which is close to
human being thought mode better. The satisfying result is
obtained by emulational test. Its advantages embodies in
the following aspects: avoiding the effects of subjectivity
and randomicity in traditional evaluating methods and
ensure the preciseness and objectivity of results; accord-
ing with the empirical situation along with the increasing
number of training samples; and overcoming the possible
linear problems among the indexes and enhancing the
reliability and adaptability of evaluation. Therefore,
compared with the traditional evaluating methods of
USRC, the one based on BP Neural Network is of better
practicability. BP Neural Network can learn by random
sample parameter and construct diverse evaluation model.
It can get reliable evaluation result on the basis of practi-
cal test sample after learning successfully; meanwhile
analysis result will be accurate and factual when training
samples increase gradually, therefore, the method has
more comprehensive applicability.
REFERENCES
[1] V. M. Giménez and J. L. Martínez, “Cost efficiency in the
university: A departmental evaluation model,” [J].
Economics of Education Review, Vol. 25, pp. 543-553,
2006.
[2] P. Boccardelli and M. G. Magnusson, “Dynamic capabilities
in early-phase entrepreneurship,” [J]. Knowledge and Process
Management, Vol. 13, pp. 162-174, 2006.
[3] D. J. Teece, “Explicating dynamic capabilities: The nature
and microfoundations of (sustainable) enterprise performance,”
[J]. Strategic Management Journal, Vol. 28, pp. 1319-1350,
2007.
[4] C. E. Helfat and M. A. Peteraf, “The dynamic
resource-based view: Capability lifecycles,” [J]. Strategic
Management Journal, Vol. 24, pp. 997-1010, 2003.
[5] B. McEvily and A. Marcus, “Embedded ties and the
acquisition of competitive capabilities,” [J]. Strategic
Management Journal, Vol. 26, pp. 1033-1055, 2005.
[6] D. Hussey, “Company analysis: Determining strategic
capability,” [J]. Strategic Change, Vol. 11, pp. 43-52, 2002.
[7] G. S. Wang, “Enterprise theory: capability theory [M],”
China Economic Publishing House, PP. 69-70, 2006.
[8] Teece, Pisano, and Shuen, “Dynamic capabilities and
strategic management”[J], Strategic Management Journal,
18(7): pp. 509-533, 1997.
[9] J. W. Dong, J. Z . Huang, and Z. H. Chen, “An
intellectual model of the evolution of dynamic capabilities;
and a Relevant Case Study of a Chinese Enterprise” [J],
Management Word, (4): pp. 117-127, 2004.
[10] J. Z. Huang and L. W. Tan, “From capability to dynamic
capability: transformation of the points of stratagem” [J],
Economy Management, (22): pp. 13-17, 2002.
[11] K. M. Eisenhards and Martin, “Dynamic capabilities:
what are they,” [J], Journal of Evolutionary Economics,
(3): 519-523, 2000.
[12] T. M. Crimmins, J. E. de Steiguer, and Donald Dennis,
“AHP as a means for improving public participation: A
pre-post experiment with university students,” [J]. Forest
Policy and Economics, Vol. 7, pp. 501-514, May 2005.
[13] A. Golec and E. Kahya, “A fuzzy model for competency-based
employee evaluation and selection,” [J]. Computers &
Industrial Engineering, Vol. 52, pp. 143-161, February
2007.
[14] J. C. Glassa, G. McCallionb, and D. G. McKillopc, “Implications
of variant efficiency measures for policy evaluations in
UK higher education,” [J]. Socio-Economic Planning
Sciences, Vol. 40, Issue 2, pp. 119-142, June 2006.
[15] L. Bodin and S. I. Gass, “On teaching the analytic hierarchy
process,” [J]. Computers & Operations research, pp.
1487-1497, 2003.
[16] D. Du, Q. H. Pang, and Y. Wu, “Modern synthesis
evaluation methods and cases choiceness,” [M], TSINGHUA
University Press, pp. 34-40, 2008.
[17] H. Hao and J. F. Zhong, “System analysis and evaluation
methods,” [M], Economic Science Press, pp. 185-239.
2007.
[18] S. L. Sun, “The application study of supplier evaluation
model in supply chain,”[J], Modernization of
Management, Vol. 1, pp. 62-64, 2007.
[19] F. S. Wang and S. Z. Peng,“Study on the measurement of the
activity-based value output of industry value chain based
on BP neural network,” [J], Journal of Harbin Institute of
Technology, Vol. 1, pp. 91-93, 2006.
[20] Y. J. Guo, “Theory, method and application of comprehensive
evaluation,” [M]. Bei Jing: Science Press, pp. 14-29,
2007.
[21] W. Z. Zhu, “The establishment of systematized evaluation
indexes to university science research capacity,” [J].
Journal of Anhui University of Technology and Science.
Vol. 18, pp. 40-44, 2003.
[22] J. Q. Cai and L. M.Cheng, “Research on evaluation of
university science research capability”[J]. R & D
Management, Vol. 10, pp. 76-80, 1998.
[23] X. Min, C. R. Dai, and H. Bin, “Research on evaluation of
university science research capability based on theory of
fuzzy mathematics,” [J]. Science and Technology Management
Research. Vol. 8, pp. 185-187, 2006.
[24] B. Liu and X. L. Wang,“Research on evaluation methon of
university sceince research capability,” [J] Science of Science
and Management of S. & T. Vol. 12, pp. 83-85, 2003.
[25] Y. H. Li and Y. Q. Hu, “Comprehensive evaluation based
on bp artificial neural network on enterprises core
competence in the old industrial base,” [J], Commercial
Research, Vol. 5, pp. 34-36, 2006.
[26] L. C. Jiao, “Systemic theory of neural network,” [M]. Xi
An: Xidian University Press, 1996.
[27] Y. F. Wu and Z. Z. Li, “Application of improved BP
nerve network model in forecasting dam safety monitor-
ing values,” [J], Design of Hydroelectric Power Station,
Vol. 2, pp. 21-24, 2002.