1 Introduction
The feature of the world economy has transformed from large-scaled industrial production to new high tech like the information technology, and the industrial convergence mainly promoted by the information technology has become the new growth point of the world economy. Many countries provide support through legislation for the network convergence, and a series of policies have been put forward to encourage technological innovation and market competition. After more than a decade’s development, a new convergence competition pattern forms. Convergence of the telecommunication network, the internet and the radio and television network has been basically implemented in some countries, which greatly promoted the development of information industry, and played a significant role in pushing the economic growth and promoting the social progress.
The development of network convergence fits to certain rules. The evolution of network convergence can be divided into four stages: Nicholas Negroponte (1973) [
1] described the marginal overlay phenomenon of the computer industry, radio and movie industry, and the printing and publishing industry by using three overlapping circles. He indicated that the overlapping area is the field with the fastest developing speed and the brightest future vision. It is the first stage. In the 1980s, of the second stage, the concept of convergence of computer network, radio and television network, and telecommunication network was proposed. In the 1990s, of the third stage, network convergence referred to telecommunication network, radio and television network, and the Internet. With more and more attention being paid on the media prosperity of communication, network convergence refers to the convergence of telecommunication network, the Internet, and media network. The popularity of the mobile Internet and the stepping of the Internet of things from concept phase to implementation phase constantly expand the scope of the network, while continuing to promote the convergence of the network. Network convergence on the physical level refers to the transmission of voice, data, and images on the same network. For users, it means to realize the function of calling, watching TV, surfing the internet, and so on through the same access.
At present, the network convergence in China refers to that of the digital Internet represented by the Internet, the telecommunication network represented by the telephone network including the mobile communication network, and the radio and television network represented by the cable television network. The networks represent three different industries in the modern information industry, namely, the basic facility in the telecommunication industry, the internet industry, and the cable television industry. The current the network convergence does not mean the physical combination of the telecommunication network, the computer network, and the cable television network, but the application integration in business application. Technically, on the network level, the realization of mutual communication forms a seamless coverage. On the business level, mutual infiltration and crossing can be practiced. On the application level, the unified IP agreement can be used gradually. On the operation level, mutual competition and cooperation can be developed and crossed towards the same aim of providing diversified, mediatized, and personalized services. Finally, the industrial management and the policies can be gradually unified.
Traditionally, communication and media are completely different. However, with the development of technology, the boundaries between them are gradually becoming fussy. The communication becomes more important as a media method and the modern communication technology becomes a base of the new media, so the media prosperity of communication becomes increasingly emphasized.
In recent years, new media represented by the Internet develops fast. New media mainly refers to the media form which appears with the support system of new technologies. In addition to the Internet, it also includes digital magazine, mobile phone SMS, and the touch media. [
2]. New media is completely different with the traditional one in form. It submerses the communication studies in the traditional meaning. The American magazine
Wired gave a vivid definition of the new media as the transmission from all people to all people.
New media grows extremely fast from its production to its development. According to the definition given by an American transmission scholar that only when the using population of one media reaches the one fifth of the whole population around the country can it be referred to the public media. In America, the usage period of public media that reached the standard including the radio, the television (wireless), and the cable television is, respectively, 38, 13, and 10 years, but that of the Internet only takes 5 years. In China, according to the report data of iResearch Consulting, the new media accounted for 35% of the media industry in 2012, in which the proportion growth compared to the same period of the network media including the network game and the network advertisement is 72.5%. The growth speed is far more than that of the traditional media of 13.6%, and the industrial scale of the new media industry expands quickly.
With the gradual deepening of the influence produced by the new media on the mainstream media, the network convergence can be promoted to a new stage until that of N networks [
3]. The network convergence in the next stage may be practiced by taking the production, processing, transferring, and storage of information as the core. Through that, all kind of information including the data, video, and message can reach to any consumers through any physical network. Any legitimate information can get through the network easily. Data can be transferred, stored, and processed on the network. The intelligence trend of the network and the toolized degree become much higher; thus, the media prosperity of the communication network can be adequately represented. However, the N network convergence is the extension and development trend of the network convergence, which can make anyone enjoy any information service through any physical network. It would be the network that enables everyone’s compound roles as the information producer, the propagator, and the consumer to be maximally realized.
From one network convergence to N network convergence, it is actually the direction of Ubiquity information development [
4]. The ubiquity information refers to the intelligent comprehensive network information society, also the ubiquitous network society. In 1994, Japanese scholar Tadao Umesao proposed the concept of information society [
5]. With the development of almost half a century, the information society is developed towards the direction of the ubiquitous intelligent comprehensive network information society, namely, the U information direction.
Currently, the technologies represented by the Bluetooth, RFID, WiFi, 4G, and the Internet of things are mutually converged with the ADSL and the FTTH networks [
6]. Users can obtain information in the forms of text, voice, picture, and so on through the ubiquitous networks including the fixed telephone, the mobile telephone, the television, the computer, and diversified informational zed terminal equipment. U information direction is the trend of future network development, which is the development direction of the ubiquitous intelligent comprehensive network information society. Its physical representation is the appearance of ubiquitous network or the named universal network. It represents the ubiquitous prosperity that all people, all organizations, and industries can get benefits from the U information direction.
As Internet applications become more sensitive to latency time, the needs for network management increase. Different applications need different levels of QoS for Internet access services decided by consumer preferences [
7]. Solutions for future mobile front haul networks, as well as those for future convergence of the mobile front haul with optical broadband access networks, are frequently discussed [
8]. The concept of network convergence for providing the overall architectural framework to bring together all the different technologies within a unifying and coherent network ecosystem has been proposed in the background of future 5G services vision [
9].
Network convergence is the necessary trend for the development of the global information, with the basic reason as the progress of information technology and the pursuing for network scope economy from the network industry. This article studies the evolution of internal and external structures of network service model under the background of network convergence and introduces technology dynamics, demand dynamics, competition, and security as adjustment variables. Then, the evaluation model of network service based on network convergence is proposed and the test is implemented to prove the validity of the evaluation model.
The paper applies the analysis technology of structural equation model (SEM) to test the hypotheses in the research. Structural equation model invokes a measurement model that defines latent variables using one or more observed variables and a structural model that imputes relationships between latent variables. The model consists of latent variables, measured variables, and a path. Latent variables cannot be directly observed but are rather inferred from other variables. Measured variables (manifest variable) can be directly measured and are usually used to explain latent variables [
10].
In the analysis process of SEM, the first problem to be judged is the validity of the measurement model, meaning whether the observation variables can be measured to define the corresponding latent variables; otherwise, the analyzed result through the SEM method would be invalid. Therefore, the trust and efficiency evaluation problem of the measurement table are involved.
2.1 Reliability test
The measurement model part in the SEM model includes the latent variables and several observable variables used to test the latent variables (shown variables). The internal consistency between the tested items is an index to be measured. To carry out the trust test for the variables is a significant aspect in the evaluation of the SEM model.
To make the reliability test on the variables is an important aspect in the evaluation of the SEM model. Cronbach’s alpha is used often to test to what degree can the observed variables belong to the same group [
11]. Therefore, this paper first carried out the reliability test on the questionnaire, meaning to verify the internal consistency between the tested items to evaluate the reliability of the measurement table through adopting the method of Cronbach’s alpha. The coefficients are used as the trust evaluation indexes, and the software of SPSS is used to finish the reliability test of the measurement table.
2.2 Factor analysis
The principle of the factor analysis can be explained as follows. In a research, some unrelated comprehensive indexes may be used to analyze each kind of information in the variables. These comprehensive indexes are called factors. The factor analysis is to describe the relationship between many indexes or elements by several factors, indicating to reflect most information only by several factors. When testing the SEM model, this paper sets up two or more observation variables, which are subordinate to a certain potential variable, and then, the trust and efficiency of this kind of variable design is tested.
In this study, potential variables are designed in the SEM model, and each potential variable can be measured by several observable variables. Then, whether these measurement items can really reflex the real features of the measured potential variables requires a factor analysis. The result of the factor analysis on the measurement table can reflect the structural efficiency, while the main evaluation indexes include the factor loading and the cumulative explaining variance. The factor loading represents the relative degree between the original variable and some certain common factor. The cumulative explaining variance indicates the cumulative effective degree of the common factor to the measurement table or model.
2.3 Descriptive statistics and simple correlation analysis
Structural equation model analysis is based on relative coefficients, and the obvious correlation between variables is required. So, to make the correlation analysis on the variables is the premise of carrying out the SEM analysis.
The descriptive statistical analysis mainly checks the mean and the standard derivation of the collected data. The standard derivation shows the average discrete degree of a group of data around the average. The higher the standard derivation, the higher the difference degree of the variables is. The Pearson correlation coefficients are used in analyzing the correlation degree between each variable, and the values of these correlation coefficients should be between − 1 and 1. A positive value shows a positive relationship, while a negative value indicates a negative relationship. The higher the absolute value, the higher degree of correlation it shows [
12].
The analyzing tools of this paper include AMOS7.0 and SPSS16.0. AMOS7.0 is a software package that uses the potential variables based on the variance matrix structure to evaluate the structure model. This method tests the mutual relationship between variables (whether the relationship is direct, indirect, or rational or irrational) through a multi-level analysis. It applies to the models with potential variables for indicating the mutual relationships and testing the model convergence.
From the above introduction, we can see that using the structural equation model technology and the AMOS software package is beneficial for analyzing the proposed problems in this paper. First, the structural equation model reflects not only the separate relationship between the elements in the model but also the mutual influence between them. Second, using the software package of AMOS7.0 to make an analysis can give a full play to the multi-route analysis of this method for presenting the features between variables’ relationship. The clear route of AMOS can help to understand the structural equation model and avoid the interferences brought by measurement errors in a better way.
The survey includes two stages. First, we carry out a small-range questionnaire survey in three representative enterprises. Fifty pieces of questionnaires were provided, and 34 pieces were fed back, with the collecting proportion of 68%. Second, we conducted a larger and more extensive survey in Beijing, Changsha, and Shanghai, with more than 500 pieces of questionnaire being provided, while 397 pieces have been collected back, taking a proportion of 79.4%. Among the collected questionnaires, 350 pieces are valid, and the validity proportion is 88.2%.
4 Results and discussion
The structural equation model includes the measurement model and the structure model. The former describes the relation between the obvious variables and the potential variables, indicating the observable variables of the potential variables [
10]. The simple measurement model is the confirmed factor analysis model. The calculation result of the measurement model, which is the load coefficient between the factor and the index, shows the explaining degree of the measurement index to the total variable of the factor. Meanwhile, the statistical significance of the load coefficients is judged, which is also the
t value.
To judge whether a model can be accepted in the testifying factor analysis, the analysis of the model fitting should be made for judging the validity of the model. The frequently used fitting indexes include the χ2 of the chi-square test and χ2/df test, root mean square error of approximation (RMSEA), standard root mean square residual (SRMR), Normal Fit Index (NFI), Comparative Fit Index (CFI), GFI, and AGFI.
4.1 Evaluation model of internal factors
According to the proposed concept model and the research hypotheses in the former part of this paper, the AMOS7.0 software is used to represent the structural equation model of the internal factors adopted in the service model selection. It can show clearly the contained variables and the variable structure in this research.
The parameter estimation of the SEM model in this paper adapts the method of maximum likelihood estimation, and the model evaluation and revision are carried out from the following three aspects. First, use each fitting index to make an overall evaluation of the model. This research mainly checks the degree of freedom, the chi-square value, the CFI, the NFI, and the RMSEA. Secondly, test the significance of the parameter, the meaning and rationality of the evaluation parameters, such as the coefficient of each route, the residual value, and the statistical significance. Thirdly, determine the statistic value of the variable model fitting degree.
Table
8 shows the overall fitting statistic value of the original model. Here, the representatives of the absolute fitting indexes include the chi-square value
χ2 = 535.232, and the degree of freedom
df = 128, and then the ratio of
χ2/
df = 4.179. The ratio value is between 2.0 and 5.0, indicating that the model is acceptable. The value of RMSEA is 0.098. Steiger (1980) [
21] believed that when the value of RMSEA is lower than 0.1, it shows a good fitting degree; when it is lower than 0.05, the fitting degree is better; and when the degree is lower than 0.01, the fitting degree is excellent. The value of CFI and NFI is all more than 0.9, achieving the requirement level of incremental fit. The above indexes all reflect that the model of the internal influencing factors in the service model selection has a good fitting degree.
Table 8
Fitting degree of the internal structure model
Chi-square value of χ2 | 535.232 |
CFI | 0.913 |
NFI | 0.927 |
RMSEA | 0.098 |
As seen from the data result in Table
9, we get that most indexes have a statistical significance when the factor load coefficient satisfies the condition of
p < 0.05, indicating that each index has a high explaining degree to the total variance of the factor, also that the observed variables can finely explain the potential variables. From the route coefficient table, it can be seen that there are two routes that have passed the significance test.
Table 9
Evaluation of the internal structure
Competitive advantages ←value proposition | .069 | .096 | .713 | .476 |
Competitive advantages ←value transfer | .888 | .237 | 3.740 | *** |
Competitive advantages ←value creation | .050 | .159 | .316 | .752 |
Competitive advantages ←interface rule | .327 | .157 | 2.083 | .037 |
4.2 Evaluation model of external factors
According to the proposed concept model and the research hypotheses in the former part of this paper, the AMOS7.0 software is used to represent the structural equation model of the external factors adopted in the service model selection. It can show clearly the contained variables and the variable structure in this research.
The overall evaluation of the model in this paper also adapts the method of max 1ikelihood estimates and a variety of fitting indexes.
Table
10 shows the overall fitting statistic value of the external influencing factor model. Here, the representatives of the absolute fitting indexes include the chi-square value
χ2 = 1699.9, the degree of freedom
df = 350, and then the ratio of
χ2/
df = 4.86. The ratio value is between 2.0 and 5.0, indicating that the model is acceptable. The value of NFI is 0.832, showing a good fitting degree. The comparative fitting index belongs to the third type of comparative fitting indexes, which also uses the expectation value of the chi-square of the theoretical model or the standard model under the non-central chi-square distribution for regulation besides the first type of information. The CFI value is 0.859, reflecting a good fitting degree.
Table 10
Fitting degree of the external structure model
Chi-square value of χ2 | 1699.9 |
CFI | 0.859 |
NFI | 0.832 |
RMSEA | 0.10 |
Table
11 shows the data result of the factor load coefficients and the corresponding C.R. value between each potential variable and the potential variable in the measurement model. Use the fixed load method, and the calculation result show that the route coefficients of the three elements of capital support, consumer demand, and technical progress have passed the test with a statistical significant under the condition of
p < 0.05.
Table 11
Evaluation of the internal structure
Service model selection capital support | .274 | .089 | 3.068 | .002 |
Service model selection consumer demand | .586 | .132 | 4.433 | *** |
Service model selection technical progress | .531 | .149 | 3.674 | *** |
Service model selection industrial development | .668 | .802 | .833 | .405 |
4.3 Discussion of adjustment variables
Based on the features of the environment faced up by the network enterprises in the network convergence, this paper will select the environmental variables from the four dimensions such as the technical change, the market change, the environmental competition, and the environmental security. The analysis of this part involves the dependent variable, the independent variable, and the regulation variable. Here, the dependent variable is the competitive advantages (CA) of enterprises, and the independent variables are the value proposition (X1), the value creation (X2), the value transfer (X3), and the interface rule (X4).
Suppose the dependent variable of the enterprise’s competitive advantage (CA) is a linear function of multiple independent variables of X1, X2, ……Xn and the error term, we get the multiple linear regression model as follows:
$$ \mathrm{CA}=\upbeta 0+\upbeta 1\ \mathrm{X}1+\upbeta 2\ \mathrm{X}2+\upbeta 3\ \mathrm{X}3+\upbeta 4\ \mathrm{X}4+\varepsilon $$
Here, CA is the dependent variable; X1, X2, X3, and X4 are the independent variables; and ε is a random disturbance term.
This part divides the sample into two statuses according to the technical dynamics, the demand dynamics, the environmental competitiveness, and the security. One is with high changing frequency with a measurement value larger than 4, and the other is with low changing frequency with a measurement value smaller than 4. When the measurement value equals to 4, it would be neglected. Then, the multiple linear regression analysis on effective samples is carried out, see Tables
12,
13,
14, and
15.
Table 12
Regression model fitting of technical dynamics
Value proposition | −.389 | .001 | .366 | .072 |
Value creation | .377 | .000 | −.806 | .000 |
Value transfer | .555 | .000 | .525 | .078 |
Interface rule | .167 | .030 | −.311 | .003 |
R square | 0.891 | 0.770 |
Adjusted R square | 0.875 | 0.723 |
Table 13
Regression model fitting of demand dynamics
Value proposition | .014 | .063 | .495 | .000 |
Value creation | .126 | .121 | .104 | .039 |
Value transfer | .496 | .000 | .623 | .000 |
Interface rule | .417 | .000 | −.459 | .000 |
R square | 0.828 | 0.709 |
Adjusted R square | 0.795 | 0.701 |
Table 14
Regression model fitting of competitiveness
Value proposition | −.293 | .001 | .325 | .025 |
Value creation | .192 | .013 | .855 | .000 |
Value transfer | .663 | .000 | −.120 | .340 |
Interface rule | .241 | .001 | .614 | .001 |
R square | 0.817 | 0.579 |
Adjusted R square | 0.751 | 0.557 |
Table 15
Regression model fitting of security
Value proposition | −.021 | .862 | −.218 | .186 |
Value creation | .156 | .219 | .620 | .002 |
Value transfer | .447 | .000 | .108 | .454 |
Interface rule | .113 | .307 | .478 | .019 |
R square | 0.371 | 0.824 |
Adjusted R square | 0.334 | 0.780 |
As seen from the Table
12, we get that in the environment with high changing frequency, the value proposition and the competitive advantages of enterprises are negatively related, while being positively related with other elements, and in the environment with low changing frequency, the value proposition and value transfer and the competitive advantages of enterprises are positively related, but being invalid in the significance test (
p ≤ 0.05), while the value transfer and the interface rule are negatively related to the CA. Meanwhile, with the technical changing degree changes from a low level to a high level, the relation degree of the service and the enterprise’s competitive advantage increases, indicating that the technologies change faster and the influence of the service model selection to the company’s competitive advantage would be larger.
Therefore, the technological change can influence the relationship between the service model and the competition advantages of enterprises, and in the environment with high technology changing frequency, the service model that has more significant influence to the competition advantages can be confirmed.
As seen from the Table
13, we get that in the environment with high changing frequency, the value proposition, value creation, value transfer, and the interface rule positively relate to the competitive advantage of enterprises, but the value proposition and the value creation have not passed the significance test (
p ≤ 0.05). In the environment with low changing frequency, the value proposition, value creation, value transfer, and the interface rule positively relate to the competitive advantage of enterprises, while the interface rule negatively relate to the CA. Meanwhile, with the technical changing degree changes from a low level to a high level, the relation degree of the service model and the enterprise’s competitive advantage increases, indicating that the market changes faster and the influence of the service model selection to the company’s competitive advantage would be larger.
Therefore, the market change can influence the relationship between the service model and the competition advantages of enterprises, and in the environment of high market changing, the service model that has more significant influence to the competition advantages can been confirmed.
As seen from the Table
14, we get that in the environment with high frequency of change, the value proposition negatively relates to the CA, while value creation, value transfer, and the interface rule positively relate to the CA. In the environment with low frequency of change, the value proposition, value creation, and interface rule positively relate to the CA, while the value transfer negatively relates to the CA, but the value transfer did not pass the significance test (
p ≤ 0.05). Meanwhile, with the technical changing degree changes from a low level to a high level, the relation degree of the service model and the enterprise’s competitive advantage increases, indicating that the fiercer the competition degree, the influence of the service model selection to the company’s competitive advantage would be larger.
Therefore, the environment competition can influence the relationship between the service model and the competition advantages of enterprises, and in the environment of high competition degree, the service model has more significant influence to the competition advantages can be confirmed.
As seen from the Table
15, we get that no matter what the environmental degree is, the results of the significant test are not ideal. However, in the environment with low environmental security, the service model and the competitive advantage of enterprises show a better correlation. Therefore, the environment security can influence the relationship between the service model and the competition advantages of enterprises, and in the environment of low security, the service model that has more significant influence to the competition advantages can be denied.