Towards Discriminant Analysis Modeling of Web 3.0 Design and Development for Students, Faculty and IT Professionals

Document Sample
Towards Discriminant Analysis Modeling of Web 3.0  Design and Development for Students, Faculty and IT Professionals Powered By Docstoc
					                                                 (IJCSIS) International Journal of Computer Science and Information Security,
                                                 Vol. 9, No. 9, September 2011




Towards Discriminant Analysis Modeling
of Web 3.0 Design and Development for
 Students, Faculty and IT Professionals
                              S.Padma and Dr.Ananthi Seshasaayee


S..Padma,
Research scholar,
Bharathiar University                                                          Dr.Ananthi Seshasaayee
Coimbatore.                                                                    Associate Prof. &Head,
Assistant Professor                                                            Quaid-e-Millath Govt.
Vels University                                                                College for women,
Chennai .                                                                      Chennai
INDIA.                                                                         INDIA
padmanivasan@gmail.com                                                         ananthiseshu@gmail.com



                                                                applications to different domains. Web 3.0
Abstract                                                        provides integrated real time application
                                                                environment to the user. The applications are
Web 3.0 is an evolving extension of the web 2.0                 majorly involved in searching using semantic
scenario. The perceptions regarding web 3.0 is                  web, 3D web and are media centric. Web 3.0
different from person to person . Web 3.0                       supports pervasive components. Each component
Architecture supports ubiquitous connectivity,                  and its relations are represented below.
network computing, open identity, intelligent web,
distributed databases and intelligent applications .
                                                                In web 3.0, web is transformed into database or
Some of the technologies which lead to the design
and development of web 3.0 applications are                     Data Web wherein the data which are published
Artificial intelligence, Automated reasoning,                   in the web is reusable and can be queried. This
Cognitive architecture, Semantic web . An attempt is            enables a new level of data integration and
made to capture the requirements of Students,                   application interoperability between platforms. It
Faculties and IT professionals regarding Web 3.0                also makes the data openly accessible from
applications so as to bridge the gap between the                anywhere and linkable as web pages do with
design and development of web 3.0 applications and              hyperlinks. Data web phase is to make available
requirements among Students, Faculties and IT                   structured data using RDF[1]. The scope of both
professionals. Discriminant modeling of the
                                                                structured and unstructured content would be
requirements facilitate the identification of key areas
in the design and development of software products              covered in the full semantic web stage. Attempts
for Students, Faculties and IT professionals in Web             will be to make it widely available in RDF and
3.0.                                                            OWL semantic formats.

Keywords : Web 3.0, Discriminant analysis , Design              The driving force for web 3.0 will be artificial
and Development ,Model                                          intelligence. Web 3.0 will be intelligent systems
                                                                or will depend on emergence of intelligence in a
                I INTRODUCTION                                  more organic fashion and how people will cope
                                                                with it. It will make applications perform logical
Web 3.0 is an extension of www, in which the                    reasoning operations through using sets of rules
information can be shared and interpreted by                    expressing logical relationships between
other software agent to find and integrate                      concepts and data on the web. With the




                                                          101                               http://sites.google.com/site/ijcsis/
                                                                                            ISSN 1947-5500
                                            (IJCSIS) International Journal of Computer Science and Information Security,
                                            Vol. 9, No. 9, September 2011




realization of the semantic web and its concepts           services architecture. It surveys base fault
web 3.0 will move into Service Oriented                    tolerance mechanisms and shows how they are
Architecture.                                              adapted to deal with the specifics of the web in
                                                           the light of ongoing work in the area[5]. Barry
The evolution of 3D technology is also being               Norton, Sam Chapman and Fabio Ciravegna
connected to web 3.0 as web 3.0 may be used on             discussed in the paper developing a Service-
massive scale due to its characteristics.                  Oriented Architecture to Harvest information for
                                                           the Semantic web which discusses about the
                                                           Armadillo architecture, how it is reinterpreted as
Web 3.0 is media centric where users can locate
                                                           workow templates that compose semantic web
the searched media in similar graphics and sound
                                                           services and show how the porting of Armadillo
of other media formats.
                                                           to new domains, and the application of new
                                                           tools, has been simplified[6].
The pervasive nature of web 3.0 makes the users
of web in wide range of area be reached not only
in computers and cell phones but also through                          III PROBLEM DEFINITION
clothing, appliances, and automobiles.
                                                           The Design and Development of web 3.0
                                                           products are on the course. Due to the existence
        II REVIEW OF LITERATURE
                                                           of the ambiguity in the requirements of Students,
                                                           Faculty and IT professionals for structuring the
Claudio Baccigalupo and Enric Plaza discussed              web 3.0 products , bridging the gap between web
in the paper poolcasting : a social web radio              3.0 developers and Students, Faculty and IT
architecture for Group Customization about Pool            professionals is required. The key factors for
casting a social web radio architecture in which           each of these three categories students ,faculty
groups of listeners influence in real time the             and it professionals are to be identified and their
music played on each channel. Pool casting users           preference order is to be extracted.
contribute to the radio with songs they own,
create radio channels and evaluate the proposed
music, while an automatic intelligent technique            Let G1, G2, G3 denote the three groups in web
schedules each channel with a group customized             3.0 . The problem is to find the order of
                                                           preferences of the three groups for the three
sequence of musically associated songs[2] .
                                                           categories Students, Faculty and IT professionals
M.T.Carrasco Benitez discussed in the paper
                                                           based on the attributes v1 , v2 , ….. vn included
Open architecture for multilingual social
                                                           in these three groups G1, G2 and G3 to facilitate
networking about an open architecture for all the
multilingual aspects of social networking. This            the design and development of applications in
architecture should be comprehensive and                   web 3.0 for the categories.
address well-trodden fields such as localization,
and more advanced multilingual techniquesto                          IV MATERIALS AND METHODS
facilitate the communication among users[3] .
Autona Gerber, Alta van der Merwe, and                     We collected the perceptions of students
Andries Barnard discussed in the paper A                   ,Faculties and IT professionals inline with web
functional Semantic web architecture about the             3.0 attributes. A five point scale was adapted
CFL architecture which depicts a simplification            which ranges from very low satisfaction , low
of the original architecture versions proposed by          satisfaction,    Medium        satisfaction, high
Bernes-Lee as a result of the abstraction of               satisfaction to very high satisfaction.
required functionality of language layers. Gerber
argues that an abstracted layered architecture for              a.    Block diagram of Web 3.0 discriminant
the semantic web with well defined                                    modeling
functionalities will assist with the resolution of
several of the current semantic web research
debates such as the layering of language
technologies [4]. Ferda Tartanoglu val’erie
Issarny, Alexander Romanovsky and Nicole
Levy discussed in the paper Dependability in the
web services architecture which lists about how
to build dependable systems based on the web



                                                     102                               http://sites.google.com/site/ijcsis/
                                                                                       ISSN 1947-5500
                       (IJCSIS) International Journal of Computer Science and Information Security,
                       Vol. 9, No. 9, September 2011




                                      b. Steps in Web 3.0 Discriminant modeling

                                           a.   Start
                                           b.   Collect the perceptions regarding the
                                                attributes of web 3.0 among the three
     Start                                      categories Students, Faculty and IT
                                                professionals
                                           c.   Classification of attributes into three
                                                groups G1, G2 and G3.
                                           d.   Compute Mean and Standard Deviation
                                                for G1, G2 and G3
     Collection of                         e.   Correlation Coefficient among the
 Perceptions of web                             groups G1, G2 and G3
                                           f.   Discriminant Modeling for the three
 3.0 among students,                            categories Students, Faculty and IT
    Faculty and IT                              professional
     professionals                         g.   stop




Classification of                     c. Preprocessing
attributes into G1,
G2 and G3                             The data collected are verified for completeness.
                                      The missing values are replaced with the mean
                                      value.

                                      d. Classification
Mean and Standard
                                      The data collected from the three categories
Deviation for G1,                     Students , Faculty and IT professionals based on
G2 and G3                             the attributes 2D, 3D, Audio, Custom mash up, E
                                      decisions, Multilingual, Result as Mash up,
                                      Semantic Maps, Semantic Wiki, Software
                                      Agents, Speech recognition. Based on the
Correlation among                     functionality of the attributes they are grouped
                                      into G1 , G2 and G3. G1 comprises of
G1, G2 and G3                         Multilingual, Semantic maps, Edecisions,
                                      Semantic wiki and Software agents . G1 is
                                      termed as Applications . G2 comprises of 3D,
Discriminant                          Audio, 2D and Speech recognition. G2 is termed
modeling for three                    as Media. G3 comprises of Custom Mash up,
                                      Result as Mash up . G3 is termed as Output.
categories Students,
Faculty and IT
professionals



   Stop



                                103                               http://sites.google.com/site/ijcsis/
                                                                  ISSN 1947-5500
                                                 (IJCSIS) International Journal of Computer Science and Information Security,
                                                 Vol. 9, No. 9, September 2011




                                                                The correlation coefficient for all pairs among
                                                                the Groups are calculated using the following
                                                                formula.[7]

e. Mean and Standard Deviation                                  Correlation(r) =[ NΣXY - (ΣX)(ΣY) /
                                                                Sqrt([NΣX2 - (ΣX)2][NΣY2 - (ΣY)2])]
                                                                where
                                                                         N = Number of values or elements
TABLE 1. COMPARISION OF MEAN FOR THE THREE                               X = perception weightage for 1st group
               CATEGORIES                                                Y = perception weightage for 2nd group
                                                                         ΣXY = Sum of the product of first and
                                                                Second group perceptions
                                                                         ΣX = Sum of 1st group
                                  MEAN
                                                                         ΣY = Sum of 2nd group
 CATEGORY                 G1            G2        G3                     ΣX2 = Sum of square 1st group
 STUDENTS               3.94049        3.5794    2.92                    ΣY2 = Sum of square 2nd group
 FACULTY                 3.17           2.95     2.49
                                                                   TABLE III . CORRELATION AMONG GROUPS FOR
 ITPROFESSIONALS         3.97           3.9      3.36                               STUDENTS


                                                                            STUDENTS
                                                                             G1      G2          G3
For all the three categories G1 Applications has
higher mean when compared to all others.                            G1        1      0.56        0.74
                                                                    G2       0.56     1          0.51
TABLE II . COMPARISION OF STANDARD DEVIATION                        G3       0.74    0.51         1
            FOR THE THREE CATEGORIES




                                                                   TABLE IV .CORRELATION AMONG GROUPS FOR
                                                                                   FACULTY
                             STANDARD
                             DEVIATION
     CATEGORY           G1        G2       G3
                                                                                   FACULTY
     STUDENTS          0.57     0.52      0.53
                                                                             G1       G2          G3
      FACULTY          0.57     0.49      0.45
                                                                    G1        1      0.34        0.34
 ITPROFESSIONALS       0.58     0.55      0.39
                                                                    G2       0.34      1         0.28
                                                                    G3       0.34    0.28          1




The standard deviation for G3 are comparatively
lower for faculty and IT professionals . Faculty
and IT professionals have similar opinions about
G3 – output. There is no significant difference in
the standard deviation of students.                               TABLE V. CORRELATION AMONG GROUPS FOR IT
                                                                                PROFESSIONALS
                                                                            ITPROFESSIONALS
f. Finding the Correlation Coefficient
                                                                              G1          G2            G3
Correlation Coefficient reveals          the     nature             G1         1          0.3          0.42
relationship between the attributes.                                G2        0.3           1          0.32
                                                                    G3       0.42      0.32             1



                                                          104                                   http://sites.google.com/site/ijcsis/
                                                                                                ISSN 1947-5500
                                            (IJCSIS) International Journal of Computer Science and Information Security,
                                            Vol. 9, No. 9, September 2011




It is evident that all the three groups G1, G2 and
G3 are positively correlated to each other in all
the three categories Students, Faculty and IT
professionals.

g. Discriminant modeling on groups:
                                                           Based on the Canonical Discriminant Function
Discriminant Function analysis is a two step               coefficient, the linear discrminant equation can
process.                                                   be written as

Step 1: A set of discriminant functions are tested            TABLE VII CLASSIFICATION RESULTS
for its significance.
                                                           Y = -2.339+ 3.247 x1 + 0.885 x2– 0.547 x3             (1)
1.a. A Matrix of total variances and Co variances
are constructed                                            Y = -7.452 -1.017x1 + 0.408x2 + 1.887x3               (2)

1.b. A matrix of pooled within group variances             Based on (1) the following are the classification
and Co variances are constructed.                          results.

1.c. F test is performed on the two matrices
constructed.

1.d Variable which have significantly different
means across the groups are identified.

Step 2. Classification.

In the classification step classification of
variables are done. DA automatically determines
some optimal combination of variables so that
the first function provides the most overall
discrimination between groups and the second
provides the second most and so on. The
functions are independent or orthogonal [8].

       V RESULTS AND DISCUSSION

TABLE VI . CANONICAL FUNCTION COEFFICIENTS



                            Function
                   1             2
 Applications      3.247         -0.01
 (x1)
 Media (x2)        0.885         0.408
 Output (x3)       -0.547        1.887
 (Constant)        -2.339        -7.452




                                                     105                               http://sites.google.com/site/ijcsis/
                                                                                       ISSN 1947-5500
                                                                 (IJCSIS) International Journal of Computer Science and Information Security,
                                                                 Vol. 9, No. 9, September 2011




                                                                          Classification Results

                                                   Group                                           Predicted Group Membership
                                                                                   Students         Faculty           IT Professionals           Total
                                                  Students                           218               7                     8                    233

                     Count                         Faculty                            4              347                     6                       357
                                               IT Professionals                       3               12                    421                      436

  Original
                                                  Students                           93.5             7.3                   4.7                      100

                        %                          Faculty                           1.1             97.2                   1.7                      100
                                               IT Professionals                      0.7              2.8                   96.6                     100


a. 95.8% of original grouped cases correctly The order of preferences for the three categories is
classified.                                  given below based on the above Classification
                                             Function Coefficients.

               Percentage of correct classification


  98                                                                           Order of Preferences among groups for Students
  97
  96                                                     Percentage of
  95                                                     correct
  94
  93                                                     classification
  92
  91
                                         al
                      lty
         ts




                                     io n
       en



                   cu
     ud



                 Fa



                                   ss
   St




                                ofe
                             pr
                            IT




TABLE VIII CLASSIFICATION FUNCTION COEFFICIENTS



                                        Categories
                   Students          Faculty       ITProfessionals

 Applications        14.048            3.743            4.818

    Media             9.374            6.097            7.488
                                                                               :
   Output             8.074             7.66            12.521

  (Constant)        -46.475           -29.192          -49.519




                                                                            106                               http://sites.google.com/site/ijcsis/
                                                                                                              ISSN 1947-5500
                                          (IJCSIS) International Journal of Computer Science and Information Security,
                                          Vol. 9, No. 9, September 2011




                                                      From the above three tables the design and
                                                      development of web 3.0 products specifically
                                                      related to Students, Faculty and IT professionals can
                                                      ensue the group preference orders and attributes .
                                                      The products can be designed with the maximum
                                                      attributes in the first group preference followed by
                                                      lesser attributes in the second and third group.

                                                                        VI CONCLUSION

                                                      The perceptions inline with web 3.0 are collected
                                                      from students, Faculty and IT professionals. The
                                                      data’s are preprocessed , classified, Mean, Standard
Order of preferences among groups for Faculty         deviation and correlation coefficient are computed
                                                      to understand the descriptive and Discriminant
                                                      modeled. At the outset of evolving growth in Web
                                                      3.0 this model is an initiative for the of web 3.0
                                                      product design for Students , Faculty and IT
                                                      professionals.




                                                                          REFERENCES

                                                      [1].
                                                      en.wikipedia.org/wiki/Resource_Description_Frame
                                                      work

                                                      [2] Baccigalupo C.; Plaza E.; Poolcasting: a Social
                                                      Web      Radio       Architecture      for    Group
                                                      Customisation.Computer.org
                                                      2007.http://portal.acm.org/citation.cfm?id=1332471.
                                                      1332755.

Order of preferences among groups for IT [3]. ]. Carrasco Benitez M.T;, Open architecture for
Professionals                            multilingual social networking. Position paper for
                                         the W3C workshop on the Future of Social
                                         Networking [FSN], 15-16 January 2009.

                                                      [4]. . Aurona Gerber; Alta van der Merwe ; Andries
                                                      Barnard; A functional semantic architecture.

                                                      http://www.eswc2008.org/final-pdfs               -for-web-
                                                      site/fisr-1.pdf.

                                                      [5]. Ferda Tartanoglu val’erie Issarny; Alexander
                                                      Romanovsky and Nicole Levy er; Dependability in
                                                      the      web      services    architecture. 2007
                                                      .http://citeseerx.ist.psu.edu/viewdoc/download?
                                                      doi=10.1.1.14.771&rep=rep1&type=pdf.




                                                   107                               http://sites.google.com/site/ijcsis/
                                                                                     ISSN 1947-5500
                                              (IJCSIS) International Journal of Computer Science and Information Security,
                                              Vol. 9, No. 9, September 2011




[6]. Barry Norton; Sam Chapman ; Fabio Ciravegna;
Developing a Service- Oriented Architecture to
Harvest information for the Semantic web .2005.

[7].       http://easycalculation.com/statistics/learn-                  S.Padma , is a research scholar in
correlation.php                                                          Bharathiar university , Coimbatore.
                                                                         She has published 2 international
[8].                                                                     journals . Her area of interests are
http://www.statpower.net/Content/312/Lecture                             Web mining.
%20Slides/Discriminant%20Analysis.pdf

[9]. Production of Cross Media Content for Multi-
Channel Distribution, 2007, AXMEDIS apos;07                                Dr. Ananthi Seshasaayee received her
Third International Conference Volume ,Issue , 28-                         Ph.D in Computer Science from Madras
30 Nov . 2007 Page(s):115-122.                                             University. At   present she is working
                                                                           as     Associate professor and Head,
                                                                           Department of computer science, Quaid-
[10 ] Mantovani.F; VR Learning : Potential and                             e-Millath Government College for
Challenges for the Use of 3D Environments in                               Women, chennai. She has published 17
Education and Traning . 2001.                             international journals. Her area of interest involve the
                                                          fields of Computer Applications and Educational
[11]. Andr.P.Freire; Renata; P.M.Fortes ;An               technology.
Evaluation of Web Accessibility Metrics based on
their Attributes .2008.

[12]. Nuno Laranjeiro; Towards Fault Tolerance in
Web Services Compositions .2007.

[13]. Dhiraj Joshi ; Ritendra Datta ; Ziming Zhuang
;PARAgrab : A Comprehensive Architecture for
Web Image .2006.

[14]. Antonio Tapiador; Antonio Fumero, Joaqu’m
Salvach’ua; Sandra Aguirre; A Web Collaboration
Architecture .2006.




                                                       108                               http://sites.google.com/site/ijcsis/
                                                                                         ISSN 1947-5500

				
DOCUMENT INFO
Shared By:
Stats:
views:67
posted:10/12/2011
language:English
pages:8