Project Title: Ranking in
Academic Social Networks
Principle
Investigator:
Dr Ali Daud
Executive
Summary: With the growth of academic social networks ranking of academic objects
(authors, papers, and conferences/journals) is important in academic social
networks for academic recommendation tasks. This proposal investigates the
challenges regarding different flavors of ranking in academic social networks,
which are; the problem of ranking based on author’s contribution, inclusion of
temporal aspect for ranking, semantic ranking and learning to rank for academic
objects. In past several link based (e.g. PageRank)
and citation count based (e.g. H-Index) algorithms have been proposed. The
citation based methods totally ignore the link structure of the environment,
while some link based methods have incorporated few measures like number of
publications, number of citations etc as weights, along with the basic link
structure. The successful incorporation of all important weights, temporal
dimension, semantics and learning can significantly improve the performance of
ranking methods. In this project we will conduct a series of experiments to
work out improved algorithms. The dataset from CiteSeer
and DBLP online publication repositories of computer science literature will be
used. They both provide computer science related publications gathered from
different famous publishers such as ACM, IEEE, Elsevier, Springer, and Wiley.
The data variables will be publication text, authors, citations (in-links and
out-links), publication year, journal or conference.
Focus
Areas: Ranking based on author’s contribution, temporal
ranking of academic objects, semantic ranking of academic objects, and learning
to rank methods for academic objects.
Project Modules: There are four
modules in this project related to ranking of academic objects.
a)
Authors
Contribution based Ranking
b)
Time
Weighted Ranking
c)
Semantics
based Ranking
d)
Learning
to Rank
Scope of the
Project:
In this project we are going to address different challenges involved in
ranking of academic social networks. It is being investigated by the provisions
of the academic challenges of the various flavors of social networks, including
the problem of ranking based on the author's contribution and involvement of
the time factor while ranking, ranking of semantic objects and learning to
rank. We’ll integrate all the important weights, dimension of time to result in
significant improvement in the performance. In this project we will conduct a
series of experiments to improve the algorithms.
Research Objectives:
1)
To
ensure the maximum credit attribution to the researchers. Users can be
interested to find papers and articles written by a particular author
2)
Academic
promotion and grant funding requires measuring research work of a particular candidate
3)
Properly
ranked author names are also used in many tasks such as searching homepage and
finding the topics as a particular author is interested in single or few topics
4)
Time-
weighted ranking to produce the results for queries relevant to a particular
time as some researchers require time oriented results
Academic Objectives:
1)
To
attract the students towards latest approaches for information retrieval in
academic social networks
2)
To
indulged the student and the faculty members in research and development work
3)
To
make a batter cooperation between Academia and industry
4)
To
enhance the capabilities of research and development and team work management
Publications on different achievements in this project will enhance the repute
of the university, ICT (R&D) Fund, and HEC which will automatically attract
the foreign students as well as the well qualified foreign faculty members
5)
This
project will encourage the applied research rather than theoretical research. MS
and PhD students can complete their thesis by becoming a part of this project