This document outlines a proposed system to filter unwanted messages from online social networks. It discusses the existing problems of misuse on social media platforms. The proposed system would use machine learning techniques like SVM for text categorization and identification of fake profiles to filter content by category (e.g. abusive, vulgar, sexual). It presents the system architecture as a three-tier structure and provides results of testing the filtering mechanism and classifier. The conclusion is that the "Filtered wall" system could address concerns around unwanted content on social media walls.
Coefficient of Thermal Expansion and their Importance.pptx
ONLINE SOCIAL NETWORK
1. CONTENTS
• Problem Definition
• Existing System
• Proposed System
• Literature Survey
• System Architecture
• Algorithm
• Mathematical Model
• Results
• Conclusion
• References
• Pulication
2. PROBLEM DEFINITION
• Online Social Media Has Become Face Of Internet Today.
• On-line Social Networks (OSNs) are most popular interactive medium to
communicate and share data, views and so much more .
• According to a rough estimate 30 billion pieces of content (web
links, news stories, blog posts, notes, photo albums, etc.) are shared
each month.
3. PROBLEM DEFINITION
• For the numerous positives there are a few negatives
which spoil the user experience .
• Primary among them is the inability to control what is
being posted on the wall.
• Our system exploits this weakness and comes up with a
solution for it .
4. PROBLEM DEFINITION (CONTD..)
• Blatant misuse of social media platforms
• Recent examples (Shivaji Issue)
• Unfiltered messages on wall can be a tool used to bully.
• The system aims at addressing such concerns and disallow the misuse.
5. EXISTING SYSTEM
• Platforms like Facebook ,Google do not provide filtering of any sort.
• Don’t have any mechanism to control fake profile menace .
• Can’t restrict updation and has been on the receiving end for a lot of such
issues.
6. EXISTING SYSTEM ARCHITECTURE
The architecture in support of OSN
services is a two-tier Structure.
Graphical User Interfaces
Social Network Manager
7. PROPOSED SYSTEM
• Propose and evaluate an automated system, called filtered wall (fw), able
to filter unwanted messages from osn user walls.
• We exploit machine learning (ml) text categorization techniques.
• Fake profile identification. IP monitoring.
• All existing features of the social media
8. PROPOSED SYSTEM (CONTD..)
• The concept of blocklist to further the utility value of the system .
• The blacklist will be a list where repeated offenders will be listed and
will be blocked from further usage
• The use of a THRESHHOLD VALUE for the same will be done
9. LITERATURE SURVEY
SR.
NO
PAPER TITLE AUTHORS ADVANTAGES LIMITATIONS PROPOSED SOLUTIONS
[1] A Comparison of
Classifiers and
Document
Representations for the
Routing Problem
[1995]
H. Schutze, D.A.
Hull, J.O. Pedersen
Error minimization Detail approach for
dimensional reduction
is required
Provides for detail
approach to dimensional
reduction
[2] Boos Texter: A Boosting-
based System for Text
Categorization
[2000]
Robert E Schapire,
Yoram Singer
Predict a good
approximation of
the set of labels
associated with a
given document
Issues in
implementation of
multilabel text
categorization
Provides for multilevel
text categorization
[3] Content-Based Book
Recommending Using
Learning for Text
Categorization[1995]
Raymond J. Mooney,
Loriene Roy
Being able to
recommended
previously unrated
items to users with
unique interests
Ratings are based on
random sampling of
items, so are having
less accurate
considerations
Ratings are based on a
group of experts
suggestions and review
and as such are more
accurate.
10. LITERATURE SURVEY (CONTD..)
SR.
NO
PAPER TITLE AUTHOR ADVANTAGES LIMITATIONS PROPOSED SOLUTIONS
[4] UNWANTED
MESSAGES ARE
FILTERED USING
CONTENT MINING
[Jan 2014]
Ms. Shruti C. Belsare,
Prof. R.R. Keole
A system will
automatically filter
unwanted messages
from OSN user
walls on the basis
of both message
content and the
message creator
relationships and
characteristics.
No assistance to users
for FR specification.
a different semantics for
filtering rules to better fit
the considered domain, an
online setup assistant
(OSA) to help users in FR
specification.
[5] A System Approach
to Avoid Unwanted
Messages
from User Walls
[feb 2014]
Dipali D. Vidhate,
Ajay. P. Thakare
concerns both the
rule
layer and the
classification
modules.
The analysis of related
work has highlighted
the lack of a publicly
available benchmark for
comparing different
approaches to content
based classification of
user walls short
texts.
To cope with this lack, we
have built and made
available a dataset D of
messages taken from
Facebook. The group of
experts has been chosen in
an attempt to
ensure high heterogeneity
concerning sex, age,
11. SYSTEM ARCHITECTURE
• The architecture in support of OSN services is
a three-tier Structure.
• The first layer Social Network Manager
(SNM)
• The second layer Social Network Applications
(SNAs).
• The third layer Graphical User Interfaces
(GUIs).
13. ALGORITHM
• SVM (Support vector machine) are supervised learning models that
analyze pattern and recognize pattern.
• Used for classification.
• SVM builds model that assigns examples in category or other .
• Representation of examples as points in space
14. MATHEMATICAL MODEL
• Set theory to make classifications of the content into relevant categories.
• Each of the categories will contain words which can be classified as
containing in those categories .
• The universal set will contain the input string which will be the text which
is posted on the wall .
• The sets will be AB(abusive ),V(vulgar),P(political),S(sexual) .
15. MATHEMATICAL MODEL (CONTD..)
[1] ML based classifier:
Input: Short Text
Output: contentSpec[showing which message belongs to particular class]
[2] Filtering mechanism:
Input: (author, creatorSpec, contentSpec, action)
Process:
FM=[UserSpec,contentSpec==category(Violence,Vulgar,offensive,Hate,Sexual)]
Output: {ContentSpec, M||Y}
19. CONCLUSION
• Filtered wall is a system to filter undesired messages from OSN walls.
• This system approach decides when user should be inserted into a black
list.
• Filtered wall has a wide variety of applications in OSN wall.
20. REFERENCES
• [1] A. Adomavicius, G.and Tuzhilin, “Toward the next generation of
recommender systems: A survey of the state-of-the-art and possible
extensions,” IEEE Transaction on Knowledge and Data Engineering,vol.
17, no. 6, pp. 734–749, 2005.
• [2] M. Chau and H. Chen, “A machine learning approach to web page
filtering using content and structure analysis,” Decision Support Systems,
vol. 44, no. 2, pp. 482–494, 2008.
• [3] R. J. Mooney and L. Roy, “Content-based book recommending using
learning for text categorization,” in Proceedings of the Fifth ACM
Conference on Digital Libraries. New York: ACM Press, 2000, pp. 195–
204.
• [4] F. Sebastiani,“Machine learning in automated text
categorization,”ACM Computing Surveys, vol. 34, no. 1, pp. 1–47, 2002.
21. PUBLICATIONS
• Published literature survey paper in National conference on inter
disciplinary research on recent trends in information technology and
computer engineering (NCRTIT).
• Published in IJAFRC journal (ISSN:2348-4853)
• Published in Indira college of engineering and management under
QUANTONIUM ’15.
• Presented and placed FIRST in DY.PATIL COLLEGE OF ENGINEERING
paper presentation competition judged by industry experts.
• Submitted paper Asian Journal for Engineering and Technological
Innovation (AJETI) (ISSN 2347-7385).