Transcript Data Mining: Crossing the Chasm

ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
ELEMENTARY MACHINE LEARNING FOR WEBMINING
A Technical paper
National
Institute
Submitted
byOf Science And Technology
Satya narayan Sahu
Satya narayan Sahu
ROLL:IT200117269
[1]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
INTRODUCTION
Algorithms that automatically discovers all pages
in a website whose location is different from the
location where visitors expect to find them and
for selecting the
set of navigation links to
National Institute Of Science And Technology
optimize the benefit to the website are proposed.
 It is hard to organize a website such that pages
are located where visitors expect to find them.

Satya narayan Sahu
[2]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
APPROACHES
New search mechanism called SmartSeek is
introduced
Machine learning
concepts Employs genetic
National Institute Of Science And Technology
algorithm (GA)
System accepts user feedback
Satya narayan Sahu
[3]
National Institute of Science & Technology
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
FUNCTIONING OF SMARTSEEK
 SmartSeek sets up a database about user
objective
 It extracts information about the common words
and phrases
 Looks for similarNational
pagesInstitute
(metasearch
engine)
Of Science And
Technology
 Comparison with other pages
 Feedback
 Further enhancement
Satya narayan Sahu
[4]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
SMARTSEEK MODEL
National Institute Of Science And Technology
Satya narayan Sahu
[5]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
IN DETAIL
Most Relevant
 Moderate Relevancy
 Irrelevant
National Institute Of Science And Technology
 Operators AND (&), OR (|),NOT (~) and CONCAT
(-).
 After n such iteration

Satya narayan Sahu
[6]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
CROSSOVER
National Institute Of Science And Technology
Satya narayan Sahu
[7]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
THE AI
 Artificial Intelligence (AI) is simply a way of
thinking intelligently aims at




Natural
language
processing
National
Institute Of
Science And Technology
Knowledge representation
Automated reasoning
Machine learning
Satya narayan Sahu
[8]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
WEB MINING
Web Mining can be said to have three operations
of interest:
finding natural grouping
of users
pages
National Institute
Of or
Science
And Technology
 finding URLs which tend to be grouped together
 finding the order in which URLs tend to be accessed

Satya narayan Sahu
[9]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
ALGORITHMIC APPROACH
Finding Expected
Locations
National Institute Of Science And Technology
Optimizing the set of Navigation Links
Satya narayan Sahu
[10]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
FINDING EXPECTED LOCATIONS

Model of Visitor Search Patterns
1(a) Single Target
National Institute Of Science And Technology
1(b) Set of Targets

Identifying Target Pages
Satya narayan Sahu
[11]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
EXAMPLE
National Institute Of Science And Technology
Website and Search Pattern
Satya narayan Sahu
[12]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
OPTIMIZING SET OF NAVIGATION LINKS
1) First Only
2)Optimize Benefit
National Institute Of Science And Technology
3)Optimize time
Satya narayan Sahu
[13]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
EXPERIMENTS
National Institute Of Science And Technology
Wharton Website Structure
Satya narayan Sahu
By Rakesh Agrawal
[14]
National Institute of Science & Technology
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
EXAMPLE:
National Institute Of Science And Technology
Satya narayan Sahu
[15]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
CONCLUSION & FUTURE DIRECTIONS
Algorithms have been proposed
Websites without
clear
separation
National
Institute
Of Scienceof
Andcontent
Technology
and index page
Time Threshold to distinguish between two
activities
Satya narayan Sahu
[16]
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute of Science & Technology
REFERENCES
1.www.ibm.com
2.www.almaden.com
Of Science And Technology
3.Mastering Data National
MiningInstitute
by Helenz
Hulki
4.How internet works by K.Perl
Satya narayan Sahu
[17]
National Institute of Science & Technology
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
THANK YOU
National Institute Of Science And Technology
Satya narayan Sahu
[18]
National Institute of Science & Technology
ELEMENTARY MACHINE LEARNING
FOR WEBMINING
National Institute Of Science And Technology
Satya narayan Sahu
[19]