Providing Intelligent Content by Using Semantic Web and

Download Report

Transcript Providing Intelligent Content by Using Semantic Web and

Providing Intelligent Content by
Using Semantic Web and Web
Mining
Pinar SENKUL
METU
Computer Eng. Dept.
17/07/2015
Providing Intelligent Content Pinar SENKUL - METU CENG
1
Intelligent Content
What is content?
Anything published in the web.
Context management systems deal
with handling organizing and
presenting the content.

17/07/2015
Providing Intelligent Content Pinar SENKUL - METU CENG
2
Intelligent Content

How about intelligent content?
Adaptable content according to the
needs and habits of the user.
17/07/2015
Providing Intelligent Content Pinar SENKUL - METU CENG
3
Semantic Web


to create a universal medium for
information exchange by putting
documents with computer-processable
semantics on the World Wide Web.
It is a vision of web pages that are
understandable by computers, so that
they can do searching and calling web
services automatically in a standardized
way.
17/07/2015
Providing Intelligent Content Pinar SENKUL - METU CENG
4
Semantic Web

Related standards
●
XML (the basic platform of the semantic web)
●
●
RDF (Resource Description Language – gives power to
build web in a machine-processable manner)
OWL (Ontology Web Language – provides a language
for defining structured, Web-based ontologies which
delivers richer integration and interoperability of data
among descriptive communities)
17/07/2015
Providing Intelligent Content Pinar SENKUL - METU CENG
5
Web Mining
web mining is to discover interesting patterns
from the web data by using data mining
techniques.
web documents data
●web structure data
●web log data
●user profiles data.
●
17/07/2015
Providing Intelligent Content Pinar SENKUL - METU CENG
6
Web Mining
Web Content Mining (Web content mining is the
extraction of useful information from the content of the
web documents.)
●
Web Usage Mining (Web usage mining is application of
data mining techniques to discover user access patterns
from web data.)
●
Web Structure Mining (Web structure mining can be
described as discovering web structure and link topology
information from web.)
●
17/07/2015
Providing Intelligent Content Pinar SENKUL - METU CENG
7
Web Usage Mining


automatic discovery of user access
patterns from one or more Web servers
can help to determine




17/07/2015
to determine the life time value of
customers,
cross marketing strategies across
products,
and effectiveness of promotional
campaigns
how to better structure a Web site in
order to create a more effective
Intelligent
Content presenceProviding
for
the
organization.
Pinar SENKUL - METU CENG
8
Web Usage Mining
Smart-SRA:
 a new approach for session
construction
 a reactive session reconstruction
algorithm. Time and navigation
oriented heuristics are used with the
site topology.
 improvement on the quality on the
constructed session
17/07/2015
“A New Approach for Reactive Web Usage Data Processing", by M. A. Bayir et.
al., ICDE-WIRI, April
2006. Intelligent Content Providing
Pinar SENKUL - METU CENG
9
Web Usage Mining
Enhancement with Semantic Information
 consider the semantic information of
the visited sites in site construction and
finding associations
 New and semantically correct
associations and profiles can be
extracted
17/07/2015
Providing Intelligent Content Pinar SENKUL - METU CENG
10
Multi-relational Data Mining





17/07/2015
Finding associative rules over multiple
relations (tables)
Integrating data from multiple tables into
a single table can cause loss of semantics
and information
New techniques for multiple relations.
Inductive Logic Programming (ILP) is a
frequently used approach.
Used in semantic web mining for ontoloy
extraction
Providing Intelligent Content Pinar SENKUL - METU CENG
11
Multi-relational Data Mining
ILP – based approach:
 Given:




17/07/2015
a set of examples, E
background knowledge, BK
produce a set of relations (clauses)
using BK that describe E.
Strong language bias : precise
syntactical description of acceptable
clauses
Providing Intelligent Content Pinar SENKUL - METU CENG
12
Multi-relational Data Mining

A new ILP-based association rule
discovery system
 No negative examples
 Purely relational
 Rules generated under less
language bias and user-defined
declarations
“A hybrid technique for multi-relational rule mining”, Seda Daglar-Toprak et.al,
Technical Report, 2005
17/07/2015
.
Providing Intelligent Content Pinar SENKUL - METU CENG
13
Multi-relational Data Mining

17/07/2015
Can be used for
extracting/generating the
semantics of the personalized
content
Providing Intelligent Content Pinar SENKUL - METU CENG
14
Providing Intelligent Content

A framework for intelligent content
generation that makes use of



17/07/2015
Semantically extended session
generation and web mining approach
ILP-based multi-relational association
rule mining for generating/extending the
semantics of the content
Applicable to various domains
Providing Intelligent Content Pinar SENKUL - METU CENG
15