Traversing Documents by Using Semantic Relationships

hurriedtinkleΤεχνίτη Νοημοσύνη και Ρομποτική

15 Νοε 2013 (πριν από 3 χρόνια και 8 μήνες)

90 εμφανίσεις

Traversing Documents by Using
Semantic Relationships

Presenter:
Bilal
Gonen

Semantic
Browser


Semantic browser is a tool that supports
browsing and navigation of a document space
by utilizing the semantic relationships.

Physical Links vs. Virtual Links

co_occurs_with

analyzes

affects

is_result_of

href

href

href

href

affects

affects

affects

co_occurs_with

co_occurs_with

co_occurs_with

co_occurs_with

co_occurs_with

is_result_of

is_result_of

is_result_of

is_result_of

A Real Example

How are these
articles related?

How do we find
other
documents
related with
“melanoma”?

One common option
is
to
use statistical techniques.

Recommendation Systems

Amazon, the best
-
known commercial recommender system,
recommends books to customers based on the statistical
similarity between customers' previous purchases.

The product:
Digital Camera

Customers who bought this
item also often bought

Digital Memory Card

Statistical proximity

A Real Example

Such a statistical technique may
return these terms.

sun's harmful rays

skin

skin cancer

leg

ankle

skin pigment

melanin

aneuploidy

There are no named
relationships.

Ontology

Pathologic
Function


Body
Substance

Substance

Diagnostic
Procedure

Amino Acid
Peptide or
Protein


Disease or
Syndrome


Neoplastic
Process

affects

Instance level

Schema level

Ureteral
Calculi

Kidney
Neoplasms

InstanceOf

InstanceOf

InstanceOf

analyzes

rdfs:subClassOf

rdfs:subClassOf

rdfs:subClassOf

Ontology is at the heart of Semantic Web.

Relationships In Ontology

breast cancer

bone cancer

non
-
melanoma

melanoma

blood cancer

skin cancer

cancers

aneuploidy

euploidy

monoploidy

chromosomal
disorder

is_result_of

A Real Example

Our approach is to offer
several relationships to
the user.

aneuploidy

allelic imbalance

chromosome
aberrations

This is what
user is
interested in.

affects

co_occurs_with

occurs_in

is_result_of

Return files
which includes
“aneuploidy”

Chromosomal Aneuploidies

Identification of Aneuploidy

Definition of Aneuploidy

Aneuploidy and Deletions

Name of files in which
“aneuploidy” occurs.


JSP

(Java Server Page)


Java Script

AJAX

user


Lucene Index
for documents


PubMed
dataset


Ontology



SemDis API

Lucene indexing is used to index
the documents with the 21,945
MESH terms when they occur in
the documents.


User Interface

(HTML page)

The advantage of the AJAX
technology is to send and receive
only needed information between
the client and server.

request

Built in LSDIS Lab. This API
is used to process the
triples in the ontology.

Contains 135 classes and
49 relationships in
schema level. And 21,945
entity instances in the
instance level

Contains 48,252
documents

Because we also used the
synonyms of the 21,945 MESH
terms, therefore we used
~104,000 terms to index the
documents.

AJAX (Asynchronous JavaScript And XML)

AJAX (Asynchronous JavaScript And XML)

Only this part is loaded to the client
side.


User Interface

(HTML page)

The MESH term is sent to
server as a request to get its
types from the ontology by
using the SemDis API.


Java Script

AJAX


JSP

(Java Server Page)

user


Lucene Index
for documents


PubMed
dataset


Ontology



SemDis API

SemDis API gets the types
of the instance term from
the ontology.

request

response

keyword

related documents

List of the documents are
returned from the Lucene
index.

Questions, Comments

Thank you…

Email: bilalgonen@gmail.com

Web: www.bilalgonen.com