1 d.3

When Metadata is the Content
From Articles to Knowledge
SSP 2009 Annual Meeting
Chris Beguel – Director of Sales – TEMIS
Baltimore, MD – May 09

Where are we? Semantic Age!

Copyright © 2009 TEMIS –All rights reserved 2

From Words to Meaning…
Trimilax 500 mg makes me feel dizzy after ingestion

Term Prop. Num. Abrev. Verb /3rd Pron. Verb Adj. Prep. Noun

Entity Product Dosing Action Target State Event Action

Fact Drug Symptom Condition

Potential Adverse Effect
Drug = Trimilax
Knowledge Dosing = 500mg
Symptom = Tireness
When = After administration

Metadata? Understand!
Metadata
Title: Google gives drivers a hand
at the gas pumps
Source: InformationWeek
Author: Antone Gonsalves
Date: November 7, 2007

Entities

Facts


Metadata
Entities
Companies
Gilbarco Veeder-Root Gilbarco

Google InformationWeek

T-Mobile HTC

Qualcomm Motorola
Persons
Lucy Sackett Sackett

Locations
Atlanta United States

Organizations
National Association of Conveni…

Technologies
Internet Linux Open-source …
Product
New Service Google Service

Copyright © 2009 TEMIS –All rights reserved Facts 5

Who: Gilbarco
Whom: unknown
What: New Service Metadata
When: unknown

Announcement Entities
Who: Gilbarco Facts
What: Google Service
When: early next week Announcement
Gilbarco New service
Who: Sackett
Launch Whom: InformationWeek
When: unknown Sackett InformationWeek
What: unknown
Launch
Gilbarco Google Service

Function Function
Announcement
Who: Gilbarco Sackett Gilbarco
With whom: Google
Who: Sackett When; unknown
State: Negative
Partnership
Company: Gilbarco
Who: Google
Function: spoke woman Gilbarco Google
With whom: T-
Mobile, HTC,
Partnership Qualcom, Motorola Alliance
When: unknown T-Mobile
Google HTC
Alliance
Qualcomm

Motorola

From Metadata to Knowledge!


What is Text Mining?

v Text Mining is an information access technology…
v Text Mining generates Knowledge
v Text Mining serves information consumers & producers

Text Mining Back-End

Data
Repository

Text Mining Front-End
(Text Analytics)


1. Enhanced Search Experience
From standard keyword search….

Simple recognition of words…


1. Enhanced Search Experience
… to Entity & Fact search!

•Make comprehensive and precise search
End-User
•Get more relevant documents
Benefits •Find what you don’know!
t

2. Faceted Navigation
From “
narrow your search”
….


2. Faceted Navigation
… to multi-dimensional faceted navigation

Point & Click
filtering

Ability to combine
several filters at once
(and/or)

Self-adjusting
filters to refine the
search

•Get a quick vision of document content
End-User
•Navigate within context-relevant information
Benefits •Rapidly focus on targeted documents

3. Data Analysis and Reporting
From bug view ….


3. Data Analysis and Reporting

… to bird-
eye view!

•Visualize key Entities & Facts (pie/bar charts)
End-User
•Detect Entities & Facts dependencies (matrix charts)
Benefits •Zoom in & out by drilling anywhere

4. Information Discovery
From flat list of documents ….


4. Information Discovery
… to
information
network
Discovery
Search
Tools
Panel

Entities

Proofs
Facts

•Search in knowledge, not in documents
End-User
•Get a graphical representation of knowledge
Benefits •Discover information by navigating within Facts

Semantic Enrichment at the Core

Automatic Entity & Facts Taxonomy Content
Categorization Extraction Management Editors

Related Topics
Editorial Web Content Extraction
& Content Management
Similarity
Management Detection

Smart
Text Mining Linking

Content Enrichment Trends Analysis
Product & Charting
Visitors &
customers
Management
Sentiment
Analysis

Content Metadata
Annotation Extraction

Original Content
Journal Scans
Expert Interviews
Event Reports


Benefits to Information Producers
Increase stickiness of website to maximize
ad revenue or subscription utilization!
v Create more engaging, longer lasting user visits
• Richer user experience with context sensitive information
• Enhanced page views per visits
• Exposing the “long tail” through suggestions and linking
• Integrate more content at a fraction of the cost
v Establish your web properties as a community
gateway
• “70% of all searches do NOT start on Google/MSN/Yahoo”
says Sue Feldman at IDC Research
• Smart search and navigation are critical to user’ experience
s

Re-Packaging Content – Elsevier
v Objective
• Develop a revolutionary database indexing the last 28 years
in chemistry patent
• Provide an exceptional users’experience by using “smart
content”
v Results
• ~20 Million Chemistry Patent documents
• Searchable by chemical reactions, solvents, reactants directly
extracted from the documents
• Released by Elsevier-MDL in Nov. 2004
v Currently
• TEMIS distributes the Chemical Entities Relationships
Annotator in partnership with Elsevier

Exposing the Long Tail – Springer
v Objective
• Mapping of meaningful words and phrases
in journal articles to encyclopedia entries
• Identification of related documents in a pool of over
three million journal articles
v Solution
• Indexing of incoming journal articles to link journal
articles with the related encyclopedia entry
• Creation of semantic fingerprint for each journal article
to allow search engine calculate degree of relationship
• Integration with Springer’ search engine
s
v Benefits
• Increased product sales by improving content linking

Answering Burning Questions – EFL

v Objectives
• Extract numerical data
from case law to enhance
information access
for lawyers.
v Solution
• Luxid® with custom annotators (address, activity,
compensation, age, turnover… )
• Export numerical data as metadata to a search engine.
v Benefits
• Productivity gain to extract and validate metadata
• Allowing to treat huge amount of case law


Questions?
Thank you!
SSP 2009 Annual Meeting
Chris Beguel – Director of Sales – TEMIS
Baltimore, MD – May 09

1 d.3

More Related Content

Similar to 1 d.3 (20)

More from Society for Scholarly Publishing (20)

1 d.3