2. Agenda
- Introduction to our
solutions at Elsevier
- Emerging needs around
content metrics
- Emerging trends in
visualization needs
3. OUR CUSTOMERS OPERATE IN AN INTEGRATED ECOSYSTEM WITH
DRUG R&D AT ITS CORE
Academia &
Government
Medical
Devices
Diagnostics
Companies
Pharma &
Biotech
TARGET ID &
VALIDATION
LEAD ID &
VALIDATION
PRE-CLINICAL
/ CLINICAL
POST-
MARKET
Characterize &
understand disease
Design effective
approach & validate
lead
Cull leads more
quickly for safety &
efficacy
Manage risk &
compliance &
improve patient care
4. Target ID &
Valid
Pre-clinical Clinical
Post-
Launch
Lead ID &
Valid
Target Lead (drug)
Questions
addressed
How do we
monitor the
commercialized
leads for adverse
events?
How do we assess and prioritize
the drug ideas for safety, delivery
and efficacy?
What drugs
can hit the
identified
targets?
What are the
potential
targets related
to the
disease?
BUT FIRST SOME BACKGROUND: OUR CUSTOMER WORKFLOW
Leveraging capability
- Text and data mining
- Integration
- Informatics
5. | 5
SCIENTIFIC LITERATURE IS EXPLODING
• Rapidly approaching 1M new citations/year in Medline – HOW TO KEEP UP?
0
100000
200000
300000
400000
500000
600000
700000
800000
900000
1000000
1940 1950 1960 1970 1980 1990 2000 2010 2020
Yearly Citation Count Totals from 2014 MEDLINE
(Publication Date Used for Categorization)
6. EMERGING TRENDS - CASE STUDY FROM BIOLOGY
Research facts
(literature)
Collated
facts
(extracted)
Quality
Metrics to
select
EVERY YEAR MILLIONS OF PAPERS ARE
PUBLISHED
WE EXTRACT 2-3 MILLION “FACTS” PER
YEAR
UP TO 70% OF RESEARCH CANNOT BE
REPRODUCED
HOW DO WE HELP OUR RESEARCHERS
DECIDE WHAT DATA TO TRUST?
7. EMERGING TRENDS - CASE STUDY FROM BIOLOGY
Research facts
(literature)
Collated
facts
(extracted)
Quality
Metrics to
select
• EVERY YEAR MILLIONS OF PAPERS ARE
PUBLISHED
• WE EXTRACT 2-3 MILLION “FACTS” PER
YEAR
• UP TO 70% OF RESEARCH CANNOT BE
REPRODUCED
How can I
decide which
facts to
believe?
Can you tell me
which methods
were used to
get to this fact?
I need the raw
data so I can re
run the
statistics
I need to know
who the author
worked with
I need to know
how well
respected this
author is
8. EXAMPLE USE CASE: BIOMARKERS
Perform
search for
all relevant
facts (text
mining)
Select
criteria to
sort or
facet data
further
Re mine
data for
new
criteria
Visualize
and select
for best
output
Metrics commonly requested in this approach are
• number of times cited
• methods used
• h-index of an author
• Cluster of publication year and tend
*
9. WHICH METHODS WERE USED – BIOMARKER EXAMPLE
• Gaphically view
research trends on
biomarkers in the
universe of facts
• Summarizes method
used, and
publication and
number of
references
• Further allows user
to dig down into
overlapping
publications and
verify utility of the
biomarker
10. KNOWLEDGE MAPS TO FIND THE RIGHT COLLABORATORS
Identified potential collaborators
in specific scientific
technologies by mapping
research landscape