Scientific Profile Generation
K-Extractor™: Scientific landscape at a glance
Customer Scenario
Need
Know the top contributors in the virology and microbiology field, their current work, their affiliations, and who is funding their work.
Big Data: Volume, Variety & Velocity. Thousands of new papers are published every day, collecting this information and keeping it up to date is no longer feasible to do by hand.
Challenge
Lymba's Solution
Solution
K-Extractor™ automatically extracts knowledge centered on authors, their research topics and research funding to generate consolidated interlinked profiles for each person.
Input
Thousands of scientific research publications in PDF format, web pages with bio pages, conference schedules.
Scientific profiles - creating in depth "baseball cards" for each person with the following information:
Employment and affiliation history
Projects, co-workers and co-authors
Funding and awards
Education
Topics of interest
Family and personal information
Contact information
Each item on profile has a direct link to supporting documents from input collections. Profiles are connected to each other with hypertext links.
Output
Key Features
Document structure recognition for PDFs and other complex file formats to detect the following elements:
– title, authors and affiliations
– headers and footers
– section titles and content
– citations and references
– tables and illustrations
Deep semantic processing of the text fragments to extract the concepts and semantic relations of interest.
Resolving aliases of people, organizations and other concepts: spelling variations, short forms, synonyms, etc. Merging concepts together enables the combination of knowledge from multiple sources.
Extracted knowledge is saved into an RDF store that is available to the customer. Lastly, the RDF store is automatically queried to generate connected profiles for each person mentioned in the input collection.
Check out K-Extractor for more details on the product and its features.
Or request a demo with us: