際際滷

際際滷Share a Scribd company logo

Lost In Translation: When Machines Meet STM Content

scrazzl
scrazzl

This is an edited version of the presentation that I gave at STM innovations #ukinno on December the 4th, 2013. It covers the Resources Identification Initiative and highlights some of the partners that are using the scrazzl product API to power research product discovery on their websites.

1 of 12
Download to read offline
Lost In Translation when machines meet STM content This presentation was designed to be delivered live. To help you understand the content I have added these notes International Association of Scientific, Technical & Medical Publishers The Voice of Academic and Professional Publishing
The Economist published a piece last month on the reproducibility crisis. The article was wide ranging in scope and explored a number of themes thought to contribute to the lack of experimental reproducibility
Referenced in the Economist piece was interesting work by Melissa Haendal et al., which speci鍖cally examined the relative identi鍖ability of material resources in a relevant sample of biomedical literature
Resource Identification Initiative This study reported that less than 50% of mentioned resources were uniquely identi鍖able by humans. The gradual recognition of this identi鍖ability problem as a signi鍖cant contributor to poor reproducibility has given rise to the Resource Identi鍖cation Initiative (#RII)
1. Unique Identi鍖er AB_1234578 A central aim of the RII is to develop a standard identi鍖cation nomenclature for all experimental resources (Commercial and Non-commercial)
2. Editor Awareness ≒ Drive adoption ≒ Better XML standards ≒ Content machine friendly Publishers have a signi鍖cant role to play in these e鍖orts. By simply improving and enforcing material and method author guidelines the relative identi鍖ability of experimental resources could be dramatically improved
3. Distribution As a company scrazzl/CompareNetworks are supporting the e鍖orts of the RII. Our business goals are to have third parties like scholarly publishers use our product data in their web applications
Quartzy This type of collaboration allows our partners to avoid the costs of managing their own product databases while growing ancillary revenues by using our product URLs.
JoVE JoVE and Quartzy are both using our product API to power product discovery on their websites. This saves them data management costs and allows them to grow ancillary revenue through syndication
Biocompare Our focus on product analysis in research content allows us to make the content of publishers we work with more discoverable. The Citation Feed service is used by product vendors to manage citations on their websites
Interested In ≒ Discussion about best practice in material and method sections. ≒ Users of product information APIs. ≒ Please join the e鍖orts to improve resource reporting by Data mining access to material and the joining the RII. You can learn more about our APIs and other ways that we work method sections. with publishers by going to developers.scrazzl.com
Questions? E: david.kavanagh@scrazzl.com T: +353 (0) 863-867-990 Twitter: @dkavanagh www.scrazzl.com

More Related Content

Lost In Translation: When Machines Meet STM Content

  • 1. Lost In Translation when machines meet STM content This presentation was designed to be delivered live. To help you understand the content I have added these notes International Association of Scientific, Technical & Medical Publishers The Voice of Academic and Professional Publishing
  • 2. The Economist published a piece last month on the reproducibility crisis. The article was wide ranging in scope and explored a number of themes thought to contribute to the lack of experimental reproducibility
  • 3. Referenced in the Economist piece was interesting work by Melissa Haendal et al., which speci鍖cally examined the relative identi鍖ability of material resources in a relevant sample of biomedical literature
  • 4. Resource Identification Initiative This study reported that less than 50% of mentioned resources were uniquely identi鍖able by humans. The gradual recognition of this identi鍖ability problem as a signi鍖cant contributor to poor reproducibility has given rise to the Resource Identi鍖cation Initiative (#RII)
  • 5. 1. Unique Identi鍖er AB_1234578 A central aim of the RII is to develop a standard identi鍖cation nomenclature for all experimental resources (Commercial and Non-commercial)
  • 6. 2. Editor Awareness ≒ Drive adoption ≒ Better XML standards ≒ Content machine friendly Publishers have a signi鍖cant role to play in these e鍖orts. By simply improving and enforcing material and method author guidelines the relative identi鍖ability of experimental resources could be dramatically improved
  • 7. 3. Distribution As a company scrazzl/CompareNetworks are supporting the e鍖orts of the RII. Our business goals are to have third parties like scholarly publishers use our product data in their web applications
  • 8. Quartzy This type of collaboration allows our partners to avoid the costs of managing their own product databases while growing ancillary revenues by using our product URLs.
  • 9. JoVE JoVE and Quartzy are both using our product API to power product discovery on their websites. This saves them data management costs and allows them to grow ancillary revenue through syndication
  • 10. Biocompare Our focus on product analysis in research content allows us to make the content of publishers we work with more discoverable. The Citation Feed service is used by product vendors to manage citations on their websites
  • 11. Interested In ≒ Discussion about best practice in material and method sections. ≒ Users of product information APIs. ≒ Please join the e鍖orts to improve resource reporting by Data mining access to material and the joining the RII. You can learn more about our APIs and other ways that we work method sections. with publishers by going to developers.scrazzl.com
  • 12. Questions? E: david.kavanagh@scrazzl.com T: +353 (0) 863-867-990 Twitter: @dkavanagh www.scrazzl.com