Q: Is it possible                           to automate                            METADATA                            CRE...
Or, alternatively:                          Will I be replaced by a computer?                                          -or...
How does it work?                          There are 2 ways of automatically                          creating metadata:  ...
Extraction vs.                                  Har vesting                          Metadata extraction involves the mini...
What kind of metadata can be                    automatically created?           Best: Technical or Structural           (...
Why bother?           Lessens time and effort required (Burk et al.,           2007).           “The enormous volume of on...
Tim Berners-Lee            Inventor of the World Wide WebThursday, March 8, 2012
“It’s really important to have a lot of data.”                           “We haven’t got data on the Web as data.”        ...
A more efficient way to                     present data                          An example of the                       ...
A: Kind of/                          it depends...Thursday, March 8, 2012
Conclusions           No artificial intelligence yet!           Automated metadata creation can be used, but           onl...
Questions?Thursday, March 8, 2012
Upcoming SlideShare
Loading in …5
×

Metadata extraction

1,320 views

Published on

Can metadata extraction be automated?

Published in: Education, Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
1,320
On SlideShare
0
From Embeds
0
Number of Embeds
2
Actions
Shares
0
Downloads
17
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Metadata extraction

  1. 1. Q: Is it possible to automate METADATA CREATION?Thursday, March 8, 2012
  2. 2. Or, alternatively: Will I be replaced by a computer? -or- Should I have gone to school for computer science?Thursday, March 8, 2012
  3. 3. How does it work? There are 2 ways of automatically creating metadata: 1) Text mining/clustering “Extraction” 2) Machine learning techniques “Har vesting”Thursday, March 8, 2012
  4. 4. Extraction vs. Har vesting Metadata extraction involves the mining of resource content (text-mining) and employs sophisticated automatic indexing techniques to produce structured (“labelled”) metadata for object representation. Metadata har vesting relies on machine capabilities to collect tagged metadata previously created by humans, machine processing, or both. Library of Congress, AMeGA Project ReportThursday, March 8, 2012
  5. 5. What kind of metadata can be automatically created? Best: Technical or Structural (format, date, page #s)* OK: Descriptive (title, abstract)* Not-so-good: Semantic (keywords, subject matter) *Not so effective for when documents have special layouts or structures.Thursday, March 8, 2012
  6. 6. Why bother? Lessens time and effort required (Burk et al., 2007). “The enormous volume of online and digital resources makes semi-automatic metadata generation a critical need” (Park, & Lu, 2009). Alleviate the problems associated with “metadata bottleneck”. Better to start with something rather than nothing.Thursday, March 8, 2012
  7. 7. Tim Berners-Lee Inventor of the World Wide WebThursday, March 8, 2012
  8. 8. “It’s really important to have a lot of data.” “We haven’t got data on the Web as data.” “Data can... help us understand the world.” Tim Berners-Lee. (2009, February). “Tim Berners-Lee on the next Web.” TED Talk. <http://www.ted.com/talks/lang/eng/tim_berners_lee_on_the_next_web.html>Thursday, March 8, 2012
  9. 9. A more efficient way to present data An example of the automatic creation of data to be reused. Dates are extracted by Google and rearranged into a timeline.Thursday, March 8, 2012
  10. 10. A: Kind of/ it depends...Thursday, March 8, 2012
  11. 11. Conclusions No artificial intelligence yet! Automated metadata creation can be used, but only with human inter vention. Some metadata types are easier to automate. Automation of metadata creation is not widely used in libraries yet.Thursday, March 8, 2012
  12. 12. Questions?Thursday, March 8, 2012

×