XXIX Charleston Semantic Web (5 Nov 2009) Hulbert
1 like•657 views
The document discusses the importance and benefits of semantic technologies and the semantic web for scientific publishers and researchers. It notes that researchers prefer online journals but are reading fewer articles and for less time per article. Researchers want to quickly evaluate articles and isolate those worth deeper reading. Semantic tagging of content allows for more precise searching, related content linking, and discovery of hard to find information. Publishers are implementing semantic tools to enrich content by tagging entities, facts, and relationships to improve discoverability and navigation. Examples from the American Institute of Physics and Royal Society of Chemistry are provided that leverage semantic tagging of content.
![Thank you. Terry Hulbert Director of Business Development XXIX Charleston Conference Charleston, SC 5 November 2009 [email_address] www.twitter.com/thulbert www.friendfeed.com/tcjh007 www.delicious.com/tcjh007](https://image.slidesharecdn.com/xxixcharlestonsemanticweb5nov2009hulbert-12578038236603-phpapp02/85/XXIX-Charleston-Semantic-Web-5-Nov-2009-Hulbert-22-320.jpg)
XXIX Charleston Semantic Web (5 Nov 2009) Hulbert
- 1. The Semantic Web: what you need to know and why it’s important for your user community Terry Hulbert Director of Business Development American Institute of Physics XXIX Charleston Conference Charleston, SC 5 November 2009
- 2. Contents Why is semantic technology a hot topic? What can we do and what are some doing? Some examples at AIP
- 3. Why? 98% of STM researchers prefer & use online journals (Hemminger et al, 2007) No. of articles read per researcher was 30% higher in ‘06 than ’96 (Tenopir & King, 2008) Reading time per article is falling, from 32 minutes in ‘96, to 24 in ‘06 (Tenopir & King, 2008)
- 4. Why? Researchers are actually trying to avoid reading as much as possible during “discovery phase”, preferring instead to evaluate article “components”…indexing terms, related content, citations, and other “indicators of value” to isolate articles worth their “real” reading time.
- 5. Why? “ Scientists skim journal articles to discover valuable information. They scan for terminology, segments, diagrams, and summaries of particular interest. But they don’t read individual articles left-to-right, top-to-bottom” Allen H. Renear and Carole L. Palmer, Science 325 (5942), 828 (14 August 2009) doi: 10.1126/science.1157784
- 6. Online reading “ Reading” an online journal approximates a bookstore visit – a place to skim, flit, discover, fritter, twitter, connect, talk, be surprised, and socialize – but most of all, not to read too deeply during initial search and discovery mode .
- 7. Strategic reading; power browsing Preference for atomised/deconstructed article “nuggets” that reveal the essence of the science within. Trend toward scanning for critical ‘cues & clues’, ‘hooks & hints’ to “flash determine” the value of an article – is it worthy of a closer, deeper read?” Scanning for accurate signposts to the literature, not the literature… not yet (“close reading” reserved for articles of interest found along the way)
- 8. How to help strategic reading Faceting Authors, topics, popularity,… Data visualisation Citation density, author connections, geo viz Related linking, ‘More Like This’ “ Bite”-serving Article extracts, snippets, image thumbnails, video/audio, trending
- 9. How are publishers doing this? Reengineering production workflows Implementation of semantic engine(s) & tools e.g. Temis, Silverchair, Access Innovations, Collexis automatic categorisation intelligent association taxonomy management Enhanced XML repository (MarkLogic) Content enrichment, text mining, entity & fact extraction, etc.
- 10. Content enrichment Highlight the meaning of content Enhances granularity e.g. book, chapter, image, topics, etc. Taxonomy is key for normalisation e.g. PACS, MeSH, MSCs
- 11. What is the output? Entities products, people, places, technologies, organisations, companies infer facts related to relationships, functions,… Richer, semantic content linking and tagging easier navigation better discoverability
- 12. What will this enable? Search precision reveal hard-to-find objects and information New products e.g. image/video/animation/multimedia collections Microproducts; mash-ups VJs, learning objects, etc. Profiling; social networking; personalisation
- 13. AIP examples AIP UniPHY social networking site for physical sciences some obvious uses of granularity plus ‘scientific context”
- 18. AIP examples Geotagging research by geo area geophysics – seismic activity plotted or plot an area and highlight research published Tag products e.g. Review of Scientific Instruments
- 20. Other examples - chemistry Royal Society of Chemistry ChemSpider (acquired May 2009) aggregating and indexing chemical structures and associated information literature, chemical vendor catalogs, molecular properties, environmental data, toxicity data, analytical data, etc. Search by chemical reactions, solvents, reactants – all extracted from documents
- 21. Summary Rapid review of: why semantic tagging, semantic technologies and the semantic Web is becoming increasingly important top level view of what’s possible some examples more detailed knowledge & expertise from my colleagues
- 22. Thank you. Terry Hulbert Director of Business Development XXIX Charleston Conference Charleston, SC 5 November 2009 [email_address] www.twitter.com/thulbert www.friendfeed.com/tcjh007 www.delicious.com/tcjh007