Web 3.0 > Semantic Web > Web Object Retrieval Zhezhu Wen 2008.11.20
1
Brief Review - Web 2.0 • Collective wisdom oriented – Folksonomy, social classification, social tagging. – Wikipedia,
• People have own contents production tools – Digital camera, cam coder, high performance PC. – Youtube, user generated contents, flickr,
• Different contents access platform – Kindle, iPod, PC, Internet Tablet, Mobile phone, TiVo, game console,
• Age of the long tail – Even the weakest voice have its own stage – Blog, social network, podcast, 2
How to define Web 3.0 Various destination has been envisioned by researchers and industry leaders. • Tim Burners-Lee: Well organized web data • Dr. Eric Schmit: Cross platform, free, lightweight applications • Jerry Yang: Customizable application, blurred distinction in terms of S/W dev. • Reed Hastings: 10M of bandwidth to download movie. (founder of Netflix) 3
How to define Web 3.0 (ctnd.) • Idealized web world that is yet to come. • Some of the aspects are impossible to be true in current technological capacity. • But, different opinions converge in following point. – Web 3.0 as a set of technologies that offer different new ways to help computers organize and draw conclusions from online data.
4
Web of Data • Today’s web is nothing but data. • Rich contents, but overflowing, disorganized and chaotic. • Heavy noises preventing signals.
5
Semantic Web • Syntax: How you say something. • Semantic: Meaning behind what you say. – I love technology Vs. I technology. – Even syntax changes, still understand the meaning.
• Internet + Web + Search engine. How we can get better? • A web that is smarter, more intelligent, that actively help people with less noise, more signals. 6
Semantic Web (cntd.) • Categorizing web data. – Vision of Dewey and today’s web. – Metadata – data describes the contents. – Something we could learn from porn. website. • Systems of computer readable labels about content.
– RDF(Resource Description Framework) • A new system for locating and describing information.
– Ontologies • That would define relationships between classification categories. 7
Critics • Unrealistic to expect busy people and business to create enough metadata to make the work. – Countermeasure: make better tools for metadata
• How to relate those controversial words in ontologies? – Marriages Vs. monogamy, polygamy, samesex relationships etc. 8
Wrap up the semantic part. • Under the semantic web concept, web is… – Becoming more like a database – With Result oriented search.
• Business are already adapt the technologies in organizing and mining information for themselves. • Two big steps: – Making the World Wide Database – Training artificial intelligence to software. 9
Web Object Retrieval • Object has attributes and actions. • Object is multi-faceted concept. • Today’s search: – Searching single instance of attributes or actions. – One dimension data. – Organized search results again, in order to build up that object.
10
Web Object Retrieval (contd.) • Web object: – People, product, papers, places, – http://products.live.com - product – http://libra.msra.cn/Default.aspx - papers
• Revolutionary Example: – Social relation search: http://renlifang.msra.cn/ – “Obama’s cabinet ”, “Bush’s cabinet”, “Jesus” – Reminding of influence in social network search. 11
Conclusion • Common understanding resides in development world. – Whatever reason drives it… – Forms voluntarily…
• It is a wishful trend that we are going to… • Not required to every development entity. – Data mining technologies are backing up to fill the gap for those who didn’t or couldn’t participate. – With certain degree of errors… 12
REFERENCES Web 3.0, wikipedia.org A Smarter Web, MIT Technology Review, March 2007. Web Object Retrieval, WWW Conference, May 2007. Future of the World Wide Web, Tim Berners-Lee March 2007. • The Long Tail: Why the Future of Business is Selling Less of More, Chris Anderson, 2006 • • • •
13