Semantic Web:
In Quest for the Next
Generation Killer Apps
Jie Bao
Tetherless World Constellation
Rensselaer Polytechnic Institute
baojie@cs.rpi.edu
http://www.cs.rpi.edu/~baojie
Oct 22nd, 2010 @ UMass Lowell 1
Outline
• Why Semantic Web?
• Key SW Standards
• Opening Data for SW
• Building SW Applications
2
What is a Killer App ?
3
A Killer App is
• …Any computer program that is so
necessary or desirable that it proves
the core value of some larger
technology, […]. A killer app can
substantially increase sales of the
platform on which it runs.
-- Wikipedia
4
http://en.wikipedia.org/wiki/Killer_application
Visicalc
5
Excel
6
Netscape
7
Facebook
8
Web Itself
9
Sir Tim Berners-Lee
As an app of the Internet
Picture Source: http://commons.wikimedia.org/wiki/File:InternetProtocolStack.png
ATM
10
A killer app of database and network technologies
Picture source: http://en.wikipedia.org/wiki/File:ATM_750x1300.jpg
Where are the killer apps
for the Semantic Web?
11
What is Semantic Web?
12
Web of Documents
13
from a
Web of Documents
to a
Web of Data
14
Web of Data
15
Travel Data
Web of Data
16
Financial Data
Web of Data
17
Housing Data
But data integration is difficult
18
and is often ad-hoc
19
and there are other issues
20
Inconsistency
21
Inconsistency
22
Inconsistency
23
Different Naming
24
#iswc2010 is #iswc
Different Naming
25
Inference
26
From: hotwire.com
Inference
27
From: travelocity.com
Can this be
automated?
(Li Ding teaches me this trick)
What We Need
• A standard data interchange format
• A standard representation of the
meaning of data
• A standard way to link data
28
Semantic Web Languages
29
Source: W3C's Semantic Web Activity / Semantic Web overview
http://www.w3.org/2007/Talks/0130-sb-W3CTechSemWeb/#%2824%29
The Layer Cake
RDF = Resource Description
Framework
30
affiliation
affiliation
knows
RDF
31
swrc:affiliation
swrc:affiliation
foaf:knows
<http://www.cs.rpi.edu/~baojie>
<http://www.rpi.edu>
< http://www.cs.rpi.edu/~hendler>
swrc:=http://swrc.ontoware.org/ontology#
foaf:=http://xmlns.com/foaf/0.1/
RDF
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
@prefix swrc: <http://swrc.ontoware.org/ontology#> .
@prefix foaf: <http://xmlns.com/foaf/0.1> .
@prefix rpics: < http://www.cs.rpi.edu/~> .
rpics:baojie swrc:affiliation <http://www.rpi.edu> .
rpics:baojie foaf:knows rpics:hendler .
rpics:hendler swrc:affiliation <http://www.rpi.edu> .
32
SPARQL= SPARQL Protocol And RDF
Query Language
A query language for RDF
SELECT ?person ?org
WHERE{
?person foaf:knows rpics:hendler .
?person swrc:affiliation ?org .
}
(find all people (and their affiliations) who know
Hendler)
33
OWL = Web Ontology
Language
• A much more powerful ontology language
• Examples (informally)
– If twitter:baojie and linkedin:baojie both use email
baojie@cs.rpi.edu, then they belong to the same
person.
– If Westin hotel is in Palo Alto, and Palo Alto is in
the Bay Area, then Westin hotel is in the Bay Area.
• OWL 1 (2004), OWL 2 (2009)
34
How they work in the real world?
35
You may not be aware that there are
already plenty of semantic data around.
36
BestBuy
37
http://www.bestbuy.com/site/The+Matrix+-+DVD/9316073.p?id=29857&skuId=9316073&st=9316073&lp=1&cp=1
BestBuy (GoodRelations)
38
http://products.semweb.bestbuy.com/products/9316073/semanticweb.rdf
Facebook
39
Facebook (Open Graph)
40
LinkedIn
41
http://www.linkedin.com/in/jiebao
LinkedIn (Microformat)
42
http://microformatique.com/optimus/?uri=http://www.linkedin.com/in/jiebao
SlideShare
43
http://www.slideshare.net/baojie_iowa/semantic-history-towards-modeling-and-publishing-changes-of-online-semantic-data
SlideShare (RDFa)
44
http://www.w3.org/2007/08/pyRdfa/extract?uri=http://www.slideshare.net/baojie_iowa/semantic-history-towards-modeling-and-publishing-
changes-of-online-semantic-data
IMDB
45
http://www.imdb.com/name/nm0000125/
IMDB (OG+Microformat)
46
http://www.w3.org/2007/08/pyRdfa/extract?uri=http://www.imdb.com/name/nm0000125/
Sig.ma (Data Aggregation)
47
http://sig.ma/search?q=Jie+Bao
Sig.ma
48
http://sig.ma/search?q=Jie+Bao
Data From Spreadsheet
49
Dominic DiFranzo and Li Ding - http://data-gov.tw.rpi.edu/wiki/Demo:_White_House_Visitor_Search
Data From Spreadsheet
50
6.46 billion RDF triples now.
Data From Relational DB
51
http://demo.openlinksw.com/about/html/http/demo.openlinksw.com/Northwind/Customer/ALFKI
52
Many Many More
Then, how semantic data help us to build
(killer) apps?
53
Semantic Twitter
54
Joshua Shinavier, TwitLogic SPARQL widget
Financial Data
55
Perry Grossman, Devin Mcqueeney, Graham G Rong, Lorin Wilde, Danny Yuan, Jie Bao
(coach). FinanceSphere, MIT LinkedData IAP 2010 Project
Financial Data
56
Bao, J., Rong, G., Li, X., and Ding, L. Representing Financial Reports on the Semantic Web - A
Faithful Translation from XBRL to OWL. In The 4th International Web Rule Symposium (RuleML).
2010
XBRL= eXtensible Business Reporting Language
RPI Map
http://map.rpi.edu
Jie Bao , Jin Guang Zheng, Rui Huang & Li Ding. Mesh-up Map and Events on Semantic Wiki: Applications in e-Science and Campus Information
System. SemanticWiki mini-series Session-4. Jan. 22, 2009. ontolog.cim3.net
Semantic Email
58
Data-gov Wiki
Li Ding and James Michaelis and Deborah L. McGuinness and Jim Hendler, Making Sense of
Open Government Data, in Proceedings of WebSci2010, 2010.
Data Mashup and Visualization
60
DiFranzo, D.: Developer Diary: CASTNET Ozone Map Demo. In WebSci 2010 poster.
http://data-gov.tw.rpi.edu/wiki/Demo:_Clean_Air_Status_and_Trends_-_Ozone
Data Mashup and Visualization
61
James Michaelis: http://data-gov.tw.rpi.edu/wiki/Demo:_Comparing_US-USAID_and_UK-DFID_Global_Foreign_Aid
Where are we?
62
Where are we?
63
Picture source: Wikipedia (http://en.wikipedia.org/wiki/Technology_adoption_lifecycle)
Adapted from “Semantic Web Adoption and Applications”, Ivan Herman, W3C. 2010-10-07 Slide 5
2005
Where are we?
64
Picture source: Wikipedia (http://en.wikipedia.org/wiki/Technology_adoption_lifecycle)
Adapted from “Semantic Web Adoption and Applications”, Ivan Herman, W3C. 2010-10-07 Slide 5
2010
The 2007 Gartner predictions
• By 2012, 80% of public Web sites will use some
level of semantic hypertext to create SW documents
[…] 15% of public Web sites will use more extensive
Semantic Web-based ontologies to create semantic
databases
• By 2017, we expect the vision of the Semantic Web
[…] to coalesce […] and the majority of Web pages
are decorated with some form of semantic hypertext.
65
“Finding and Exploiting Value in Semantic Web Technologies on the Web”, Gartner Report, May 2007
Imagine a world where data are linked
and make sense
66
On ATM
67
You can check transactions by their categories, and the ATM knows that not all items
from BestBuy are electronics (e.g., office supplies), since it is connected to the BestBuy
product database.
On TV
68
It can generate a personalized program list of movies starred by Sean Connery,
since it is connected to IMDB
In Your Car
69
Picture from: http://electronics.howstuffworks.com/gadgets/automotive/car-gps-accidents.htm/printable
The in-car GPS tells you attractions of the revolutionary war era, since it can read a semantic
version of Wikipedia and a geo-location database from the US government.
We are only limited by our imaginations
70
That’s why I believe the Semantic Web is
a beautiful thing
71
Thank you!
Slides are available @
http://slidesha.re/aimhRY
72

Semantic Web: In Quest for the Next Generation Killer Apps

Editor's Notes

  • #6 One of the first examples of a killer application is generally agreed to be the VisiCalc spreadsheet on the Apple II platform.[1] The machine was purchased in the thousands by finance workers (in particular, bond traders) on the strength of this program
  • #7 For Windows
  • #8 Once the most popular browser Promote WWW
  • #9 The currently most successful Social Network
  • #19 HTML, XML
  • #20 Expedia, Mint, Zillow all provide different solutions Reinvent the wheel
  • #39 http://products.semweb.bestbuy.com/products/9316073/semanticweb.rdf
  • #59 Lunch is social event
  • #61 For this demo, we used RDF data derived from ozone and visibility readings provided by the EPA&amp;apos;s Castnet project. This RDF combines raw data on Castnet readings (ozone and visibility) with corresponding geographic information on the sites of readings, which the raw Castnet data lacks. On the provided map of the US, every yellow dot represents a single Casetnet site and dot size corresponds to an average Ozone reading for that site, where larger dots represent larger averages. When a Castnet site is clicked, a small pop-up opens, displaying more information on that site, along with a link. This link takes the user to another page that displays in a timeline all the Ozone and Visibility data available for that site. This timeline uses the Google Visualization API giving users added functionality in how they view the data in the timeline. To allow users to filter through Castnet readings, a faceted browsing interface (from the MIT Simile Exhibit API) is provided.
  • #62 This application presents a mashup of foreign aid data (represented in US Dollars) from the United States Agency for International Development (USAID) and UK Department for International Development (DFID) for the 2007 US Fiscal Year.
  • #66 Gartner, Inc. (NYSE: IT) is the world&amp;apos;s leading information technology research and advisory company