Engineering a Semantic Web (Spring 2018)

Engineering a Semantic Web
ITWS Capstone (Spring 2018)
John S. Erickson, Ph.D
The Rensselaer IDEA
Tetherless World Constellation

Understanding...
Consuming...
Building...
Doing Science with...
...the Web of Data

Objectives
● Deeper understanding of Web architecture
● Understand the Semantic Web "stack"
● Review Linked Data principles & practices
● Build cool applications
● Contribute to the Web of Data
● Make better IT management decisions

1. From the Web of Documents to the Web of Data
2. Linked Data: Building Blocks of the Web (of Data)
3. Consuming: Semantics-driven analytics
Perspective: Consuming the Web of Data

Web Architecture
First Principles of The Web:
● A standard system for identifying resources
● Standard formats for representing resources
● A standard protocol for exchanging resources

Web Architecture
First Principles of The Web:
Relevant core standards:
● URIs (URLs): Uniform Resource Identifiers
● HTML: Hypertext Markup Language
● HTTP: Hypertext Transfer Protocol
https://www.w3.org/History/1989/proposal.html

What's a "Resource?"
A resource can be anything that has identity. Familiar examples
include an electronic document, an image, a service (e.g., "today's
weather report for Los Angeles"), and a collection of other resources.
Not all resources are network "retrievable"; e.g., human beings,
corporations, and bound books in a library can also be considered
resources. The resource is the conceptual mapping to an entity or set of
entities, not necessarily the entity which corresponds to that mapping at
any particular instance in time. Thus, a resource can remain constant
even when its content---the entities to which it currently
corresponds---changes over time, provided that the conceptual
mapping is not changed in the process. [1]
1. Tim Berners-Lee, "Uniform Resource Identifiers (URI): Generic Syntax" (RFC 2396) Aug 1998

Architecture of the World Wide Web, Volume One http://www.w3.org/TR/webarch/

https://www.wunderground.com/mx/oaxaca
Identifies
Represents

Data Mining: Mapping the Blogosphere http://bit.ly/18MuXdD

Identifying Web Resources (1)
● A global identification system is essential...
○ ...to share information about resources
○ ...to reason about resources
○ ...to modify or exchange resources
● Resources are anything that can be linked to or spoken of
○ Documents, weather reports, cat videos, people, ideas...
● Not all resources are "on" the Web
○ They might be referenced by Web resources...
○ ...while not being retrievable from it
○ These are (so called) information resources
Les Carr, et.al. http://slidesha.re/142MFrV

● A global standard is required; the URI is it
● Others systems are possible...
○ ...but added value of a single global system of identifiers is high
○ Enables linking, bookmarking and other functions across
heterogeneous applications
● How are URIs used?
○ All resources have URIs: “A resource...has identity”
○ Each URI identifies a single resource
in a context- independent manner
○ URIs act as names and (usually) addresses
○ In general URIs are "opaque"
Uniform Resource Identifier (URI): Generic Syntax (RFC 3986) http://www.ietf.org/rfc/rfc3986.txt

● "URIs identify and URLs locate..."
○ ...and identify
● URLs are URIs aligned with protocols
○ URLs include the "access mechanism" or "network location"
e.g. http:// or ftp://
○ How to "dereference" the URI and retrieve the thing
● URL examples
○ ftp://ftp.is.co.za/rfc/rfc1808.txt
○ http://www.ietf.org/rfc/rfc2396.txt
○ mailto:John.Doe@example.com
○ telnet://192.0.2.16:80/
Uniform Resource Identifier (URI): Generic Syntax (RFC 3986) http://www.ietf.org/rfc/rfc3986.txt

Representing Resources (1)
● "Resources" are manifest as digital files
○ More precisely: serializations that look like files...
● The Web recognizes a (growing) set of {file | serialization} formats
○ The original and workhorse is HTML...
○ ...but there are many others
● Retrievable resources on the web serve multiple purposes
○ Resources encode information and data
○ Resources aggregate links to other resources
● This is what makes The Web(tm) a "web..."

Resources (nodes)
aggregate links to
other resources to
create a Web

Retrieving Resources (1)
● Review: URLs refer to retrievable resources
○ ie URIs that specify some protocol for retrieval
● The original and most common Web protocol is HTTP
● Specialized protocols are possible but resources may
appear "off the grid..."
● More common case is HTTP w different formats...

URIs, HTTP, many formats...
http://www.w3.org/2006/Talks/0521-sb-AC-management/ReCTechStack-bg.png

Principles for creating a "healthy" Web
Tim Berners-Lee http://www.w3.org/DesignIssues/LinkedData.html
● Use URIs as names for things
● Use HTTP URIs so "people" can look up those names
● When "someone" looks up a URI, return useful information
○ use standard representation formats to express it
● Include links to other URIs, so consumers can discover more things
○ By "consumers" we mean people or applications
Why is linking important???

Implications of a well-connected Web
Links to other nodes
are a "vote" of quality
and/or relevance
PageRank https://en.wikipedia.org/wiki/PageRank
Google PageRank

Implications of a well-connected Web
Links to other nodes
are a "vote" of quality
and/or relevance
PageRank https://en.wikipedia.org/wiki/PageRank
Google PageRankhttp://bit.ly/2mpeADL

What is this "Semantic Web?"
...and where can I get one???

Engineering a Semantic Web (Spring 2018)

Web of "meaning”
Web of Data
Linked Data
Linking ideas...
Semantic Web?

http://bit.ly/195LrMz

subject object
predicate
RDF: Resource Description Framework

subject object
“article”
“James
Hendler”
predicate
“has creator”
RDF: Resource Description Framework

http://dbpedia.org/resource/James_Hendler
doi:10.1109/MC.2009.30
http://purl.org/dc/elements/1.1/creator
http://dx.doi.org//10.1109/MC.2009.30

http://dbpedia.org/resource/James_Hendle
r
doi:10.1109/MC.2009.30
http://dx.doi.org//10.1109/MC.2009.30

We're missing something...
● Check: URIs for names: S, P, O can be URIs
● Check: HTTP URIs: all of our examples are resolvable
● Now: "Return something useful" when we resolve URIs
○ Format: How do we serialize RDF?
○ Protocol: How do we retrieve RDF?
Let's go to the graph...

Source: Programming the Web
http://bit.ly/1aZwr40

"Raw" Triples
Via the W3C RDF VCalidator Service:
http://www.w3.org/RDF/Validator/

N-Triples...
<http://kiwitobes.com/toby.rdf#ts> <http://xmlns.com/foaf/0.1/homepage> <http://kiwitobes.com/>.
<http://kiwitobes.com/toby.rdf#ts> <http://xmlns.com/foaf/0.1/nick> "kiwitobes".
<http://kiwitobes.com/toby.rdf#ts> <http://xmlns.com/foaf/0.1/name> "Toby Segaran".
<http://kiwitobes.com/toby.rdf#ts> <http://xmlns.com/foaf/0.1/mbox> <mailto:toby@segaran.com>.
<http://kiwitobes.com/toby.rdf#ts> <http://xmlns.com/foaf/0.1/interest> <http://semprog.com>.
<http://kiwitobes.com/toby.rdf#ts> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type>
<http://xmlns.com/foaf/0.1/Person>.
<http://kiwitobes.com/toby.rdf#ts> <http://xmlns.com/foaf/0.1/knows> _:jamie .
<http://kiwitobes.com/toby.rdf#ts> <http://xmlns.com/foaf/0.1/knows> <http://semprog.com/people/colin>.
_:jamie <http://xmlns.com/foaf/0.1/name> "Jamie Taylor".
_:jamie <http://xmlns.com/foaf/0.1/mbox> <mailto:jamie@semprog.com>.
_:jamie <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://xmlns.com/foaf/0.1/Person>.
<http://semprog.com/people/colin> <http://xmlns.com/foaf/0.1/name> "Colin Evans".
<http://semprog.com/people/colin> <http://xmlns.com/foaf/0.1/mbox> <mailto:colin@semprog.com>.
<http://semprog.com/people/colin> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type>
<http://xmlns.com/foaf/0.1/Person>.
<http://semprog.com> <http://www.w3.org/2000/01/rdf-schema#label> "Semantic Programming".
<http://semprog.com> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://xmlns.com/foaf/0.1/Document>.

N3...
@prefix foaf: <http://xmlns.com/foaf/0.1/>.
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>.
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>.
@prefix semperp: <http://semprog.com/people/>.
@prefix tobes: <http://kiwitobes.com/toby.rdf#>.
tobes:ts a foaf:Person;
foaf:homepage <http://kiwitobes.com/>;
foaf:interest <http://semprog.com>;
foaf:knows semperp:colin,
[ a foaf:Person;
foaf:mbox <mailto:jamie@semprog.com>;
foaf:name "Jamie Taylor"];
foaf:mbox <mailto:toby@segaran.com>;
foaf:name "Toby Segaran";
foaf:nick "kiwitobes".
<http://semprog.com> a foaf:Document;
rdfs:label "Semantic Programming".
semperp:colin a foaf:Person;
foaf:mbox <mailto:colin@semprog.com>;
foaf:name "Colin Evans".

RDFa...
:
<div xmlns:foaf="http://xmlns.com/foaf/0.1/" about="http://kiwitobes.com/toby.rdf#ts" typeof="foaf:Person">
Name: Toby Segaran 
Nickname: kiwitobes 
Interests: <a rel="foaf:interest" href="http://semprog.org">
Semantic Programming</a>
Homepage: <a rel="foaf:homepage" href="http://kiwkitobes.com/">KiwiTobes</a>
Friends: 
<ul rel="foaf:knows">
<li about="http://semprog.com/people/colin"
typeof="foaf:Person" property="foaf:name">Colin Evans</li>
<li typeof="foaf:Person">
Jamie Taylor 
Email: <a rel="foaf:mbox" href="mailto:jamie@semprog.com">
jamie@semprog.com</a> 
</li>
</ul>
</div>
:

RDF/XML... <rdf:RDF
xmlns:foaf='http://xmlns.com/foaf/0.1/'
xmlns:rdf='http://www.w3.org/1999/02/22-rdf-syntax-ns#'
xmlns:rdfs='http://www.w3.org/2000/01/rdf-schema#'>
<foaf:Person rdf:about="http://kiwitobes.com/toby.rdf#ts">
<foaf:name>Toby Segaran</foaf:name>
<foaf:homepage rdf:resource="http://kiwitobes.com/"/>
<foaf:nick>kiwitobes</foaf:nick>
<foaf:mbox rdf:resource="mailto:toby@segaran.com"/>
<foaf:interest>
<foaf:Document rdf:about="http://semprog.com">
<rdfs:label>Semantic Programming</rdfs:label>
</foaf:Document>
</foaf:interest>
<foaf:knows>
<foaf:Person rdf:about="http://semprog.com/people/colin">
<foaf:name>Colin Evans</foaf:name>
<foaf:mbox rdf:resource="mailto:colin@semprog.com"/>
</foaf:Person>
</foaf:knows>
<foaf:knows>
<foaf:Person>
<foaf:name>Jamie Taylor</foaf:name>
<foaf:mbox rdf:resource="mailto:jamie@semprog.com"/>
</foaf:Person>
</foaf:knows>
</foaf:Person>
</rdf:RDF>

JSON-LD...
{
"@graph": [
{
"@id": "http://kiwitobes.com/toby.rdf#ts",
"@type": "http://xmlns.com/foaf/0.1/Person",
"http://xmlns.com/foaf/0.1/homepage": {
"@id": "http://kiwitobes.com/"
},
"http://xmlns.com/foaf/0.1/interest": {
"@id": "http://semprog.com"
},
"http://xmlns.com/foaf/0.1/knows": [
{
"http://xmlns.com/foaf/0.1/mbox": {
"@id": "mailto:jamie@semprog.com"
},
"http://xmlns.com/foaf/0.1/name": "Jamie Taylor"
},
{
"@id": "http://semprog.com/people/colin"
}
],
"@id": "mailto:toby@segaran.com"
},
"http://xmlns.com/foaf/0.1/name": "Toby Segaran",
"http://xmlns.com/foaf/0.1/nick": "kiwitobes"
},
{
"@id": "http://semprog.com/people/colin",
"@id": "mailto:colin@semprog.com"
},
"http://xmlns.com/foaf/0.1/name": "Colin Evans"
},
{
"@id": "http://semprog.com",
"@type": "http://xmlns.com/foaf/0.1/Document",
"http://www.w3.org/2000/01/rdf-schema#label": "Semantic Programming"
}
]
}

RDF/JSON...
{
"http://kiwitobes.com/toby.rdf#ts": {
"http://xmlns.com/foaf/0.1/nick": [
{
"type": "literal",
"value": "kiwitobes"
}
],
"http://www.w3.org/1999/02/22-rdf-syntax-ns#type": [
{
"type": "uri",
"value": "http://xmlns.com/foaf/0.1/Person"
}
],
"http://xmlns.com/foaf/0.1/interest": [
{
"type": "uri",
"value": "http://semprog.com"
}
],
"http://xmlns.com/foaf/0.1/knows": [
{
"type": "uri",
"value": "http://semprog.com/people/colin"
},
{
"type": "bnode",
"value": "_:N40b366148cfc4c48a80f4e15acbd2858"
}
],
"http://xmlns.com/foaf/0.1/mbox": [
{
"type": "uri",
"value": "mailto:toby@segaran.com"
}
],
"http://xmlns.com/foaf/0.1/homepage": [
{
"type": "uri",
"value": "http://kiwitobes.com/"
}
],
"http://xmlns.com/foaf/0.1/name": [
{
"type": "literal",
"value": "Toby Segaran"
}
]
},
"http://semprog.com/people/colin": {
{
"type": "uri",
"value": "mailto:colin@semprog.com"
}
],
{
"type": "literal",
"value": "Colin Evans"
}
],
{
"type": "uri",
}
]
},
"_:N40b366148cfc4c48a80f4e15acbd2858": {
{
"type": "uri",
"value": "mailto:jamie@semprog.com"
}
],
{
"type": "literal",
"value": "Jamie Taylor"
}
],
{
"type": "uri",
}
]
},
"http://semprog.com": {
{
"type": "uri",
"value":
"http://xmlns.com/foaf/0.1/Document"
}
],
"http://www.w3.org/2000/01/rdf-schema#label": [
{
"type": "literal",
"value": "Semantic Programming"
}
]
}
}

RDF Serialization: Summary
● N-Triples: Verbose, "pedagogical"
● N3: Concise, in common use;
● RDFa: Commonly used for embedded RDF
● RDF/XML: Some use in government & "enterprise"
● JSON-LD: Fast-rising LD standard
● RDF/JSON: Older convention for LD applications

RDF Serialization: Summary
● N-Triples: Verbose, "pedagogical"
● N3: Concise, in common use;
● RDFa: Commonly used for embedded RDF
● RDF/XML: Some use in government & "enterprise"
● JSON-LD: Fast-rising LD standard
● RDF/JSON: Older convention for LD applications
Google, "Introduction to Structured Data" http://bit.ly/2mpaJXa

Things we still haven't discussed...
● How to retrieve this "linked data" of which I speak
● How (and where?) to query RDF "graphs"
● How to use LD in applications
● How to create visualizations, "mashups," etc
Also:
● How to create and publish linked data...

Consuming Linked Data
● Querying RDF using SPARQL
● Endpoints and triple stores
● Mashing data in the query
● Mashing data in the application

CBIS Analytics:
Studying the Impact of
Active Bioremediation on
Respiration, Heart Rates
and Executive Function
using an Integrated
Knowledge Graph /
Analytics / Visualization
Environment

CBIS AMPS Experimental Layout
DATA
CASE Hadatac

Data INCITE CBIS Analysis Framework (Summer 2017)
CASE Hadatac
UI-based CBIS Data Selection (non-cognitive)
RDF Load
SPARQLQueries
(manual,viaUI)
CSV Download
CBIS Experiment
Knowledge Graph
Critical metadata
relating measures
across sources
SPARQLQueries
(DirectfromR)
"Auxiliary" Data
Experimental Log
Executive Function Data
Cortisol and sAA Data
CSV
SPARQL
Query
Results
Native R
Data Frame
Linked Data (RDF) Conversion
(Python, SETLr, et.al.)
Multi-Window Shiny
On RPI Campfire

Data representation challenges
1. Data collected at different rates and time
a. Environmental data from HOBO sensor were collected every 10 seconds.
b. Physiological data from Hexoskin vest collected every second.
c. Physiological data from Sphygmomanometer were collected twice per day.
d. Cognitive only have test results ("Total Correct" etc)
2. Missing data especially physiological data
a. People didn’t show up
b. Hexoskin vest didn’t work properly
3. Different data types
a. "Real time" readings from HOBO sensor and Hexoskin
b. Cognitive test results
c. Saliva swabs for stress measurement (Cortisol, sAA)
4. Different experimental duration each day
a. Day 1: 85 minutes
b. Day 2: 75 minutes
c. Day 3: 60 minutes

The CBIS Analytics Graph
Physical
Environment
Executive
Function (Cognitive)
Physiological
Stress (Saliva)
Unified CBIS
Knowledge Graph

@prefix foaf: <http://xmlns.com/foaf/> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
@prefix hadatac: <http://hadatac.org/2017/05/hadatac#> .
@prefix oboe: <https://semtools.ecoinformatics.org/oboe#> .
<http://incite.org/cbis_feb_2017/SGQ/Sample_1> a oboe:Measurement ;
hadatac:hasTimestamp "2017-02-07T12:45:28"^^xsd:dateTime ;
hadatac:hasDay_Date <http://incite.org/cbis_feb_2017/Day_1_2_7_17> ;
hadatac:hasExperiment_Phase <http://incite.org/cbis_feb_2017/Group_A> ;
hadatac:hasObject_id <http://hadatac.org/kb/case#SBJ-CBIS-Feb-2017-04> ;
hadatac:hasValue "13" ;
hadatac:hasElevation "0.0" ;
hadatac:hasCharacteristic "STICSA-G Somatic" ;
hadatac:hasCharacteristic_uri <http://hadatac.org/ont/case#ATTYP-SGQ-somatic> ;
hadatac:hasUnit "dimensionless unit" ;
hadatac:hasUnit_uri <http://geneontology.org/GO.format.obo-1_2.shtml#UO_0000186> ;
hadatac:hasInstrument_model "SGQ: SG Questionnaire" ;
hadatac:hasInstrument_uri <http://hadatac.org/kb/case#INS-SGQ> ;
hadatac:hasPlatform_name "The Platform" ;
hadatac:hasPlatform_uri <http://hadatac.org/kb/case#CH-AQ-CBIS-Feb-2017> ;
hadatac:hasEntity "human" ;
hadatac:hasEntity_uri <http://semanticscience.org/resource/Human> .
Example N3 Encoding (Executive Function result)

SPARQL: Pattern matching over RDF
graphs

?s ?blackboard
dbpedia2:blackboard
http://bit.ly/RumkhW

http://dbpedia.org/resource/Double,_Double,_Boy_in_Trouble

CBIS Analytics Example Queries

Available Variables and Measurements
All subject types (no differentiation between human, air, etc)
"9 variables available"
List of available variables
Available samples/variable

Querying Available Categories of Indicator
* Requires annotation of indicator to aggregate by category
"Available samples from this category"
Indicators from this
category
Instruments producing this
category of measurement

Available Variables by Category
Physiology Indicator (based on ontology*)
* Requires annotation of indicator to
aggregate by category
Select all characteristics of type
"Physiological Indicator"

Physical Environment Indicator (based on ontology*)
"Physical Environment Indicator"
"3 variables of this category available"
Available samples/variable

Executive Function Indicator (based on ontology*)
"Executive Function Indicator"
"82 variables of this category available"
Available samples/variableToo many variables in this
category to list completely...

Available Subjects and Measurements
Humans as "Subjects"
Selecting a particular subject
(ie entity) type: Human
Artifact of our graph
loading process (due
to missing URIs in the
dataset)

Available Subjects and Measurements
Air as the "Subject"*...
There is actually only one unique room; we see six as an artifact of missing URIs combined with upload issues

List all available Instruments

Available Date Range
min/max dateTimes by subject (ie entity) type
MAX timestamps for each type
of subject ("Human" and "Air")
MAX timestamps for each type
of subject ("Human" and "Air")

Variable Range (max/min/avg)
Aggregation
Selected
Aggregation Selected
Variable Selected
Variable MAX (per-subject) Variable MIN (per-subject)

Calling from R is easy!
Define query Submit query
Plot dataframe

"Next Time": Building the Web of Data...
● Converting datasets to RDF
● Hosting: Triplestores & endpoints
● Enterprise use cases
● Advanced techniques
● Web Science...

Bonus Slides:
Building the Web of Data

Bonus: Building the Web of Data
1. Review: the Web of Data
2. Publishing the Web of Data
3. Engineering the Web of Data in the Enterprise
4. Enterprise Applications of Semantic Technologies
5. Advanced "Semantic Web" concepts
6. Web Science: Observing and (re)Engineering the Web

First Principles of the Web...
Relevant core standards:
● URIs (URLs): Uniform Resource Identifiers
● HTML: Hypertext Markup Language
● HTTP: Hypertext Transfer Protocol
Review: Web Architecture

Review: Linked Data Principles
Tim Berners-Lee http://www.w3.org/DesignIssues/LinkedData.html
● Use URIs as names for things
● Use HTTP URIs so people can look up those names
● When someone looks up a URI, return useful information
○ use standard representation formats to express it
● Include links to other URIs, so consumers can discover more things
○ By "consumers" we mean people or applications

Now: Publishing the Web of Data
Recall our triples from previous lecture...

Publishing the Web of Data
Recall our triples from previous lecture...
To be useful, this data must be
loaded in a triple store and
published via a web-accessible
SPARQL endpoint

Industrial-strength Triple stores
1. AllegroGraph (1+Trillion)
2. OpenLink Virtuoso v6.1 - 15.4B+ explicit; uncounted virtual/inferred
3. BigOWLIM (12B explicit, 20B total); 100,000 queries per $1
4. Garlik 4store (15B)
5. Bigdata(R) (12.7B)
6. YARS2 (7B)
7. Jena TDB (1.7B)
8. Jena SDB (650M)
9. Mulgara (500M)
10. RDF gateway (262M)
11. Jena with PostgreSQL (200M)
12. Kowari (160M)
13. 3store with MySQL 3 (100M)
14. Sesame (70M)
15. Others who claim to go big
TWC uses Virtuoso Open Source edition

Industrial-strength Triple stores
1. AllegroGraph (1+Trillion)
2. OpenLink Virtuoso v6.1 - 15.4B+ explicit; uncounted virtual/inferred
3. BigOWLIM (12B explicit, 20B total); 100,000 queries per $1
4. Garlik 4store (15B)
5. Bigdata(R) (12.7B)
6. YARS2 (7B)
7. Jena TDB (1.7B)
8. Jena SDB (650M)
9. Mulgara (500M)
10. RDF gateway (262M)
11. Jena with PostgreSQL (200M)
12. Kowari (160M)
13. 3store with MySQL 3 (100M)
14. Sesame (70M)
15. Others who claim to go big
You can install Apache Jena yourselves!

Publishing: RDBMS to RDF
● Advantage: Leveraging "legacy"
sources
● Challenge: Complexity...
● Example: D2RQ Platform
○ D2RQ Mapping Language
○ D2RQ Engine
○ D2R Server
See also: http://d2rq.org/

Publishing: Linked Data API
Motivations:
● SPARQL, RDF have high learning curves
● RDF support in the common web
development tool stacks is scarce
● Solution is the Linked Data API
Advantages:
● Easy to use web API on linked data
● Allows publisher to provide URIs for lists of
things
● Allows users to get back the data as JSON,
XML, or RDF
● Easy to filter data using simple URL query
parameters
Makes it easy to create web applications over the
published data using standard tools
https://code.google.com/p/linked-data-api/wiki/Specification

Linked Data
API Example
UK Bathing Water Data Explorer
Live: http://environment.data.gov.uk/lab/bwq-web
Details: http://www.epimorphics.com/web/projects/bathing-water-quality

Architecture of Linked Data Applications
● The Crawling Pattern
● The On-The-Fly Dereferencing Pattern
● The Query Federation Pattern
The architecture of a Linked Data application
depends on its driving use case

The Crawling Pattern
● Applications "crawl" the Web of Data in advance by traversing RDF links
● Integrate and cleanse discovered data
● Provide higher layers of the application with an integrated view of the original data
● Mimics the architecture of classical Web search engines like Google and Yahoo
● Suitable for implementing applications on top of an open, growing set of sources
○ new data sources are discovered by the crawler at run-time.
● Separates the tasks of building up the cache and using cache later
○ enables applications to execute complex queries with reasonable performance over large
amounts of data
Disadvantages:
● Data is replicated
● Applications may work with stale data; crawler only re-crawls sources at certain intervals
The crawling pattern is implemented by Linked Data search engines

"Crawling Pattern" in the Wild
Google Rich Snippets

The On-The-Fly Dereferencing Pattern
● URIs are dereferenced and links are followed the moment the application requires the data
● Applications never process stale data
Disadvantages:
● More complex operations are very slow as they might involve dereferencing thousands of URIs
in the background
● Architectures have been proposed for answering complex queries over the Web of Data by
relying on on-the-fly dereferencing pattern
● Results show that data currency and a very high degree of completeness are achieved at the
price of very slow query execution
The crawling pattern is implemented by Linked Data browsers

"On-the-fly" examples
● Our previous example (dynamic version)
● Tabulator, Marbles

The Query Federation Pattern
● Relies on sending complex queries (or parts) directly to a fixed set of data sources.
● Useful if data sources provide SPARQL endpoints in addition to serving their data on the Web
via dereferenceable URIs
● Enables applications to work with current data without needing to replicate complete data
sources locally
Disadvantages:
● Finding performant query execution plans for join queries over larger numbers of data sources
is complex (i.e. a research topic)
● Query performance slows down significantly when number of data sources grows
● Query federation pattern should only be used in situations where the number of data sources
is known to be small
Applications could follow links between data sources, examine voiD descriptions provided by these
data sources and then include data sources which provide SPARQL endpoints into their list of targets
for federated queries
http://linkeddatabook.com/editions/1.0/

Query Federation Example
SELECT ?birthDate ?spouseName ?movieTitle
?movieDate {
{ SERVICE <http://dbpedia.org/sparql>
{ SELECT ?birthDate ?spouseName WHERE {
?actor rdfs:label "Arnold
Schwarzenegger"@en ;
dbpo:birthDate ?birthDate ;
dbpo:spouse ?spouseURI .
?spouseURI rdfs:label ?spouseName .
FILTER ( lang(?spouseName) = "en" )
} } }
{ SERVICE <http://data.linkedmdb.org/sparql>
{ SELECT ?actor ?movieTitle ?movieDate WHERE {
?actor imdb:actor_name "Arnold
Schwarzenegger".
?movie imdb:actor ?actor ;
dcterms:title ?movieTitle ;
dcterms:date ?movieDate .
} } } }
Application Code
Federated
SPARQL Service
DBPedia.org LinkedMDB.org
e.g. Jena ARQ
See also bobdc.org http://bit.ly/HLdQ4S

Enterprise Use Cases
http://www.w3.org/2001/sw/sweo/public/UseCases/

Enterprise Use Cases
http://www.w3.org/2001/sw/sweo/public/UseCases/
"Enterprise Energy Intelligence" (DERI)
"A Semantic Web Content Repository for Clinical Research"
(Cleveland Clinic)

Cleveland Clinic Use Case
● Improve the Clinic’s ability to use patient data for generating new
knowledge to improve future patient care through outcomes-based and
longitudinal clinical research.
● Leverage expressiveness and versatility of formats to provide
individual patients an appropriate terminology and accessible view of
summary data.
● Over 4 years, Cleveland Clinic has developed a representational
methodology for bridging data collection, document management, and
knowledge representation.
● The result is a unified content repository called SemanticDB.
http://www.w3.org/2001/sw/sweo/public/UseCases/ClevelandClinic/

Cleveland Clinic Use Case
● SemanticDB internally deployed for production on top of an open source
XML & RDF content repository and Firefox (with extensions).
● Methodology realized through a core set of terms that facilitate creation of
a domain vocabulary (or domain model)
○ instances of the vocabulary managed automatically by the system.
● Patient records available as both uniform, structured markup and RDF.
● Coordinated use of both representation languages enables a variety of
operations on patient record:
○ form-based data entry, transformation to reporting formats, document
validation, targeted inference, and querying
○ Operations can be dispatched on the patient record documents and
RDF graphs over a uniform set of interfaces.

DERI "Enterprise Energy Intelligence"
http://dgsit.deri.ie/?q=node/15 and http://slidesha.re/195Gnrr

http://slidesha.re/195Gnrr

More about "Enterprise Linked Data"
Part I: Why Link Enterprise Data?
● Semantic Web and the Linked Data Enterprise
● The Role of Community-Driven Data Curation for Enterprises
Part II: Approval and Support of Linked Data Projects
● Preparing for a Linked Data Enterprise
● Selling and Building Linked Data: Drive Value and Gain Momentum
Part III: Techniques for Linking Enterprise Data
● Enhancing Enterprise 2.0 Ecosystems Using Semantic Web and Linked Data
Technologies
● Linking XBRL Financial Data
● Scalable Reasoning Techniques for Semantic Enterprise Data
● Reliable and Persistent Identification of Linked Data Elements
Part IV: Success Stories
● Linked Data for Fighting Global Hunger
● Enterprise Linked Data as Core Business Infrastructure
● Standardizing Legal Content with OWL and RDF
● A Role for Semantic Web Technologies in Patient Record Data Collection
● Use of Semantic Web technologies on the BBC Web Sites
http://3roundstones.com/led_book/led-contents.html

● Vocabulary design/RDFS
● Knowledge Organization
● Ontology design
● Provenance
● Inference
Advanced Concepts

● Vocabulary design/RDFS
● Knowledge Organization
● Ontology design
● Provenance
● Inference
Advanced Concepts
For a poetic (and humorous!) consideration of the
evolution of the "Semantic Layer Cake" see:
Jim Hendler, "My Take on the Semantic Web Layer
Cake." http://bit.ly/195L70i

Inference: Discovering New Relationships
On the Semantic Web, data is modeled as a set of (named) relationships between resources
● Inference means using automatic procedures to generate new relationships
○ based on the data...
○ ...and some additional information in the form of a vocabulary or a set of rules
● The new relationships may explicitly added to the set of data, or may be returned at query
time (implementation issue)
● The source of additional information is defined through vocabularies or rule sets
● Both approaches draw upon knowledge representation techniques
○ Ontologies provide classification methods, putting an emphasis on defining 'classes',
'subclasses', on how individual resources can be associated to such classes, and
characterizing the relationships among classes and their instances
○ Rules define mechanisms for discovering and generating new relationships based on
existing ones, much like logic programs (Prolog)
● In the Semantic Web toolkit, RDFS, OWL, or SKOS are used for defining ontologies
○ RIF covers rule based approaches

Vocabulary Design: W3C RDFS (1)
● RDF Vocabulary Description Language
● RDF has no mechanism for:
○ describing properties
○ describing the relationships between properties and other resources
● RDF Schema defines classes and properties for describing classes,
properties and other resources
● RDF Schema vocabulary descriptions are written in RDF
http://www.w3.org/TR/2004/REC-rdf-schema-20040210/

RDF Schema: Classes
● rdfs:Resource ...is the class of everything
● rdfs:Class ...declares a resource as a class for other resources
● rdfs:Literal ...literal values such as strings and integers
● rdfs:Datatype ...the class of datatypes
● rdf:XMLLiteral ...the class of XML literal values
● rdf:Property ...the class of properties

RDF Schema: Properties
● rdfs:domain ...declares the class of the subject in a triple whose second component is the predicate.
● rdfs:range ...declares the class or datatype of the object in a triple whose second part is the predicate
○ ex:employer rdfs:domain foaf:Person
○ ex:employer rdfs:range foaf:Organization
● rdf:type ...state that resource is an instance of a class
● rdfs:subClassOf ...allows to declare hierarchies of classes.
○ e.g. "Every Person is an Agent": foaf:Person rdfs:subClassOf foaf:Agent
● rdfs:subPropertyOf ...states that all resources related by one property are also related by another
● rdfs:label ...used to provide a human-readable version of a resource's name
● rdfs:comment ...provides a human-readable description of a resource
● rdfs:seeAlso ...indicates a resource that might provide additional information about the subject resource.
● rdfs:isDefinedBy ...indicates a resource defining the subject resource

Knowledge Organization 1: W3C OWL
● Web Ontology Language
● RDFS too weak to describe resources in sufficient detail
○ No localised range and domain constraints
■ Can’t say that the range of hasChild is person when applied to persons and elephant
when applied to elephants
○ No existence/cardinality constraints
■ Can’t say that all instances of person have a mother that is also a person, or that
persons have exactly 2 parents
○ No transitive, inverse or symmetrical properties
■ Can’t say that isPartOf is a transitive property, that hasPart is the inverse of isPartOf
or that touches is symmetrical
● Difficult to provide reasoning support
○ No “native” reasoners for non-standard semantics
http://www.w3.org/2007/OWL/wiki/OWL_Working_Group or http://bit.ly/195WANj

Knowledge Organization 1: W3C OWL
Desirable features identified for Web Ontology Language:
● Extends existing Web standards
○ Such as XML, RDF, RDFS
● Easy to understand and use
○ Should be based on familiar KR* idioms
● Formally specified
● Of “adequate” expressive power
● Possible to provide automated reasoning support
KR* = knowledge representation
http://www.w3.org/2007/OWL/wiki/OWL_Working_Group or http://bit.ly/195WANj or http://bit.ly/1960964

OWL Tools: Protege-OWL Editor
http://protege.stanford.edu/overview/protege-owl.html

Knowledge Organization 2: W3C SKOS
● Simple Knowledge Organization System
● An application of RDFS and OWL
● Provides a way to represent controlled vocabularies, taxonomies and
thesauri
○ controlled vocabulary: a list of terms which a community or
organization has agreed upon
○ taxonomy: a controlled vocabulary organized in a hierarchy
○ thesaurus: a taxonomy with more information about each concept
including preferred and alternative terms.
○ A thesaurus may also contain relationships to related concepts
● SKOS is an OWL ontology; it can be written out in any RDF syntax
http://www.w3.org/2004/02/skos/ or http://slidesha.re/1etWDue or http://bit.ly/1etYLlE

Provenance: The W3C PROV Model
● A set of W3C recommendations and notes on modelling provenance
● PROV-O is the "core..."
http://www.w3.org/TR/prov-primer/

Provenance in a Nutshell
● prov:Entity is a physical, digital, conceptual, or other kind
of thing with some fixed aspects; entities may be real or
imaginary
● prov:Activity is something that occurs over a period of
time and acts upon or with entities; it may include
consuming, processing, transforming, modifying,
relocating, using, or generating entities
● prov:Agent is something that bears some form of
responsibility for an activity taking place, for the
existence of an entity, or for another agent's
activity
These three classes provide a basis for the rest of PROV-O
http://www.w3.org/TR/prov-primer/ or http://www.provbook.org/

Inference and W3C RIF
● Production Rules
○ Analogous to instruction in a program: If a certain condition holds, then some action is
carried out
○ Example: "If a customer has flown more than 100,000 miles, then upgrade him to Gold
Member status."
● Declarative Rules
○ Stating a fact about the world
○ Understood as sentences of the form "If P, then Q"
○ Example: "If a person is currently president of the United
States of America, then his or her current residence is the
White House."
● There are many rule systems, esp. in the expert systems domain
● The W3C Rule Interchange Format is an interchange format
between existing rule systems
http://www.w3.org/TR/2013/NOTE-rif-primer-20130205/

What is Web Science?
● Positions the World Wide Web as an object of scientific
study unto itself
● Recognizes the Web as a transformational, disruptive
technology
● Its practitioners focus on understanding the Web...
○ ...its components, facets and characteristics
● The Web Science Method: “the process of designing
things in a very large space..."

What does Web Science ask?
● What processes have driven the Web’s growth?
○ Will they persist?
● How does large-scale structure emerge from a simple set of
protocols?
● How does the Web function as a socio-technical system?
● What drives the viral uptake of certain Web phenomena?
Bottom line: What might fragment the Web?

Engineering a Semantic Web (Spring 2018)

More Related Content

What's hot (20)

Similar to Engineering a Semantic Web (Spring 2018) (20)

More from Rensselaer Polytechnic Institute (6)

Recently uploaded (20)

Engineering a Semantic Web (Spring 2018)