Semantic Technologies and Programmatic Access to Semantic Data

Steffen Staab Programmatic Access to Semantic Data 1Institute for Web Science and Technologies · University of Koblenz-Landau, GermanyWeb and Internet Science Group · ECS · University of Southampton, UK &

Semantic Technologies and Programmatic Access to Semantic Data

Steffen Staab

University of Southampton &

Universität Koblenz-Landau

Steffen Staab Programmatic Access to Semantic Data 2

Traditional Information System

Business Logics

Structured DataUnstructured

Data

Presentation and Interaction

Charakteristics:• Processes known• Data structures

known• Meaning of data in

schema and implicit in code


Information ecosystems nowadaysExamples• Open Data• 1000s

DBs/company• Ad-hoc data

Characteristics• Some structure• Late structure• Social context• Meaning of data

most important


How does data receive meaning?Explicit:• Formal schema/ontology

– By someone else?

Implicit:• Names are just used for

describing

Social:• Communities converge

– By discussion– By emergence

Meaning?


Talking many languages...

Sub languages• For consumer

– title,...• Global retailers

– barcode• US food industry

– serving size, calories,...• Producer

– batch number...

https://www.youtube.com/watch?v=ga1aSJXCFe0

Depending on who you are – you encounter the (un)expected, the (un)known, the (un)understandable,...


The Unknown

https://www.flickr.com/photos/wrobel/8175902444/


We know a bit.... 1. URIs as identifiers

2. http lookup

3. RDF (triples)

4. relations, also to other locations


What is/should Linked Data good for?• Data integration is (relatively) easy

– Migrating different data sources to linked data is (relatively) easy

• Late schema is easy– Just add some more fields

• Ignoring data is easy– Think of crisps

• Serendipitous use– Discover new information &

new sources by following links

• Data repurposing / pointing– Use what others have done at both schema

and data level

Dealing w

ith the unknown

data and data schema


Issue: From Data Publishing to Understanding

?

De-contextualization Re-contextualization

Publishing data the structure of which you know is easier than understanding what you do not know


Example scenario: JamendoData about license free music• ~ 1 Million triples• classes and predicates

from 18 different ontologies– FOAF, Tag ontology,

music ontology, …

Simple programming task:• List for every music artist,

all the records they made


Software Development Process Overview

data model design

revised data model design

data model prototype

data queries

final data model

Creation of initial data

model

Exploration of the data

source

Creation of model in

code

Query design / implementation

Mapping of query results


Accessing Artists Using Apache Jena


From artists to songs

Observations• SPARQL queries are strings• Results are strings• Requires good understanding of the data source

No semantics, just syntax


Programming Language Support for RDF Access

Static Typing Errors detected before

execution Misspelling discovered by

compiler! Anectode: 2nd place

because of misspelt code Static types are form of

documentation Less knowledge about

data source required Better IDE integration /

autocompletion

Code generation• Sommer• Winter• OntoMDEDynamic Typing E.g. ActiveRDF

(Oren et al 2007)) “convention over

configuration” dynamic

metaprogramming allows for slick code


Programming with Linked Data


c1

Programming with Linked Data

Tasks of the Programmer 1 Schema exploration

2 Programming code types

3 Programming queries

4 Programming procedures for

• creating, • manipulating,• persisting

objects


Node Path Query Language Using AutocompletionExploration of classes


Node Path Query Language Using AutocompletionExploration of classes

Exploration of relations


Node Path Query Language: Query FormulationExploration of classesExploration of relationsQuerying for instances

Type set of mo:MusicArtist

No definition or declaration needed


Node Path Query Language for Code DevelopmentExploration of classesExploration of relationsQuerying for instancesDeveloping code with queries

All translated into SPARQL queries at• Development time• Type inference at compile time

(but also as part of IDE)• Querying again at run time

One language to bind them all


Node Path Query Language for Code DevelopmentExploration of classesExploration of relationsQuerying for instancesDeveloping code with queriesDeveloping code with new classes

All translated into SPARQL queries at• Development time• Run time update• Persistence!


NPQLNPQL (Node Path Query Language)• Intensional Queries Describing RDF classes and properties for reuse in IDE and in host language metaprogramming

• Extensional Queries Class instances and property instances

• Compilation to SPARQL for reuse of existing endpoints

Ongoing discussion about details of NPQL


LITEQNPQL (Node Path Query Language)• Intensional Queries• Extensional Queries• Compilation to SPARQL

LITEQ (Language Integrated Types, Extensions and Queries) • Implementation of NPQL as F# Type Provider in Visual Studio• Autocompletion using NPQL queries• Automatic typing

of extensional query resultsby intensional queries


Outlook: Programming with Linked Data• More expressive query languages

– Derived data types in tractable description logics!

• More precise combined type inference– (derived) type from data source– type inference in programming language

• Programming across data sources– Federated queries– Linktraversal-based queries (the unknown sources)

• Integration of schema induction – Low quality of schema/ontologies

• Improved autocompletion


Conclusion


Issue: From Data Publishing to Unknown Data Understanding

CognitionStorytellingPragmatics

Ontology PatternsConceptual Modeling

Metamodels...

QuantityPertinenceManner


What is missing?...a lot...

• Indexing• Search• Data and schema quality• Pragmatics• ...


Semantic Web

Social Web & Web Retrieval

Interactive Web & Human Computing

Web & Economy

Software & Services

Computational Social Science

Thank You!