Piazza: Data Management Infrastructure for Semantic Web Applications

1
Piazza Data Management Infrastructure for
Semantic WebApplications

• Alon Y. Halevy, Zachary G. Ives, Peter Mork, Igor
  Tatarinov.

Speaker Sergey Chernov Tutor Jens Graupmann
2
Outline

• INTRODUCTION. SEMANTIC WEB.
• PIAZZA SYSTEM OVERVIEW
• IMPLEMENTATION DETAILS
• 3.1 MAPPING LANGUAGE
• 3.2 QUERY ANSWERING ALGORITHM
• CONCLUSIONS.

3
Introduction

• Goal
• Data Integration and Knowledge Management
• Problem
• Web data lacks machine-understandable semantics
• Solution
• Semantic Web?

4
The Semantic Web

• Web sites include structural annotations
• You can pose meaningful queries on them.
• Ontologies provide the semantic glue.
• Internal implementation of web sites left open.
• Agents perform tasks
• Query one or more web sites
• Perform updates (e.g., set schedules)
• Coordinate actions
• Trust each other (or not).
• I.e., agents operating on a gigantic
  heterogeneous distributed database.

(View by A. Halevy)
5
General requirements

• Robust infrastructure for querying
• Peer data management systems.
• Facilitate mapping between different structures.
  Need tools for
• Locating relevant structures
• Easily joining the semantic web.
• Get data into structured form
• Should we worry about the legacy web?

6
Using views for specifyingmappings

• Local-As-View (LAV).
• Data sources can be described as views over the
  mediated schema.
• Global-As-View (GAV).
• Mediated schema can be described as a set of
  views over the data sources.

Mediated Schema
Site B
Site A
Site C
Mediated Schema
Site B
Site A
Site C
7
Mapping

• Mapping AB specifies representation of structured
  data from scheme of node A into scheme of node B

Mediated Schema
Mapping MS-C
Mapping A-MS
Mapping C-MS
Mapping MS-A
Mapping AB
Mapping BC
Site B
Site C
Site A
Mapping BA
Mapping CB
8
Piazza Peer Data-Management System

• Goal
• Large scale autonomous sharing of structured data
• Peer data management system (PDMS)
• Autonomous Peers export data in their own schemas
• Pair-wise mappings between peers
• Generalization of a Data Integration system
• NOT a P2P file sharing system

9
Relationship of PDMS to

• P2P overlay networks (the Structured World)
• Data integration systems (no central logical
  mediated schema)
• Federated databases (scale, ad-hoc nature)
• Distributed databases (no central administration)

10
Representing Data

• A spectrum of possibilities
• Relational tables, some integrity constraints
• XML can encode relational, hierarchical
• Xquery emerging standard query language (SQL
  for XML)
• RDF XML on drugs.
• Sees only the logic ignores other aspects.
• DAMLOIL
• Full-blown Knowledge representation language.
• They all have semantics just different
  expressive powers.
• We keep the data simple. Mappings between data at
  different peers are more complex.

11
Peer Data Management
DB Projects
MIT
UW
UCB
Stanford

• Mappings are query expressions
• DbResearcher(x) ? Researcher(x),Area(x,DB)
• DbResearcher(x), Office(x,DBLab) DbLabMember(x)

12
Piazza mapping language (1)

• XML/XML Example

ltpubsgt ltbookgt a IN
document(source.xml)\ /authors/author
t IN a/publication/title, typ IN
a/publication/pub-type WHERE typ book
lttitlegt t lt/titlegt
ltauthorgt ltnamegt a/full-name lt/namegt
lt/authorgt lt/bookgt lt/pubsgt
Target pubs book title
author name publisher
name
Source authors author full-name
publication title
pub-type
13
Piazza mapping language (2)

• piazzaid attribute

ltpubsgt ltbook piazzaidtgt
a IN document(source.xml)\
/authors/author t IN a/publication/title,
typ IN a/publication/pub-type WHERE typ
book lttitle piazzaidtgt t
lt/titlegt ltauthor piazzaidtgt
ltnamegt a/full-name lt/namegt
lt/authorgt lt/bookgt lt/pubsgt
Target pubs book title
author name publisher
name
Source authors author full-name
publication title
pub-type
14
Piazza mapping language (3)

• Partial mapping

ltpubsgt ltbook piazzaidtgt
a IN document(source.xml)\
/authors/author t IN a/publication/title,
typ IN a/publication/pub-type WHERE typ
book PROPERTY t gtA AND t lt
B ltpublishergt
ltnamegt PROPERTY this IN
PrintersInc, PubsInc lt/namegt
lt/publishergt lt/bookgt lt/pubsgt
Target pubs book title
author name publisher
name
Source authors author full-name
publication title
pub-type
15
Query Answering Algorithm

• Problem
• Evaluate query Q at P1 given a network of
  mappings
• Reformulate the query over all relevant peers
• Chaining of mappings using a combination of query
  composition and query rewriting
• QP1(x) - DbResearcher(x)
• Query Composition
• M DbResearcher(x) ? Researcher(x),Area(x,DB)
• ? QP2 (x) ?
  Researcher(x),Area(x,DB)
• Query Rewriting
• M DbResearcher(x), Office(x,DBLab)
  DbLabMember(x)
• ? QP3 (x) ?
  DbLabMember(x)

16
Query Reformulation (1)
Mapping
Query
ltS2gt ltpeoplegt people/S1/people
ltfacultygt namepeople/faculty/name/text()
name lt/facultygt
ltstudentgt studentpeople/student/text()
ltnamegt student lt/namegt
ltadvisorgt facultypeople/faculty,
namefaculty/name/text(),
adviseefaculty/advisee/text()
where adviseestudent
name ltadvisorgt lt/studentgt
lt/peoplegt lt/S2gt
ltresultgt for faculty in
/S1/people/faculty, name in
faculty/name/text(), advisee in
faculty/advisee/text() where name
Ullman return ltstudentgt
advisee lt/studentgt lt/resultgt
17
Query Reformulation (2)
Query tree pattern
Mapping tree pattern
Query
ltS2gt
ltresultgt for faculty in
/S1/people/faculty, name in
faculty/name/text(), advisee in
faculty/advisee/text() where name
Ullman return ltstudentgt
advisee lt/studentgt lt/resultgt
S1 ltpeoplegt people
faculty name advisee adviseestudent
ltadvisorgt name
18
Query Reformulation (3)
Query tree pattern
Mapping tree pattern
Query
ltS2gt
ltresultgt for faculty in
/S2/people/student, advisor in
student/advisor/text(), name in
student/name/text() where advisor
Ullman return ltstudentgt name
lt/studentgt lt/resultgt
S1 ltpeoplegt people
faculty name advisee adviseestudent
ltadvisorgt name
19
Reformulation times

• Table 1 The test queries and their respective
  running times.

20
Current and the Future

• Current status
• Demo scenario using XML
• Looking at real domains (Bio dbs, NASA dbs)
• Future Work
• More efficient reformulation algorithm
• Semantic network analysis eliminate redundant
  mappings and inconsistent mappings
• Query caching to speed up query evaluation

21
Conclusions

• Mapping language for mapping between sets of XML
  source nodes with different document structures
• Architecture that uses the transitive closure of
  mappings to answer queries
• Algorithm for query answering over this
  transitive closure of mappings, which is able to
  follow mappings in both forward and reverse
  directions

22
Thank You!
23
Further literature

• Alon Y. Halevy, Zachary G. Ives, Dan Suciu, Igor
  Tatarinov Schema Mediation for Large-Scale
  Semantic Data Sharing
• Igor Tatarinov, Zachary Ives, Jayant Madhavan,
  Alon Halevy, Dan Suciu, Nilesh Dalvi, Xin (Luna)
  Dong, Yana Kadiyska, Gerome Miklau, Peter Mork
  The Piazza Peer Data Management Project
• Alon Y. Halevy, Zachary G. Ives, Dan Suciu, Igor
  Tatarinov Schema Mediation in Peer Data
  Management Systems
• Alon Halevy, Oren Etzioni, AnHai Doan, Zachary
  Ives, Jayant Madhavan, Luke McDowell, Igor
  Tatarinov Crossing the Structure Chasm
• Madhan Arumugam, Amit Sheth, and I. Budak
  Arpinar Towards Peer-to-Peer Semantic Web A
  Distributed Environment for Sharing Semantic
  Knowledge on the Web
• Hendler J., Berners-Lee T., Miller E.
  Integrating Applications on the Semantic Web