XMLBased Information Mediation with MIX

1
XML-Based Information Mediation with MIX
SDSC/DICE C. Baru, A.Gupta, B. Ludäscher, R.
Marciano UCSD DB-Lab Y. Papakonstantinou, P.
Velikhov, V. Chu

• MIX -- Mediation of Information using XML
• Mediator Architecture
• Homebuyer Example
• MIX components

2
The MIX VisionThe Web emerges as a distributed
database with XML as its data model
XML query language
XML documents
XML View(s)
XML View(s)
XML View(s)
Wrappers also export 1. Schemas Metadata
(XML-Data, RDF,...) 2. Description of
supported queries...
Wrapper
Wrapper
Native XML Database
Legacy Source
Legacy Source
3
Mediated XML Views
MIXm Mediator
XML View(s)
XML View(s)
XML View(s)
Wrapper
Wrapper
XML Data Source
Data Source
Computational Source
4
The Home Buyer Scenario

• Find homes
• in the Del Mar region,
• with at least 2 beds, 2 baths, and a pool, and
• an elementary school in the neighborhood

Web Client
XMAS Query
Results (XML)

• XMAS query

MIXm Mediator

• Query GUI

Homes mediator
Data
Data
Neighborhood mediator
National test scores
Data
Schools mediator
Home info (real estate)
Community info (name, ZIP)
Crime info (ZIP, stats)
Nhood info (demographics)
Schools info (address, size)
School district info (scores,spending,ZIP)
www.realtor.com
www.sandag.cog.ca.us
www.sannet.gov
www.homeadvisor.msn.com
www.asd.com
5
Home Buyer GUI
6
Home Buyer GUI (Answers)
Generated XMAS Query
XML Answer Document
7
MIX Components - BBQ

• Blended Browsing and Querying (BBQ) interface
• DTD based query composition
• generates XMAS queries against mediator or source
  views
• Browsing (navigation) and querying of XML
  documents
• generates XMAS queries modified by DOM-VXD
  operations to incrementally evaluate the result
  set, to support navigation of XML documents

8
BBQ Example (AMICO Prototype)
Q1 Find title, type, and image ID of
paintings
BBQ Interface
Q2 Find creator and related metadata of
paintings
XMAS query
XML doc
MIXm View based on AMICO DTD
Wrapper
AMICO XML Database
AMICO XML Database
MARC Database
9
BBQ AMICO Example
BBQ query composition
XSL rendered output
XML answer document
10
MIX Components - Wrappers

• Wrapper toolkit
• model resource information using XML DTD / XML
  Schema, including a mapping source data ? DTD
• provide mapping XMAS ? source QL / operations
• Wrappers for
• Relational schemas (XViews) XMAS ? SQL
• GIS (ArcView, MapObjects)
• HTML Wrappers
• Interactive Web sites (Forms, Image Maps, )

11
MIX Components - Mediator, XMAS

• Mediator toolkit
• allows definition of views across multiple
  sources
• query engine to execute queries on views
• views are expressed in a declarative XML query
  language XMAS ...
• XML Matching And Structuring QL
• set of XML sources new integrated XML
  document
• evaluation XMAS algebra (plus DOM-VXD)
• Extensions
• Mediation of spatial sources
• Support for external functions

12
MIX Components - DOM-VXD, LXP

• DOM extension for Virtual XML Documents (DOM-VXD)
• lazy, demand-driven implementation of DOM.
  Supports browsing/navigation of XML documents
  with a server-side, compute as you go model
• Lean XML Protocol (LXP)
• Minimal protocol for exchanging partially
  explored XML objects
• Related to W3C XML Fragment Interchange
• Stanford SDLIP Simple Digital Library
  Interoperability Protocol

open XML trees
13
MIX Components - DTD Inference/Mining

• DTD Inference
• given the source DTDs and the (XMAS) view
  definition, determine the tightest DTD of the
  result document
• DTD Mining
• given a set of example XML documents, derive a
  best possible DTD which captures all examples
• use DTDs for query optimization and BBQ

14
The MIX Project Team

• UCSD DB-Lab (CSE)
• Yannis Papakonstantinou
• (Project Lead)
• Victor Vianu
• Pavel Velikhov
• Andreas Giannakopoulos
• Pratik Mukhopadhay
• Michail Petropoulos

• SDSC/DICE
• Chaitan Baru (Project Lead)
• Amarnath Gupta
• Bertram Ludäscher
• Richard Marciano
• Kevin Munroe
• Paul Nguyen
• Ilya Zaslavsky

15
URLs

• MIX
• http//www.npaci.edu/DICE/MIX
• http//www.db.ucsd.edu/Projects/MIX
• AMICO http//www.npaci.edu/DICE/AMICO
• SRB http//www.npaci.edu/DICE/SRB
• HPSS http//www.sdsc.edu/hpss
• DICE http//www.npaci.edu/DICE
• UCSD DB-LAB http//www.db.ucsd.edu