cooperative query answering for semistructured data michael barg raymond k. wong reviewed by...

14
Cooperative Query Cooperative Query Answering for Answering for Semistructured Semistructured data data Michael Barg Raymond K. Wong Michael Barg Raymond K. Wong Reviewed by Reviewed by Swetha Swetha Jack Jack Christian (Absent) Christian (Absent) Chris Chris

Upload: thomasine-carr

Post on 26-Dec-2015

212 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

Cooperative Query Cooperative Query Answering for Answering for

Semistructured dataSemistructured dataMichael Barg Raymond K. WongMichael Barg Raymond K. Wong

Reviewed by Reviewed by SwethaSwetha

JackJackChristian (Absent)Christian (Absent)

ChrisChris

Page 2: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

ContentsContents

IntroductionIntroduction

What is Cooperative Query Answering?What is Cooperative Query Answering?

MotivationMotivation

Work DoneWork Done

Results with a detailed exampleResults with a detailed example

Summary and related workSummary and related work

Page 3: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

IntroductionIntroduction

Semi structured data - XML Semi structured data - XML Self describing schemaSelf describing schema Heterogeneous data sources (combination of Heterogeneous data sources (combination of

schemas)schemas) Standard XPATH queryStandard XPATH query

REQUIRES ACCURATE UNDERSTANDING REQUIRES ACCURATE UNDERSTANDING OF STRUCTUREOF STRUCTURE

Page 4: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

What is Cooperative Query What is Cooperative Query Answering?Answering?

Cooperative query answering (CQA) seeks to Cooperative query answering (CQA) seeks to provideprovide

relevant responses to queries posed by users in relevant responses to queries posed by users in cases where a direct answer is not very helpful or cases where a direct answer is not very helpful or informative. informative.

Approximate or alternative information may be Approximate or alternative information may be more useful and less misleading to a user.more useful and less misleading to a user.

Example: Example: Q: “Is Sam an associate professor?”Q: “Is Sam an associate professor?”

User believes most associate professors have tenure.User believes most associate professors have tenure.

Sam is not tenured. Sam is an associate professor.Sam is not tenured. Sam is an associate professor.

A: “Yes, but he doesn’t have tenure.”A: “Yes, but he doesn’t have tenure.”

Page 5: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

MotivationMotivation

To query XMLTo query XML Knowledge of underlying structureKnowledge of underlying structure Rich semantics come from structureRich semantics come from structure Documents adhering to different schema Documents adhering to different schema

located in same repositorylocated in same repository

Page 6: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

Work DoneWork Done

Edges -> relationshipsEdges -> relationships

Nodes -> elementsNodes -> elements

Algorithm is based onAlgorithm is based on

traversing the graphtraversing the graph

and scoring similarity.and scoring similarity.

Page 7: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

Work Done (2)Work Done (2)

Algorithm: Cooperative Query ProcessingAlgorithm: Cooperative Query Processing

Input: Input: Query Q, which is represented by the query tree, QQuery Q, which is represented by the query tree, QTT

Output: Output: Set of ranked tuples, <Set of ranked tuples, <resultresult, score >, score >

Page 8: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

ResultResultDesired Info

Find all phone numbers of restaurants in Soho

Actual Query //restaurant[//Soho]//phone number

Results Phone numbers with an ancestor ”Restaurant” who had a child named ”Soho”,

Phone numbers of restaurants with an address in Soho

Page 9: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

Result (Cont’s)Result (Cont’s)

Page 10: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

Result (Cont’s)Result (Cont’s)

Page 11: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

Related WorkRelated Work

Not well researched for semistructured Not well researched for semistructured datadata

Well established in Relation Databases.Well established in Relation Databases. Extending SQL SyntaxExtending SQL Syntax Database (CoBase)Database (CoBase) Concept SimilaritiesConcept Similarities Cooperative InterfacingCooperative Interfacing

Page 12: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

SummarySummary

Casual and non-expert users might not have Casual and non-expert users might not have accurate knowledge of the underlying structure.accurate knowledge of the underlying structure.

Cooperative Query Answering is a method which Cooperative Query Answering is a method which enables users to meaningfully query semi enables users to meaningfully query semi structured (i.e. XML) data.structured (i.e. XML) data.

The paper suggested a graph based method to The paper suggested a graph based method to do thisdo this Traversing through the nodesTraversing through the nodes Results is promisingResults is promising

Page 13: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

Questions?Questions?

Page 14: Cooperative Query Answering for Semistructured data Michael Barg Raymond K. Wong Reviewed by SwethaJack Christian (Absent) Chris

ReferencesReferencesM. Barg, R.K. Wong. Cooperative Query Answering for Semistructured Data.In Proceedings of the 14th Australasian Database Conference (ADC), Adelaide Australia,Feb 2003.

S. Abiteboul. Querying semi-structured data. In Proceedings of the InternationalConference on Database Theory (ICDT). Springer Verlag, 1997.

T. Bray, J. Paoli, and C.M. Sperberg-McQueen. Extensible markup language(XML)1.0. In W3C Recommendation, World Wide Web Consortium, 1998; availableonline at http://www.w3.org/TR/1998/REC-xml-19980210.

P. Buneman. Tutorial: Semistructured data. In International Conference onPODS, 1997.

T. Gaasterland, P. Godfrey and J. Minker. An Overview of CooperativeAnswering, Journal of Intelligent Information Systems, 1, 123-157, 1992

D. Quass, A. Rajaraman, Y. Sagiv, J. Ullman,and J. Widom. Queryingsemistructured heterogeneous information. In Proceedings of Deductive andObject Oriented Databases, 1995.