Data Warehousing Project Proposal Topic: BibHouse

1
Data Warehousing Project ProposalTopic BibHouse

• Amanda Cohen, Marcie Matejka, Bryan Wilhelm,
  Tamara Wilson

2
Background and Goals

• Each piece of scholarly work (book, article,
  etc.)  has a citation that details the aspects of
  the work
• Author, Title, Date, Publication details, etc.
• Citation acts as metadata for article
• Can be linked to other citations and data
• Goal To demonstrate how to load heterogeneous
  citation data sources into a data warehouse and
  how to perform analysis that data

3
Data Sources

• DBLP
• Format XML
• Maintained at University of Trier
• CiteSeer
• Format BibTex and OAI
• CIA
• Will be used to track geopolitical data about
  citations if time allows

4
Data Integration and Analysis

• Transformation and Cleaning
• Ensure each citation conforms to Chicago MLA
  citation style
• Missing pieces of information will be found by
  cross-referencing citation with other data
  sources
• Analysis
• Perform clustering and classification algorithms
• User will be able to see connections between
  publication based on several aspects
• Other publications by a specific author,
  citations on related topics, etc.

5
Tools

• Data Integration
• ETL Process
• Examines data to see if it can be used for other
  purposes
• Microsoft SQL Server Integration Services (SSIS)
• Potters Wheel
• BibTex Tools
• Data Analysis
• SQL Server Analysis Services and Reporting
  Services (to summarize results)
• Weka

6
Deliverables

• Final Report
• Printout of Results to sample analysis queries
• Warehouse Schema
• Public Demonstrations of Warehouse functionality

7
Data Warehouse Schema
8
Database Tables