Real Estate II

Abstract

This data set lists houses for sale. The data set is used for the purpose of schema matching, especially for the complex schema matching.

Contribution

Original Owner and Donor

Anhai Doan
Department of Computer Science 
University of Illinois, Champaign-Urbana
anhai@cs.uiuc.edu

Description

• The data is created from the Real Estate I by removing, merging, and splitting certain schema elements. The objective was to create real-estate schemas that have a fair number of complex matchings.
• This data was collected as a designed experiment for the purpose of Schema Matching.
• As of now the publications that have used this data are:
  • Learning to Map between Structured Representations of Data, A. Doan. Ph.D. Dissertation, Univ. of Washington-Seattle, 2002.
  • iMAP: Discovering Complex Semantic Matches between Database Schemas, A. Doan, Y. Lee, R. Dhamankar, A. Halevy, and P. Domingos. Proc. of the ACM SIGMOD Conf. on Management of Data. To appear.

Data Format

This data set consists of one original data set and one sample mapping over generated data from the original one.

• Original Data Set
  • The data was selected from the Real Estate I.
  • The user can browse each XML file and download one tarred and gzipped file.
• Sample Mapping
  • We asked a volunteer to examine and create complex query formulas that combine the attributes.
  • We created sample target sources by applying the query formulas over a subset of the original one.
  • The target source has been divided into two files to represent some join paths.(Here, the join path is "agent-id".)

Data Files

• Data Sources
  • Schema
  • Browse
  • One XML file per record[Tarred & gzipped (323 KB)]
• Mapping
  • Source
    • Schema
    • One big XML file[Gzipped(21 KB)]
  • Target(join path:"agent_id")
    • Agent Details
      • Schema
      • One big XML file[Gzipped(4 KB)]
    • House Details
      • Schema
      • One big XML file[Gzipped(21 KB)]