Multifaceted Exploitation of Metadata for Attribute Match Discovery in Information Integration

1
Multifaceted Exploitation of Metadata for
Attribute Match Discovery in Information
Integration

• David W. Embley
• David Jackman
• Li Xu

2
Background

• Problem Attribute Matching
• Matching Possibilities (Facets)
• Attribute Names
• Data-Value Characteristics
• Expected Data Values
• Data-Dictionary Information
• Structural Properties

3
Approach

• Target Schema T
• Source Schema S
• Framework
• Individual Facet Matching
• Combining Facets
• Best-First Match Iteration

4
Example
Car
Car
Style
has
01
0
01
01
has
has
has
Cost
Mileage
Miles
Source Schema S
Target Schema T
5
Individual Facet Matching

• Attribute Names
• Data-Value Characteristics
• Expected Data Values

6
Attribute Names

• Target and Source Attributes
• T A
• S B
• WordNet
• C4.5 Decision Tree feature selection
• f0 same word
• f1 synonym
• f2 sum of distances to a common hypernym root
• f3 number of different common hypernym roots
• f4 sum of the number of senses of A and B

7
WordNet Rule
8
Confidence Measures
9
Data-Value Characteristics

• C4.5 Decision Tree
• Features
• Numeric data
• (Mean, variation, standard deviation, )
• Alphanumeric data
• (String length, numeric ratio, space ratio)

10
Confidence Measures
11
Expected Data Values

• Target Schema T and Source Schema S
• Regular expression recognizer for attribute A in
  T
• Data instances for attribute B in S
• Hit Ratio N/N for (A, B) match
• N number of B data instances recognized by the
  regular expressions of A
• N number of B data instances

12
Confidence Measures
13
Combined Measures
Threshold 0.5
14
Final Confidence Measures
15
Experimental Results

• Matched Attributes
• 100 (32 of 32)
• Unmatched Attributes
• 99.5 (374 of 376)
• Feature ---Color
• Feature ---Body Type.

16
Conclusions

• Direct Attribute Matching feasible
• Individual-Facet Matching good
• Multifaceted Matching better

17
Future Work

• Additional Facets
• More Sophisticated Combinations
• Additional Application Domains
• Automating Feature Selection
• Indirect Attribute Matching

www.deg.byu.edu