Evaluating MLNs for Collective Classification

Abstract

Collective Classification is the process of labeling instances in a graph using both instance attribute information and information about relations between instances. While several Collective Classification Algorithms have been well studied, the use of Markov Logic Networks (MLNs) remains largely untested. MLNs pair first order logic statements with a numerical weight. With properly assigned weights, these rules may be used to infer class labels from evidence stated as logic statements. Our study evaluated MLNs against other Collective Classification algorithms on both synthetic data and real date from the CiteSeer dataset. Also whole, we encountered inconsistent and often poor performance with MLNs, especially on synthetic data where other Collective Classification algorithms performed well.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 13, 2010
Accession Number
ADA535643

Entities

People

  • Robert J. Crane

Tags

DTIC Thesaurus Topics

  • Abstracts
  • Algorithms
  • Availability
  • Classification
  • Computer Science
  • Computers
  • Contracts
  • Information Operations
  • Instructions
  • Logic
  • Logic Gates
  • Monitoring
  • Networks
  • Security
  • United States Naval Academy

Fields of Study

  • Computer science

Readers

  • Atmospheric Science/Meteorology
  • Distributed Systems and Data Platform Development
  • Systems Analysis and Design

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks