Markov logic: a unifying language for knowledge and information management

ABSTRACT

Modern information and knowledge management is characterized by high degrees of complexity and uncertainty. Complexity is well handled by first-order logic, and uncertainty by probabilistic graphical models. What has been sorely missing is a seamless combination of the two. Markov logic provides this by attaching weights to logical formulas and treating them as templates for features of Markov random fields. This talks surveys Markov logic representation, inference, learning and applications. Inference algorithms combine ideas from satisfiability testing, resolution, Markov chain Monte Carlo and belief propagation. Learning algorithms involve statistical weight learning and inductive logic programming. Markov logic has been successfully applied to a wide range of information and knowledge management problems, including information extraction, entity resolution, ontology learning, link prediction, heterogeneous knowledge bases, and others. It is the basis of the open-source Alchemy system (http://alchemy.cs.washington.edu).

INDEX TERMS

Primary Classification:
  F. Theory of Computation
  F.4 MATHEMATICAL LOGIC AND FORMAL LANGUAGES
      F.4.1 Mathematical Logic

Additional Classification:
  G. Mathematics of Computing
  G.3 PROBABILITY AND STATISTICS

  H. Information Systems
  H.2 DATABASE MANAGEMENT
  H.3 INFORMATION STORAGE AND RETRIEVAL
      H.3.3 Information Search and Retrieval

  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE

General Terms:
Algorithms, Design, Experimentation, Languages, Theory

Keywords:
first-order logic, markov logic, markov networks