Earticle

초록

영어

Context information in smart space are complex and uncertain, it increases the difficulty of information prediction. In order to increase the accuracy of information prediction, a new method is proposed in this paper. Firstly, first-order predicate logic is used to describe the context information and Markov logic network is used to achieve the unification of the first-order logic and probabilistic graphical models. Secondly Markov logic network model is established based on some particular scenes which were predicted by neural network, the purpose of this model is to realize the prediction of device information which is valuable for the task planning by relevant complex context information. At last, simulation test is done using the real data collected from our smart space laboratory, the experiment results show that method is effective.

목차

Abstract
 1. Introduction
 2. First-order Logic Description of Context Information
 3. Markov Logic Network
  3.1. Markov Network
  3.2. Definition of Markov Logic Network
 4. The Establishment and Use of Knowledge Base in Smart Space
  4.1. Markov Logic Network Based on the Specific Situation
  4.2. Parameter Learning and Prediction of Markov Logic Networks
 5. The Application of Markov Logic Network in Specific Situation
  5.1. Determination of Certain Circumstances in Smart Space
  5.2. Specific Applications of Markov Logic Network in Smart Space
 6. The Experimental Results
  6.1. Construction of the Original Markov Logic Network and Parameter Training
  6.2. Platform of Prediction Algorithm and Reasoning
  6.3. Comparison of Markov Logic Network and Bayesian Network
 7. Summary
 Acknowledgement
 References

키워드

저자

• Lu Fei [ School of Control Science and Engineering, Shandong University, Jinan, 250061, China ]
• Li Jian [ School of Control Science and Engineering, Shandong University, Jinan, 250061, China ]

참고문헌

발행기관

간행물

표지보기 관심저널 등록

• 간행물명

  International Journal of Hybrid Information Technology

• 간기

  격월간

• pISSN

  1738-9968

• 수록기간

  2008~2016

• 십진분류

  KDC 505 DDC 605