Paper Info
Title
A Sensor Selection And Learning Method For Credibility Assessment Using Sensor Data
Abstract
We herein introduce a sensor selection and learning method for credibility assessment using sensor data. To assess the credibility of information, comparing with primary resources and integrating sensor data are effective to derive the objectivity of the information. However, when we are using sensor data, if incorrect sensors are selected or if the necessary sensors are not referred, it implies that the credibility of the credibility assessment system is uncertain. This paper presents a sensor selection and learning method for information credibility assessment using sensor data. In this method, several discriminators select sensors through weighting. Additionally, the result is feedbacked to the discriminators to learn the precision of the discriminators. The primary contribution of this method is the proposal of a method of sensor selection from several sensors that is opened on the World Wide Web by weighting from discriminators such that the discriminators can learn their precision from the feedback of the selection result. This paper shows the feasibility and usefulness of the method through experiments. An application method for credibility assessment technology is also introduced.
Year
DOI
Venue
2018
10.3233/978-1-61499-933-1-400
INFORMATION MODELLING AND KNOWLEDGE BASES XXX
Keywords
Field
DocType
Credibility, SensorData, OpenData, Sensor Selection, Data Integration
Discrete mathematics,Credibility,Artificial intelligence,Sensor selection,Mathematics,Machine learning
Conference
Volume
ISSN
Citations 
312
0922-6389
0
PageRank 
References 
Authors
0.34
0
2
Name
Order
Citations
PageRank
Ken Honda100.34
Yoshida, N.211.55