Volunteer Computing systems (VC) use the idle computing resources which are composed of unreliable volunteers whose exhibit volatile behavior and often they are not available because of the autonomy nature of owners. Scientific experiments that use internet-connected computers are internet based VC projects. However, the volatility of volunteer is significant challenge in VC and it is effective on applications that are using these types of hosts. Therefore, investigating resource characterization in order to derive efficient factors of host�s availability is needed to improve VC systems. Factors analysis of hosts is fruitful in making better decision for job scheduler in order to improve system�s performance. The purpose of this paper is determination the availability score model for volatile hosts according to statistical analysis of host�s factors from actual occurrence trace data set. This study is proved the relationship across host�s factors then it is proposed Scoring Availability Model. © 2005 - 2014 JATIT & LLS. All rights reserved.