In　open　and　dynamic　computing　environments,　the　entities　that　make　connections　and　interactions　may　know　little　about　each　other　or　without　prior　knowledge.　So,　certain　level　of　trust　must　be　established.　This　may　require　that　an　entity　request　some　information　from　other　entities　that　probably　involves　privacy.　Therefore,　quantifying　privacy　loss　and　trust　gain　is　a　meaningful　and　important　subject.　Although　some　work　has　already　been　done　in　this　area,　it　has　failed　to　consider　the　relationship　between　privacy　information　and　dynamic　trust　variation　during　quantifying　process.　In　this　paper,　we　propose　a　novel　quantifying　metric　for　privacy　protection　based-on　information　theory　in　which　we　investigate　dynamic　trust　variation　to　lower　privacy　loss　while　achieving　more　trust　gain　when　exchange　information.　Simulation　results　show　that　our　metric　can　achieve　the　goal　well.